Synthesis and biological studies of novel nucleoside phosphoramidate prodrugs

Article abstract of DOI:10.1021/jm010337r

A novel approach to the intracellular delivery of nucleotides using phosphoramidate-based prodrugs is described. Specifically, we have developed phosphoramidate prodrugs of the anticancer nucleotide 5-fluoro-2′-deoxyuridine-5′monophosphate (FdUMP). These

Full text of DOI:10.1021/jm010337r

J . Med. Chem. 2001, 44, 4475-4480
4475
Syn th esis a n d Biologica l Stu d ies of Novel Nu cleosid e P h osp h or a m id a te
P r od r u gs
Sandra C. Tobias and Richard F. Borch*
Department of Medicinal Chemistry and Molecular Pharmacology and the Cancer Center, Purdue University,
West Lafayette, Indiana 47907
Received J uly 24, 2001
A novel approach to the intracellular delivery of nucleotides using phosphoramidate-based
prodrugs is described. Specifically, we have developed phosphoramidate prodrugs of the
anticancer nucleotide 5-fluoro-2′-deoxyuridine-5′monophosphate (FdUMP). These phosphora-
midate prodrugs contain an ester group that undergoes intracellular activation liberating
phosphoramidate anion, which undergoes spontaneous cyclization and P-N bond cleavage to
yield the nucleoside monophosphate quantitatively. In vitