N-Acyl-phosphoramidates as potential novel form of gemcitabine prodrugs

Article abstract of DOI:10.1016/j.bmc.2014.02.034

Gemcitabine (dFdC) is a cytidine analog remarkably active against a wide range of solid tumors. Inside a cell, gemcitabine is phosphorylated by deoxycytidine kinase to yield gemcitabine monophosphate, further converted to gemcitabine di- and triphosphate. The most frequent form of acquired resistance to gemcitabine in vitro is the deoxycytidine kinase deficiency. Thus, proper prodrugs carrying the 5′-pdFdC moiety may help to overcome this problem. A series of new derivatives of gemcitabine possessing N-acyl(thio)phosphoramidate moieties were prepared and their cytotoxic properties were determined. N-Acyl-phosphoramidate derivatives of gemcitabine have similar cytotoxicity as gemcitabine itself, and have been found accessible to the cellular enzymes. The nicotinic carboxamide derivative of gemcitabine 5′-O-phosphorothioate occurred to be the best inhibitor of bacterial DNA polymerase I and human DNA polymerase α.

Full text of DOI:10.1016/j.bmc.2014.02.034

Bioorganic & Medicinal Chemistry 22 (2014) 2133–2140
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
N-Acyl-phosphoramidates as potential novel form of gemcitabine
prodrugs
Janina Baraniak , Aleksandra Pietkiewicz, Renata Kaczmarek, Ewa Radzikowska, Katarzyna Kulik,
⇑
Karolina Krolewska, Marcin Cieslak, Agnieszka Krakowiak, Barbara Nawrot
Department of Bioorganic Chemistry, Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 112, 90 363 Lodz, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Gemcitabine (dFdC) is a cytidine analog remarkably active against a wide range of solid tumors. Inside a
cell, gemcitabine is phosphorylated by deoxycytidine kinase to yield gemcitabine monophosphate,
further converted to gemcitabine di- and triphosphate. The most frequent form of acquired resistance
to gemcitabine in vitro is the deoxycytidine kinase deficiency. Thus, proper prodrugs carrying the 50-
pdFdC moiety may help to overcome this problem. A series of new derivatives of gemcitabine possessing
N-acyl(thio)phosphoramidate moieties were prepared and their cytotoxic properties were determined.
N-Acyl-phosphoramidate derivatives of gemcitabine have similar cytotoxicity as gemcitabine itself,
and have been found accessible to the cellular enzymes. The nicotinic carboxamide derivative of gemcit-
abine 50-O-phosphorothioate occurred to be the best inhibitor of bacterial DNA polymerase I and human
Received 31 October 2013
Revised 13 February 2014
Accepted 18 February 2014
Available online 3 March 2014
Keywords:
Gemcitabine
Prodrug
Cytotoxicity
Anticancer therapy
DNA polymerase inhibitor
DNA polymerase a.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
process for antiviral activity, pdFdC is thought to be produced by
the carboxyesterase-mediated formation of a carboxylate nucleo-
phile that intramolecularly displaces the O-aryl substituent.5 The
subsequent P–N bond cleavage may be catalyzed by so far unspec-
ified enzyme with the phosphoramidase activity such as histidine
triad nucleotide binding protein (Hint-1).5 Hint-1 was found ubiq-
uitous in all organisms. It exerts phosphoramidase activity towards
Gemcitabine (20,20-difluoro-20-deoxycytidine, dFdC, Fig. 1A) ex-
erts a spectrum of antitumor activity against a wide range of solid
tumors, including pancreatic, non-small cell lung, breast and ovar-
ian cancers.1 Inside a cell, gemcitabine is phosphorylated by deox-
ycytidine kinase to yield gemcitabine monophosphate (pdFdC),
which is, in a stepwise manner, converted to gemcitabine di- and
triphosphate (dFdCDP and dFdCTP, respectively). The first phos-
phorylation, catalyzed by deoxycytidine kinase, is the rate-limiting
step. The most frequently described form of acquired resistance to
gemcitabine in vitro is deoxycytidine kinase deficiency.2 Therefore,
one of the possible ways to enhance its activity is to deliver appro-
priately protected gemcitabine 50-O-phosphate to the cell.
There are only a few reports available on pronucleotide deriva-
tives of 50-pdFdC and processes of their unmasking. O-Aryl N-alan-
inyl derivative of pdFdC (Fig. 1B) represents a recent extension of
the prodrug strategy that has been extensively investigated for
antiviral agents such as 30-azido-30-deoxythymidine (AZT).3 In
mouse xenograft models of human prostate and colon cancers
the O-phenyl-N-(benzyloxy-L-alaninyl) derivative of gemcitabine
is significantly more potent and less toxic than gemcitabine itself.4
By analogy to what has been assumed as the metabolic activation
several synthetic phosphoramidate substrates, such as AMP-N-e-
lysine, AMP-N-alanine, AMP-N-morpholine and adenosine-50-
phosphoramidate (NH2-pA).6 Recently, it has been demonstrated
that Hint-1-assisted hydrolysis of the P–N bond in P-chiral nucle-
oside 50-O-phosphoramidothioate occurs with retention of config-
uration, what is consistent with the formation of an intermediate
covalent enzyme–substrate complex.7 Hint-1 is suggested to be
responsible for activation of nucleotide prodrugs such as mono-
and diesters of nucleoside phosphoramidates derived from amino
acids.3,8,9
An O-alkyl-N-methyl-N-(4-chlorobutyl)phosphoramidate pro-
drug of gemcitabine (Fig. 1C) has been described by Borch and
co-workers.10 For this prodrug, the metabolic activation involves
intracellular generation of an unstable phosphoramidate anionic
intermediate, which undergoes spontaneous cyclization accompa-
nied by P–N bond cleavage. However, the P–N bond cleavage by an
unspecified phosphoramidase is alternatively considered. The
phospholipid gemcitabine conjugate (Fig. 1D) represents a third
class of 50-pdFdC prodrug. This conjugate is thought to undergo
the hydrolytic unmasking by phospholipase C.11 At physiological
⇑
Corresponding author. Tel.: +48 42 681 6970; fax: +48 42 681 5483.
E-mail address: bnawrot@cbmm.lodz.pl (B. Nawrot).
Equal contribution.
http://dx.doi.org/10.1016/j.bmc.2014.02.034
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

2134
J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
We have recently described a new approach for preparation
of O-alkyl-N-acyl-phosphoramidates and thiophosphoramidates12
based on the oxathiaphospholane chemistry.13 To the best of our
knowledge, 50-O-(N-acyl-phosphoramidate) derivatives of gemcit-
abine have not been so far tested as potential prodrugs, while rel-
atively low stability of the P–N bond in cellular media makes them
interesting in that context. Our goal was to synthesize and evaluate
anticancer activity of several conjugates of 50-O-(thio)phosphory-
lated gemcitabine with representative carboxamides and to check
if these compounds are hydrolyzed by the HINT-1 protein.
2. Results and discussion
2.1. Chemistry
We have prepared twelve new derivatives of gemcitabine, con-
sisting of conjugates of 50-O-thiophosphorylated and 50-O-phos-
phorylated dFdC with alkyl, aryl or amino acid carboxamides
(5a–f and 6a–f, respectively, Scheme 1). The selected amides
included benzamide (a), nicotinamide (b), isonicotinamide (c),
acetamide (d), N-acetylprolylamide (e), and N-acetylphenylalani-
nylamide (f). The syntheses were performed by condensation of
2-chloro-1,3,2-oxathiaphospholane (2) with a carboxamide (1a–f)
and subsequent sulfurization of the resulting phosphoramidite
with elemental sulfur. The resulting N-(2-thiono-1,3,2-oxathia-
phospholane) derivatives 3a–f were obtained in 40–62% yield
(Table 1) and their structures were confirmed by 31P NMR and
mass spectrometry analysis. These compounds are remarkably
stable and can be stored at ambient temperature for a long
time (>1 year). Their DBU-assisted nucleophilic ring-opening
condensation with N-4,O-30-dibenzoylgemcitabine (4) furnished
phosphorothioate derivatives, further debenzoylated with concen-
trated ammonia. The resulting derivatives 5a–f were isolated by
ion-exchange column chromatography in 27–92% yield, and
characterized by 31P NMR and FAB-MS. Since compounds 5 were
Figure 1. Gemcitabine and its prodrugs.
pH, this gemcitabine derivative is a phosphate monoanion, and
therefore, differs from the uncharged O-ester-N-alkylphosphoram-
idate derivatives shown above. The lipid moiety does not enhance
the cellular uptake of such monoanions as evidenced by the uptake
of monoanionic amino acid phosphoramidates.5
Scheme 1. Synthesis of N-acyl-thiophosphoramidate (5) and N-acyl-phosphoramidate (6) derivatives of gemcitabine. Reagents: (a) S8, pyridine; (b) DBU, CH3CN; (c) NH4OH;
(d) PhlO2.

J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
2135
Table 1
The physico-chemical data and yields of isolated compounds 3a–f, 5a–f and 6a–f
3
5
6
b
b
Yield(%) 31P NMR (d, ppm)a FAB-MS (Mꢀ1) (m/z) Yield (%) 31P NMR (d, ppm)
FAB-MS (Mꢀ1) (m/z) Yield (%) 31P NMR (d, ppm)
FAB-MS (Mꢀ1) (m/z)
a
b
c
d
e
f
62
42
46
48
51
58
88.78
89.99
89.93
88.46
88.63, 88.45
88.57, 88.59
258
259
259
196
293
343
88
80
27
92
40
42
48.42, 47.90
48.51, 48.26
47.97, 47.15
47.39, 47.17
47.96, 47.20
47.86, 47.14
461
462
462
399
496
546
52
29
21
79
35
12
ꢀ4.21
ꢀ4.65
ꢀ4.90
ꢀ4.96
ꢀ5.12
ꢀ5.26
445
446
446
383
480
530
a
In CD3OD.
In D2O.
b
obtained as mixtures of diastereomers, their 31P NMR spectra con-
tained two resonance signals representing both diastereomers.
Compounds 5a–f were oxidized with iodoxybenzene to yield 6a–
f. The desired compounds 6a–f were isolated by ion-exchange
chromatography (12–79% yield) and their structures were con-
firmed by 31P NMR and FAB-MS analysis (Table 1). N-acyl-phos-
phoramidates 5a,b,d,e and N-acyl-thiophosphoramidates 6a–f of
gemcitabine were used for further studies.
Table 2
The toxicity of compounds 5a,b,d,e and 6a–f (IC50
leukaemia (K562) cells, non-cancerous human umbilical vein endothelial (HUVEC)
cells, human cervix carcinoma (HeLa) cells and human pancreatic adenocarcinoma
(CFPAC) cells, determined by the MTT assay
#
)
towards chronic myelogenous
Compound
IC50
(
l
M)
K562
48 h
HUVEC
48 h
HeLa
CFPAC
72 h
72 h
96 h
96 h
Gemcitabine
0.05
800
>1000
500
>1000
0.07
>1000
0.1
0.003
>1000
>1000
>1000
800
0.05
50
0.05
0.09
10
10
900
0.9
650
500
700
450
2
1.4
0.9
1
0.47
425
77.5
525
350
0.47
0.42
0.62
0.57
0.62
0.42
0.35
315
62.5
475
310
0.32
0.40
0.45
0.42
0.55
0.45
5a
5b
5d
5e
6a
6b
6c
6d
6e
6f
2.2. Cytotoxicity studies
>1000
>1000
500
>1000
10
10
15
80
>1000
To check the prodrug properties of N-acyl-phosphoramidate
and N-acyl-thiophosphoramidate derivatives of gemcitabine, the
cytotoxicity of 5a,b,d,e and 6a–f towards three cancer cell lines—
human cervix carcinoma (HeLa), chronic myelogenous leukaemia
(K562) and human pancreatic adenocarcinoma (CFPAC) cell lines
was determined. The CFPAC cell line was selected because the cells
are in vitro sensitive to gemcitabine, and patients with pancreatic
cancer are successfully cured with this drug.14 Although gemcita-
bine is a standard chemotherapeutic agent for pancreatic cancer,
recent data demonstrate (as examined under in vivo conditions)
that cells of a side population of human pancreatic ductal adeno-
carcinoma are more resistant to gemcitabine than other tumor
cells.15 Human umbilical vein endothelial cells (HUVEC, isolated
from freshly collected umbilical cords) were used as control,
non-cancerous cells. The cells were exposed to the test compounds
for 24, 48, 72 and 96 hours, and the cytotoxicity was determined
by the MTT assay as described previously.16 The viability of
untreated cells was taken as 100%. The toxicity of the test com-
pounds was compared to the toxicity of gemcitabine, determined
in the same MTT test. The IC50 values (IC50 is a concentration of a
test compound required to reduce the cell survival fraction by
50%) were calculated from the corresponding dose-response curves.
At first, we screened the toxicity of all test compounds towards
four cell lines at shorter incubation times (24 and 48 h). Under
these conditions, the gemcitabine N-acyl-thiophosphoramidates
5a,b,d,e have shown no cytotoxic effect. These PS-compounds
occurred to be either non-toxic or their toxicity was rather poor
0.5
30
80
0.8
45
200
#
The IC50 values correspond to the concentration (in
lM) of the tested com-
pounds required to reduce the cell survival by 50%. Untreated cells were used as a
control. The IC50 values were calculated from the dose-response curves (mean
values from 2 independent experiments performed in triplicate).
cells with phosphoramidates 6a–f resulted in strong toxic effects
(Fig. S1, a lower panel), with the IC50 values similar to that for gem-
citabine (IC50 ca. 0.5
dates 5a–b, 5d–e occurred much less toxic, except for 5b, for
which IC50 values of 77.5 and 62.5 M were determined after 72-
and 96 h-treatment, respectively. The remaining thiophosphoram-
idates 5a, 5d and 5e showed modest toxicity towards CFPAC cells
and only at high concentrations (IC50 > 310 lM). The 96 h-expo-
sure of HeLa cells revealed that 6a–e exert cytotoxicity comparable
lM, Table 2). In contrast, thiophosphorami-
l
to that of gemcitabine, while thiophosphoramidates 5a–b, 5d–e
were non-toxic (IC50 > 450 lM). Thus, one may conclude that in
the cancer (K562, HeLa, CFPAC) and normal (HUVEC) cells all four
screened thiophosphoramidates 5 do not serve as gemcitabine pro-
drugs. In contrast, the toxicity of gemcitabine phosphoramidates
6a–f (similar to that of gemcitabine) supports the prodrug pathway
of their action, at least in CFPAC cells. The time profiles of toxicity
of 6 (Fig. S1, lower panels) suggest that these compounds require
rather long time for ‘unmasking’, but afterwards their activities
are virtually identical.
(IC50 > 500 lM) (Table 2 for K562 and HUVEC, Figure S1 for CFPAC,
data not shown for HeLa). Moreover, the HeLa and CFPAC cell lines
were insensitive to gemcitabine N-acyl-phosphoramidates 6a–f
(IC50 > 600
10 M. In contrast, derivatives 6a, 6c and 6d were toxic to the
K562 cells in the short incubation time experiments (IC50 < 0.5 M,
lM), except for 6c, for which IC50 in HeLa cells was
l
2.3. Influence of tested compounds on the activity of selected
DNA polymerases
l
Table 2). Interestingly, in HUVEC cells screened, these compounds
were much less toxic compared to gemcitabine alone. For example,
6a and 6c were 17 times less toxic than gemcitabine, while 6d—30
times.
Next, we checked the effects of longer exposure (72 and 96 h)
on the survival of CFPAC and HeLa cells upon treatment with the
compounds screened. We found that extended treatment of CFPAC
The antitumor effect of gemcitabine may result from inhibition
of ribonucleotide reductase by gemcitabine diphosphate. Such
inhibition may reduce the level of deoxyribonucleoside triphos-
phates required for DNA synthesis. Alternatively, gemcitabine
triphosphate may inhibit DNA polymerase and, in that way, may

2136
J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
Elongaꢀon product (%)
slow down DNA synthesis and repair. Incorporation of the gemcit-
abine unit into a DNA chain results in inhibition of DNA synthesis.
We analyzed the influence of compounds 5a,b,d,e and 6a–f on the
DNA elongation reaction catalyzed by the Klenow fragment
(exo+/ꢀ) of Escherichia coli DNA polymerase I, and by the
genetically engineered E. coli DNA polymerase variant D141A/
E143A/A485L (Therminator™). The latter enzyme is known for
enhanced ability to incorporate modified substrates, such as
dideoxynucleotides, ribonucleotides and acyclonucleotides, into
the growing DNA chain.17,18 In addition, human DNA polymerase
KF+
mM
a
(hPol a
) was also tested.19
2.3.1. Inhibitory activity
Elongaꢀon product (%)
In one set of experiments, a single base extension reaction (SBE)
was catalyzed either by the Klenow fragment of DNA polymerase I
(with or without 30 ? 50 exonucleolytic activity, KF+ or KFꢀ, respec-
KF-
tively), by the Therminator™ polymerase or by hPol
a. In this
assay, the compounds 5a,b,d,e and 6a–f were used as the
competitors to 20-deoxycitidine 50-triphosphate (dCTP). In a typical
procedure, the tested compound and dCTP were added simulta-
neously to the mixture of the template oligonucleotide (TG)
50-terminated with a guanosine unit, and a complementary primer
oligonucleotide (P) 50-labeled with a radioactive 32P-phosphate
group. For KF+, KFꢀ and Therminator™ polymerases, the tested
compounds were used at 2.0 mM concentration, i.e. at 5-fold molar
excess over 0.4 mM dCTP. The same ratio was maintained in the
mM
Elongaꢀon product (%)
120
experiments with hPol
a (0.5 mM concentration of the inhibitors
and 0.1 mM dCTP). The reaction mixtures were analyzed by PAGE
100
80
60
40
20
0
(Fig. S2). The relative activities of the KF+, KFꢀ, Therminator™
hPol α
and hPol
a polymerases were described as a fraction of the
obtained elongation product. None of the screened gemcitabine
derivatives inhibited the Therminator™ polymerase (100% of the
expected single-base extended product was seen). In the experi-
ments with KF+/ꢀ polymerases, 5d, 6a, 6c and 6f did not show
any inhibitory activity. Only 5a, 5b, 6b, 6d and 6e stopped the
polymerization reactions. Compound 5e to a certain extent slowed
down the KFꢀ polymerase, while did not inhibit the two other bac-
µM
500
0
100
200
5b
300
6d
400
6e
5a
6b
terial enzymes. In experiments with hPol a, compounds 5a, 5b, 6b,
6d, 6e also demonstrated inhibitory properties while 6a, 6c, 5e, 6f
did not. Thus, the results obtained for bacterial Klenow and human
DNA polymerases are very similar, contrary to the base-modified
gemcitabine triphosphate analogs, which were good inhibitors
Figure 2. Dose-response inhibition of the elongation activity of the Klenow
fragment of DNA polymerase
from E. coli with (KF+) and without 30 ? 50
exonucleolytic properties (KFꢀ), and human DNA polymerase
, by 5a, 5b, 6b, 6d
and 6e. The X axis—a molar concentration of the test compound (up to 2.0 mM); the
Y axis—the amount of the obtained elongation product (%). The SBE reaction with
dCTP and without any inhibitor was used as a control (100% yield).
I
a
only towards hPol a 19
.
For the compounds exhibiting inhibitory activity against the
KF+/KFꢀ polymerases and hPol
response curves were produced (Fig. 2). In a typical procedure,
5a, 5b, 6b, 6d and 6e were used in 1- to 5-fold molar excess over
dCTP (from 0.4 to 2.0 mM for KF+/KFꢀ and from 0.1 to 0.5 mM for
a
in the SBE reaction, the dose-
Table 3
Inhibitory activity of 5a,b and 6b,d,e derivatives of gemcitabine towards the Klenow
fragment of DNA polymerase
I
from E. coli with (KF+) and without 3’ ? 5’
exonucleolytic properties (KFꢀ), and human DNA polymerase
base extension reaction
a (hPol a) in the single
hPol a). The products were then analyzed by PAGE (20%/7 M urea)
and the amount of the n+1 product was quantified. The data indi-
cate (Fig. 2) that among the tested compounds the thiophosph-
oramidate derivative 5b is the most potent competitive inhibitor
Entry
Compound
KF+
KFꢀ
hPol a
EC50 (mM)
in experiments with KF+, KFꢀ enzymes and hPol
a. Besides, at
1
2
3
4
5
6
NA-pdA
5a
5b
6b
6d
0.54
1.03
0.28
0.75
0.71
0.68
0.49
0.99
0.13
0.52
0.51
1.16
—
0.35
0.08
0.25
0.07
0.41
ca. 2:1 ratio over dCTP compounds 6b and 6d alter significantly
the activity of the two bacterial polymerases. In experiments
with hPol a, also 6d showed inhibitory activity as good as 5b
(EC50 < 0.1 mM, Table 3). It might be plausible that the derivatives
6 (the PO-series) under cellular conditions act as prodrugs of gem-
citabine being unmasked and phosphorylated to dFdCTP, and do
6e
The dCTP substrate was used at 0.4 mM concentration for experiments with KF+ and
KFꢀ and at 0.1 mM with hPol
a
.
not inhibit nuclear human DNA polymerase
a. Since 5b is a nico-
tinic amide derivative of gemcitabine 50-O-phosphorothioate, we
checked an influence of gemcitabine 50-triphosphate (dFdCTP)
and the conjugate of nicotinic amide with 50-dAMP (NA-pdA) on
the progress of the SBE reaction. dFdCTP occurred to be a good
substrate for the tested DNA polymerases, while NA-pdA exhibited
inhibitory activity towards the KF+ and KFꢀ polymerases at
effective concentration (EC50, a concentration that causes 50% inhi-
bition) of 0.54 and 0.49 mM, respectively (Table 3).
As mentioned before, 5b exhibits the lowest EC50, which is two
times smaller than for the NA-pdA/KF+ system, and ca. 4-times
lower for NA-pdA/KFꢀ. Stronger inhibitory activity of 5b in

J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
2137
P
TG
C
6a 5a 6b 5b
C
6e 5e 6d 5d
C
6c 6f 6g
comparison to 5a allows to hypothesize that, both, nicotinic amide
and phosphorothioate function, play important roles in the interac-
n+1
n
n-1
tions of 5b with the Klenow (exo+/ꢀ) fragments and hPol
a. This
idea is supported by slightly reduced activity of a gemcitabine
derivative with nicotinic carboxamidophosphate residue (6b),
which differs from 5b by the absence of the sulfur ligand at the
phosphate moiety. Notably, the lack of a nitrogen atom in the aro-
matic ring of the carboxamide substituent (as in 6a), or its changed
position (as in the isonicotinic residue of 6c), make these deriva-
tives completely inactive towards the screened polymerases.
For the compounds other than 5a, 5b, 6b, 6d and 6e the EC50
values were assessed as >2 mM for bacterial polymerases and
Figure 3. The electrophoretic analysis of the single base extension reaction of the
50-32P-labeled primer P carried out with 5a,b, 5d,e and 6a–f and the KF+ enzyme.
Reactions were carried on using the TG template, 50-terminated with the guanosine
unit. The dCTP substrate was used as a control (lane C). The products n+1 were
expected, while the products nꢀ1, nꢀ2 etc. resulted from the 30 ? 50 exonucleolytic
activity of KF+. The arrows point to the lines where no shorter products were found.
>0.5 mM for hPol
The observed in vitro good inhibitory properties of 5b towards
bacterial polymerases and human DNA polymerase do not bring
any significant effect in human cells, as seen in cytotoxicity studies
(Table 2). Thus, cytotoxicity of 5b may be attributed to the action
a, and are not listed.
Despite of the large excess of the rHint-1 enzyme and long incu-
bation time, all tested thiophosphoramidates 5 and phosphorami-
a
dates
6 remained unhydrolyzed. Also compounds 5a,b,d,e
remained unchanged by the HeLa lysates. In contrast, the com-
pounds 6a–f (the phosphoramidate series) were partially digested
by the HeLa cellular enzymes. After 22 h of incubation, the best
substrate 6f was degraded in 29.0%, while 6b in 4.5% only (Table 4).
Interestingly, in the presence of the K562 cell lysate all tested com-
pounds were partially degraded (93.4–67.44% of intact substrates)
with the exception for N-acetylphenylalanyl-phosphoramidate
derivative 6f, which was digested in more that 50% (45.9%
remaining).
In the screened reaction mixtures, the intact substrates were
accompanied by several digestion products. Three of them, i.e.
gemcitabine (dFdC, Rt = 11.8 min), gemcitabine 50-O-phosphate
(pdFdC, 11.1 min) and 50-O-phosphoramidate (NH2-pdFdC,
10.6 min) were identified by co-injection with original samples.
As an example, an HPLC profile of the products of digestion of 6f
with the K562 lysate is shown in Figure S3.
as hPol
a inhibitor (in the intact form), while the compounds of
the series 5 do not work as prodrug equivalents of gemcitabine.
2.3.2. Substrate activity
In 2007, Herdewijn et al. reported that conjugates of pdA with
selected natural amino acids mimic the natural nucleoside triphos-
phates and are accepted by certain DNA polymerases.20 In this
case, the amino acid moiety linked to nucleoside 50-phosphate
through the P–N bond serves as a leaving group in the nucleotidyl
transfer reaction. Especially, the 50-aspartyl- and 50-histidyl-phos-
phoramidates of 20-deoxyadenosine were efficient substrates for
elongation of the DNA chain by reverse transcriptase of the
HIV-1 and the Therminator™ DNA polymerase. Therefore, having
the carboxamide conjugates 5a,b,d,e and 6a–f, a series of the SBE
assays was performed as described above, yet without dCTP added.
The experiments have shown that none of the tested compounds
served as a substrate for Therminator™ neither for KF+/KFꢀ
polymerases (no n+1 products were obtained) (see Fig. 3 for KF+
polymerase). Interestingly, 5b, 6d and 6e inhibited (to a certain ex-
tent) the 30 ? 50 exonucleolytic activity of the KF+ polymerase (the
lanes, in which the shorter products are not present, marked with
the arrows). This result is compatible with the inhibitory activity of
5b towards the Klenow fragments KF+ and KFꢀ. Thus, among the
tested compounds, the thiophosphoramidate derivative 5b is the
most potent inhibitor of the KF+ and KFꢀ enzymes, and it blocks
efficiently the 30 ? 50 exonucleolytic activity of KF+. This activity
might be responsible for higher cytotoxicity of the nicotinic car-
boxamide derivative of 50-O-phosphorothioate of gemcitabine
Unexpectedly, the addition of a phosphatase inhibitor to the
reaction mixture of 6f and K562 lysate resulted in a more complex
mixture of products. However, the amount of gemcitabine in the
reaction mixture was smaller than the amount remaining in the
reaction carried out without the inhibitor (19.6% vs 48.5%)
(Fig. S4). This result indicates that gemcitabine 50-O-phosphate is
the primary product of the degradation of the prodrugs. This con-
clusion is further supported by an observation that the treatment
of gemcitabine 50-phosphate with K562 lysate resulted in its par-
tial dephosphorylation (Fig. S5). Similarly, digestion of gemcitabine
50-phosphoramidate furnished the dephosphorylated product
(gemcitabine) and gemcitabine 50-monophosphate (Fig. S6). It is
likely that hydrolysis of NH2-pdFdC to pdFdC in the lysate is cata-
lyzed by the HINT-1 enzyme, especially that such the transforma-
tion was observed in vitro upon action of the recombinant HINT-1
protein. The hydrolytic activity of rHint-1 towards NH2-pdFdC was
(5b) compared to other thiophosphoramidate derivatives
5
screened in the SBE assay.
assigned as 1.88 0.17 nmol minꢀ1
l
gꢀ1 for 1 mM substrate, or
2.4. Stability of prodrugs in aqueous solutions, pH 3-11
0.88 0.08 nmol minꢀ1
l
gꢀ1 for 0.2 mM substrate (Fig. S7), which
The stability of representative compounds 5b, 5d, 6b and 6d
(20 mM) was checked in 0.1 M ammonium acetate buffered solu-
tions (pH 3–11), at room temperature, over 2, 8 and 24 h. The sta-
bility was monitored by RP-HPLC on a C18 column. All screened
compounds were found to be stable, even over the prolonged
(24 h) incubation time (data not shown).
Table 4
Digestion of 5a,b,d,e and 6a–f with proteins present in the HeLa and K562 cell lysates
Compound
(1 mM)
% of intact substrate
HeLa protein extract (40 g, K562 protein extract (40 lg,
l
22 h)
22 h)
2.5. Stability of test compounds in the presence of rHint-1
enzyme or lysates of the HeLa and K562 cells
5a
5b
5d
5e
6a
6b
6c
6d
6e
6f
100.0
100.0
100.0
100.0
78.7 2.2
95.5 0.7
76.7 2.9
94.1 2.9
89.6 2.3
71.0 3.8
91.5 0.3
93.4 0.1
88.4 0.3
85.3 0.2
67.4 1.7
86.8 0.5
82.8 1.2
85.1 1.4
80.2 2.4
45.9 1.8
To check whether 5a,b,d,e and 6a–f are substrates for HINT-1
phosphoramidase, the compounds were treated with either recom-
binant Hint-1 protein (rHint-1) or with lysates of the HINT-1
expressing HeLa and K562 cells. The resultant mixtures were ana-
lyzed by RP-HPLC, and the content of the intact substrates was
determined at 260 nm (Table 4).

2138
J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
is a value comparable to the rHint-1 hydrolytic activity towards
NH2-pA (0.2 mM substrate, 1.71 0.25 nmol minꢀ1 gꢀ1).6
(2.5 ꢁ 18 cm). The column was eluted with a gradient of methanol
in chloroform (0 ? 1%). Appropriate fractions were combined and
evaporated under reduced pressure. The yields of 3a–f, their chem-
ical shift values (31P NMR), and FAB-MS data are presented in
Table 1.
l
Although we assumed that the activation of acylphosphorami-
dates is done by Hint-1 phosphoramidase (the cleavage of the
P–N bond), stability of 5 and 6 towards that enzyme suggests that
the unmasking pathway of these prodrugs is different. Since cyto-
toxicity tests showed that the tested compounds reduced the via-
bility of the cells, one can conclude that they undergo some
intracellular enzymatic transformations. It is plausible that intra-
cellular proteases (e.g. numerous cathepsins present mainly in
lysosomes), cleave the C(O)–N bond to yield gemcitabine 50-O-
phosphoramidate, which is subsequently converted by Hint-1 to
gemcitabine 50-phosphate, the desired active intermediate.
Although in cellular extracts we observed further conversion of
pdFdC into dFdC, this process may be less efficient inside the cell,
because phosphatases localized in different cell compartments
may have limited access to the drugs in a cytoplasm. Moreover,
the HeLa and K562 cell extracts used in the stability tests con-
tained the protease inhibitor to prevent the total degradation of
cellular enzymes. If this inhibitor also slowed down the cleavage
of the C(O)–N bond in the prodrugs, the efficiency of the activation
of the screened compounds measured in the extracts might actu-
ally be underestimated.
4.1.2. General procedure for the synthesis of gemcitabine
N-acyl-thiophosphoramidates (5a–f, Scheme 1)
Into a solution of starting 3a–f (1 mmol) dissolved in dry aceto-
nitrile (8 mL), a solution of N-4,O-30-dibenzoyl-gemcytabine (4,
dFdC-2Bz, 1 mmol) in acetonitrile (6 mL) was added, followed by
DBU (1.2 mmol, 184 lL). The mixture was stirred for 24 h at room
temperature for the amino acid carboxamide derivatives, or for
48 h at 40 °C for the other carboxamide derivatives. Then, the sol-
vent was removed under reduced pressure and the residue was
treated with concentrated aq. ammonia (15 mL) to remove the
benzoyl protecting groups. The reaction mixture was stirred for
24 h at room temperature, concentrated in vacuo and compounds
5a–f were isolated using ion-exchange column chromatography
[DEAE-Sephadex A-25; TEAB (0.0 ? 0.1 M; pH 7.5) as the eluent].
4.1.3. General procedure for the synthesis of gemcitabine
N-acyl-phosphoramidates (6a-f)
To a solution of 5a–f (0.1 mmol) in water (1 mL), PhIO2
(0.2 mmol) was added. The suspension was stirred at room tem-
perature until got brown (15–120 min. depending on a compound)
and then was concentrated under reduced pressure. The residue
was dissolved in H2O (5 mL) and washed with chloroform (5 mL).
The aqueous layer was concentrated and the products 6a–f were
isolated using ion-exchange column chromatography [DEAE-
Sephadex A-25; TEAB (0.0 ? 0.1 M, pH 7.5) as the eluent].
20,20-Difluoro-20-deoxycytidine-50-O-[N-(benzoyl)thiophosphoram-
idate (5a): Yield 88%. 31P NMR (D2O) d: 48.42, 47.90. 1H NMR (D2O)
d: 7.62 (d, 1H, H-6), 7.45 (m, 5H, H-Ar), 6.05 (t, 1H, H-10), 5.71 (d,
1H, H-5), 4.35 (m, 4H, H-30, H-40, 2xH-50). FAB-MS m/z: (Mꢀ1) 461.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(3-carbonyl-pyridine)]thi-
ophosphoramidate (5b): Yield 80%. 31P NMR (D2O) d: 48.51, 48.26.
1H NMR (D2O) d: 8.70 (d, 1H, H-6), 8.49 (d, 1H, H-Ar), 8.01 (m,
1H, H-Ar), 7.65 (m, 1H, H-Ar), 7.39 (m, 1H, H-Ar), 6.18 (t, 1H, H-
10), 5.89 (d, 1H, H-5), 4.52 (m, 1H, H-30), 4.11 (m, 1H, H-40), 3.71
(m, 2H, H-50). FAB-MS m/z: (Mꢀ1) 462.
3. Conclusions
The experiments on cytotoxicity of gemcitabine N-acyl-thi-
ophosphoramidates 5 and N-acyl-phosphoramidates 6 indicate
that the compounds of PO-series (derivatives 6) are metabolized
by the cellular enzymes to yield active intermediates and exert
similar cytotoxicity as gemcitabine itself. Thus, these derivatives
can be considered good candidates for gemcitabine prodrugs. Their
most appreciated property is lower cytotoxicity towards non-can-
cerous HUVEC cells, than that of gemcitabine. In contrast, the com-
pounds of PS-series (the derivatives 5), are more resistant to
unmasking and to enzymatic conversion to the active form
(dFdCTP), and their cellular toxicity is lower. The only interesting
property is shown for thiophosphoramidate 5b, which in vitro (in
the intact form) exerts good inhibitory activity towards bacterial
DNA polymerase I and human DNA polymerase
a. This property,
however, if translated into cellular system (as intracellular inhibi-
tion of human DNA polymerases), results in only slight toxic effect
for cancer cells. Therefore, we conclude that N-acyl-thiophosph-
oramidates 5 rather do not serve as gemcitabine prodrugs, while
may exert some inhibitory properties towards enzymes vital for
the cells. Described here gemcitabine N-acyl-phosphoramidates 6
are suitable models for further development of better gemcitabine
prodrugs.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(4-carbonyl-pyridine)]
thiophosphoramidate (5c): Yield 27%. 31P NMR (D2O) d: 47.97,
47.15. 1H NMR (D2O) d: 8.50 (d, 1H, H-6), 7.60 (m, 2H, H-Ar),
7.54 (d, 2H, H-Ar), 6.15 (t, 1H, H-10), 5.70 (d, 1H, H-5), 4.18 (m,
4H, H-30, H-40, 2xH-50). FAB-MS m/z: (Mꢀ1) 462.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(acetyl)]thiophosphorami-
date (5d): Yield 92%. 31P NMR (D2O) d: 47.39, 47.17. 1H NMR (D2O)
d: 7.80 (m, 1H, H-6), 6.15 (t, 1H, H-10), 6.03 (d, 1H, H-5), 4.38 (m,
4H, H-30, H-40, 2xH-50), 1.98 (s, 3H, CH3). FAB-MS m/z: (Mꢀ1) 399.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(
a
-N-acetylprolyl)]
thi-
4. Experimental
ophosphoramidate (5e): Yield 40%. 31P NMR (D2O) d: 47.96, 47.20.
1H NMR (D2O) d: 7.87 (m, 1H, H-6), 6.18 (m, 1H, H-10), 6.03 (m,
1H, H-5), 4.32 (m, 2H, H-30, H-40), 4.11 (m, 2H, H-50), 3.88 (m, 2H,
CH2), 3.53 (m, 2H, CH2), 2.04 (s, 3H, CH3), 1.89 (m, 2H, CH2). FAB-
MS m/z: (Mꢀ1) 496.
4.1. General experimental methods (chemistry)
4.1.1. General procedure for the synthesis of N-(2-thiono-1,3,2-
oxathiaphospholanyl) carboxamides (3a–f)
20,20-Difluoro-20-deoxycytidine-50-O-[N-(
a
-N-acetylphenyl
ala-
To a solution of carboxamide 1a–f (1 mmol) in dry pyridine
(8 mL) containing suspended elemental sulfur (1.5 mmol), neat 2-
chloro-1,3,2-oxathiaphospholane (2)21 (1.1 mmol) was added
dropwise. The reaction mixture was stirred at room temperature
for 12 h (for carboxamide derivatives) or 3 h (for amino acid
carboxamide derivatives). Then the solvent was removed under
reduced pressure. To the residue, acetonitrile (10 mL) was added
and excess sulfur was filtered off. The solvent was evaporated
under reduced pressure, the residue was dissolved in chloroform
(2–3 mL), and the sample was applied on a silica gel column
nyl)]thiophosphoramidate (5f): Yield 42%. 31P NMR (D2O) d: 47.86,
47.14. 1H NMR (D2O) d: 7.73 (d, 1H, H-6), 7.20 (m, 5H, H-Ar),
6.18 (t, 1H, H-10), 6.13 (d, 1H, H-5), 4.30 (m, 1H, H-30), 4.17 (m,
1H, H-40), 4.11 (m, 2H, H-50), 2.98 (m, 2H, CH2), 2.76 (m, 1H, CH),
1.81 (s, 3H, CH3). FAB-MS m/z: (Mꢀ1) 546.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(benzoyl)]phosphorami-
date (6a): Yield 52%. 31P NMR (D2O) d: ꢀ4.21. 1H NMR (D2O) d: 7.63
(d, 1H, H-6), 7.40 (m, 5H, H-Ar), 6.06 (t, 1H, H-10), 5.67 (d, 1H, H-5),
4.41 (m, 4H, H-30, H-40, 2xH-50). FAB-MS m/z: (Mꢀ1) 445.

J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
2139
20,20-Difluoro-20-deoxycytidine-50-O-[N-(3-carbonyl-pyridine)]
phosphoramidate (6b): Yield 29%. 31P NMR (D2O) d: ꢀ4.65. 1H NMR
(D2O) d: 8.76 (d, 1H, H-6), 8.54 (d, 1H, H-Ar), 8.05 (m, 1H, H-Ar),
7.63 (m, 1H, H-Ar), 7.43 (m, 1H, H-Ar), 6.12 (m, 1H, H-10), 5.94
(m, 1H, H-5), 4.49 (m, 1H, H-30), 4.11 (m, 1H, H-40), 3.78 (m, 2H,
H-50). FAB-MS m/z: (Mꢀ1) 446.
4.5. Single nucleotide incorporation catalyzed by Therminator™
polymerase
A mixture (25
(NH4)2SO4, 0.1% Triton X-100, the template and 50-labeled primer
oligonucleotides (4.4 M each), O-alkyl-N-acylphosphor amidates
of gemcitabine (5a–b,5d–e and 6a-f, 400 M, 10 nmoles), and 2U
lL) containing 2 mM MgSO4, 10 mM KCl, 10 mM
l
20,20-Difluoro-20-deoxycytidine-50-O-[N-(4-carbonyl-pyridine)]ph-
osphoramidate (6c): Yield 21%. 31P NMR (D2O) d: ꢀ4.90. 1H NMR
(D2O) d: 8.53 (m, 2H, H-6, H-Ar), 7.61 (m, 2H, H-Ar), 7.44 (d, 1H,
H-Ar), 6.05 (t, 1H, H-10), 5.71 (d, 1H, H-5), 4.28 (m, 4H, H-30, H-40,
2xH-50). FAB-MS m/z: (Mꢀ1) 446.
l
of Therminator™ (BioLabs) in a 20 mM Tris-HCl buffer (pH 8.8)
was incubated for 1 h at 75 °C, and then deep-frozen in liquid
nitrogen to stop the reaction. An analogous reaction was
performed with dCTP as a control. The resulting products were
analyzed by PAGE (20%/7 M urea). The gels were autoradiographed
and the amounts of the products were quantified using a G-box
instrument.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(acetyl)]phosphoramidate
(6d): Yield 79%. 31P NMR (D2O) d: ꢀ4.96. 1H NMR (D2O) d: 7.72 (d,
1H, H-6), 6.17 (t, 1H, H-10), 5.99 (d, 1H, H-5), 4.37 (m, 1H, H-30),
4.27 (m, 1H, H-40), 4.09 (m, 2H, H-50), 1.97 (s, 3H, CH3). FAB-MS
m/z: (Mꢀ1) 383.
4.6. Inhibition of SBE catalyzed by KF+/ꢀ polymerase
20,20-Difluoro-20-deoxycytidine-50-O-[N-(
a-N-acetyl prolyl)]phos-
phoramidate (6e): Yield 35%. 31P NMR (D2O) d: ꢀ5.12. 1H NMR
(D2O) d: 7.71 (m, 1H, H-6), 6.15 (m, 1H, H-10), 6.11 (m, 1H, H-5),
4.28 (m, 2H, H-30, H-40), 4.01 (m, 2H, 2xH-50), 3.49 (m, 2H, CH2),
1.97 (s, 3H, CH3), 1.84 (m, 4H, 2xCH2). FAB-MS m/z: (Mꢀ1) 480.
A reaction mixture (25
beled primer oligonucleotides (4.4
l
L) containing the template and 50-la-
each), dCTP (400 M,
lM
l
10 nmoles), O-alkyl-N-acylphosphoramidate of gemcitabine (5a–
b, 5d–e, 6a–f, 0.4, 0.8, 1.2, 1.6 or 2.0 mM) and KF+/ꢀ (1U) in a
0.05 M Tris–HCl, 5 mM MgCl2, 1 mM DTT buffer (pH 8.0) was incu-
bated for 2 h. The resulting products were analyzed by PAGE (20%/
7 M urea). The gels were autoradiographed and the amounts of the
products were quantified using a G-box instrument.
20,20-Difluoro-20-deoxycytidine-50-O-[N-(
a-N-acetylphenyl alanyl)]
phosphoramidate (6f): Yield 12%. 31P NMR (D2O) d: ꢀ5.26. 1H NMR
(D2O) d: 7.76 (d, 1H, H-6), 7.12 (m, 5H, H-Ar), 6.18 (t, 1H, H-10), 6.03
(d, 1H, H-5), 4.37 (m, 1H, H-30), 4.19 (m, 1H, H-40), 4.01 (m, 2H, H-50),
2.88 (m, 2H, CH2), 2.66 (m, 1H, CH), 1.81 (s, 3H, CH3). FAB-MS m/z:
(Mꢀ1) 530.
4.7. Inhibition of SBE catalyzed by Therminator™ polymerase
A reaction mixture (25
beled primer oligonucleotides (4.4
l
L) containing the template and 50-la-
each), dCTP (400 M,
4.2. Chemical synthesis of oligonucleotides
lM
l
10 nmoles), O-alkyl-N-acylphosphoramidate of gemcitabine (5a–
b, 5d–e, 6a–f, 0.4, 0.8, 1.2, 1.6 or 2.0 mM), and Therminator™
(2U) in a 20 mM Tris-HCl, 2 mM MgSO4, 10 mM KCl, 10 mM
(NH4)2SO4 and 0.1% Triton X-100 buffer (pH 8.8) was incubated
for 1 h. The resulting products were analyzed by PAGE (20%/7 M
urea). The gels were autoradiographed and the amounts of the
products were quantified using a G-box instrument.
The template 50-GGAGGCCATAGCTGTTCCT-30 (TG) and the
primer oligonucleotide 50-AGGAACAGCTATGGCCTC-30 (P) were
synthesized on a Gene World synthesizer (K&A, Germany) using
commercially available phosphoramidite monomers (Glen
Research). The oligonucleotides were deprotected and RP-HPLC
purified by
a
two-step (DMT-on and DMT-off) procedure.22
Molecular mass of the oligomers was confirmed by MALDI-TOF
mass spectrometry (the primer MW 5505, m/z 5501; the template:
MW 5817, m/z 5819).
4.8. Inhibition of SBE catalyzed by human DNA polymerase
a
A reaction mixture (8
50-labeled primer oligonucleotide (1.2
alkyl-N-acylphosphoramidate derivative of gemcitabine (5a,b,
5d–e, 6a–f, at 100, 200, 300 or 500 M concentration) and 2U of
hPol (EURx, Gdan´ sk, Poland) in a buffer (60 mM Tris–HCl (pH
l
L) containing a template (1.9
l
M) and a
l
M), dCTP (100
lM), an O-
4.3. Oligonucleotide labeling
The primer oligonucleotide was labeled at the 50-end with
l
[c-
32P]ATP (Hartmann Analytic) and T4 polynucleotide kinase
a
8.0), 5.0 mM magnesium acetate, 0.3 mg/mL BSA, 1 mM DTT and
0.1 mM spermine) was incubated at 37 °C for 2 h. The reaction
was stopped by addition of formamide with bromophenol blue,
followed by heating at 95 °C for 4 min. The products were analyzed
by PAGE (20%/7 M urea). The gels were autoradiographed and the
(USB). A mixture (20
7.6, 10 mM MgCl2, 10 mM 2-mercaptoethanol), 0.55 nmol of oligo-
nucleotide, 0.032 mCi [
32P]ATP, and T4 polynucleotide kinase
(5U) was incubated for 1 h at 37 °C, then heat denatured and stored
at ꢀ20 °C.
lL) containing buffer (0.05 M Tris–HCl, pH
c
-
amounts of the products were quantified using
instrument.
a G-box
4.4. Single nucleotide incorporation catalyzed by the Klenow
fragment of Pol I from E. coli with and without 30 ? 50
exonucleolytic activity
4.9. Cells and cytotoxicity assay
Human umbilical vein endothelial cells (HUVEC) were isolated
from freshly collected umbilical cords as previously described,23
and cultured in plastic dishes coated with gelatin, in RPMI 1640
medium supplemented with 20% FBS (fetal bovine serum), 90 U/
A mixture (25
and 50-labeled primer oligonucleotides (4.4
acylphosphoramidates of gemcitabine (5a–b, 5d–e and 6a-f,
400
M, 10 nmoles) and E. coli polymerase KF+/ꢀ (Fermentas, 1U)
l
L) containing 5 mM MgCl2, 1 mM DTT, template
l
M each), O-alkyl-N-
l
mL heparin, 150
lg/mL ECGF (Endothelial Cell Growth Factor,
in a 0.05 M Tris-HCl buffer (pH 8.0) was incubated for 2 h at
37 °C, and then heat denatured. An analogous reaction was per-
formed with dCTP as a control. The resulting mixtures were ana-
lyzed by PAGE (20%/7 M urea). The gels were autoradiographed
and the amounts of the products were quantified using a G-box
instrument.
Roche Diagnostics, Mannheim, Germany) and antibiotics (100
l
g/
mL streptomycin and 100 U/mL penicillin). The HeLa (human cer-
vix carcinoma) and K562 (chronic myelogenous leukaemia) cells
were cultured in RPMI 1640 medium (developed by Moore et al.,
1976, Roswell Park Memorial Institute) supplemented with antibi-
otics and 10% fetal calf serum, in a 5% CO2—95% air atmosphere.

2140
J. Baraniak et al. / Bioorg. Med. Chem. 22 (2014) 2133–2140
The CFPAC (human pancreatic adenocarcinoma) were cultured in
RPMI 1640 medium supplemented with antibiotics (100 g/mL
containing 0.5 mM MgCl2, and either rHint-1 (14
lg) or protein ly-
l
sate from HeLa or K562 cells (40 g). The samples were incubated
l
streptomycin, 100 U/mL penicillin) and 10% fetal calf serum in a
for 22 h at 37 °C. The reactions were quenched by cooling on ice,
then the samples were heat denatured and centrifuged. The super-
natants were analyzed by RP-HPLC on a BDS Hypersil C18 column
5% CO2—95% air atmosphere.
The HUVEC, HeLa and K562 cells (in amount 7 ꢁ 103) and CFPAC
cells (104 cells) in 200
lL of the medium were seeded on each well
(5
l
m, 250 ꢁ 4.6 mm; Thermo Electron Corporation) with a gradi-
on a 96-well plate (Nunc). After 24 h, the cells were exposed to the
tested compounds. The stock solutions of the compounds
(100 mM) were prepared in water and added to each well to
achieve the required final concentration (1 mM, 1 ꢁ 10ꢀ1 mM,
1 ꢁ 10ꢀ2 mM, 1 ꢁ 10ꢀ3 mM, 1 ꢁ 10ꢀ4 mM, 1 ꢁ 10ꢀ5 mM and
1 ꢁ 10ꢀ6 mM) in the cell cultures.
ent of acetonitrile in 0.1 M triethylammonium bicarbonate (pH 7.4,
0% ? 8% over 20 min at a flow rate 1 mL/min. Quantification was
performed by electronic integration of the areas of the peaks of
interest. Each experimental point represents the mean SE from
measurements performed in at least duplicate.
All cells cultures were incubated for another 24 or 48 h except
for the CFPAC cells, which were incubated additionally up to 72
and 96 h. The toxicity of all compounds was determined by the
MTT assay which employs 3-(4,5-dimethylthiazol-2-yl)-2,5-diphe-
nyltetrazolium bromide (Sigma, St. Louis, MO). Briefly, after incu-
bation with drugs, the cells were treated with the MTT reagent
for 2 h. MTT-formazan crystals were dissolved in 20% SDS and
50% DMF at pH 4.7, and the absorbance was measured at 570
and 650 nm on an ELISA-PLATE READER (FLUOstar Omega). To
determine the viability of the cells, untreated cells were used as
a control (100% viability). The IC50 values (the concentration of
the test compound required to reduce the cell survival fraction
by 50% of the control) were calculated from the dose-response
curves.
Acknowledgements
The cytotoxicity studies were done in Screening Laboratory of
Department of Bioorganic Chemistry. The authors thank Professor
Wojciech J. Stec for valuable comments and stimulating discus-
sions. Financial support from Ministry of Science and Higher
Education (through the statutory funds of CMMS PAS, Lodz,
Poland) is gratefully acknowledged.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2014.02.034.
References and notes
4.10. Stability of prodrugs in aqueous solutions pH 3-11
1. Moysan, E.; Basttiat, G.; Benoit, J.-P. Mol. Pharm. 2013, 10, 430.
2. Bergman, A. M.; Pinedo, H. M.; Peters, G. J. Drug Resist. Updat. 2002, 5, 19.
3. Cahard, D.; McGuigan, C.; Balzarini, J. Mini-Rev. Med. Chem. 2004, 4, 371.
4. McGuigan, C. Chemical compounds. 2005, WO 2005/012327 A2.
5. Drontle, D. P.; Wagner, C. R. Mini-Rev. Med. Chem. 2004, 4, 409.
6. Bieganowski, P.; Garrison, P. N.; Hodawadekar, S. C.; Faye, G.; Barnes, L. D.;
Brenner, C. J. Biol. Chem. 2002, 277, 10852.
Stability of representative compounds 5b, 5d, 6b and 6d in the
form of 20 mM solutions in aqueous 0.1 M ammonium acetate (pH
3–11, room temperature) was checked after 2, 8 and 24 h incuba-
tion. The stability was monitored by RP-HPLC (a C18 column) in
0.1 M TEAB/acetonitrile solvent system.
7. Krakowiak, A.; Kaczmarek, R.; Baraniak, J.; Wieczorek, M.; Stec, W. J. Chem.
Comm. 2007, 2163.
8. Wagner, C. R.; Iyer, V. V.; McIntee, E. J. Med. Res. Rev. 2000, 20, 417.
9. Zemlicka, J. Biochim. Biophys. Acta 2002, 1587, 276.
10. Weidong, W.; Sigmond, J.; Peters, G. J.; Borch, R. F. J. Med. Chem. 2007, 50, 3743.
11. Alexander, R. L.; Greene, B. T.; Torti, S. V.; Kucera, G. L. Cancer Chemother.
Pharmacol. 2005, 56, 15.
4.11. Stability of prodrugs in the presence of rHint-1, or HeLa or
K562 cells lysates
4.11.1. The protein and cells lysate preparation
´
12. Baraniak, J.; Kaczmarek, R.; Wasilewska, E.; Korczynski, D.; Stec, W. J.
The rabbit Hint1 protein was expressed from
a plasmid
Tetrahedron Lett. 2004, 45, 4269.
13. Stec, W. J.; Grajkowski, A.; Koziolkiewicz, M.; Uznanski, B. Nucleic Acids Res.
1991, 19, 5883.
14. Akada, M.; Crnogorac-Jurcevic, T.; Lattimore, S.; Mahon, P.; Lopes, R.;
Sunamura, M.; Matsuno, S.; Lemoine, N. R. Clin. Cancer Res. 2005, 11, 3094.
15. Van den Broeck, A.; Gremeaux, L.; Topal, B.; Vankelecom, H. BMC Cancer 2012,
12, 354. http://dx.doi.org/10.1186/1471-2407-12-354.
16. Maszewska, M.; Leclaire, J.; Cieslak, M.; Nawrot, B.; Okruszek, A.; Caminade, A.
M.; Majoral, J. P. Oligonucleotides 2003, 13, 193.
pSGA02-HINT24 in E. coli (the BL 21*strain) and purified using
AMP-agarose (Sigma) affinity chromatography, according to a pub-
lished procedure.6 The homogenous enzyme preparation was dia-
lyzed against
a buffer containing 20 mM Tris (pH 7.5) and
150 mM NaCl, then concentrated and stored at ꢀ80 °C.
Total protein lysates from the HeLa and K562 cells (cellular
extracts) were prepared from cell cultures (75 mL flasks) with cells
grown to confluence. Cells were scraped, sonicated in a PBS buffer
containing a mixture of protease inhibitors (CompleteÒ, Roche,
Penzberg, Germany). The samples were centrifuged and the mole-
cules of molecular weight smaller than 1800 Da (e.g. nucleosides,
nucleoside mono-, di- and triphosphates) were removed from the
supernatant using Protein Desalting Spin Columns (Pierce). The
eluates were used as the protein lysates. The total protein amount
was measured by a spectrophotometric method.
17. Gardner, A. F.; Jack, W. E. Nucleic Acids Res. 1999, 27, 2545–2555.
18. Gardner, A. F.; Jack, W. E. Nucleic Acids Res. 2002, 30, 605.
ˇ
ˇ
19. Perlíková, P.; Eberlin, L.; Ménová, P.; Raindlová, V.; Slavetínská, L.; Tlouštová,
E.; Bahador, G.; Lee, Y. J.; Hocek, M. ChemMedChem 2013, 8, 832.
20. Adelfinskaya, O.; Terrazas, M.; Froeyen, M.; Marlière, P.; Nauwelaerts, K.;
Herdewijn, P. Nucleic Acids Res. 2007, 35, 5060.
21. Martynov, I. V.; Kruglyak, Y. L.; Leibovskaya, G. A.; Khromova, Z. I.; Strukov, O.
G. Zh. Obshch. Khim. 1969, 39, 996.
22. Zon, G.; Stec, W. J. Oligonucleotides and Analogues—A Practical Approach; IRL
Press: Oxford, 1991. pp. 87.
23. Jaffe, E. A.; Nachman, R. L.; Becker, C. G.; Minick, C. R. J. Clin. Invest. 1973, 52,
2745.
24. Gilmour, J.; Liang, N.; Lowenstein, J. M. Biochem. J. 1997, 326, 471.
4.11.2. Enzymatic assays
For the enzymatic digestion, 1 mM solutions of the substrates 5
or 6 were prepared in 20 mM Na-HEPES buffer (pH 7.2, 20 lL)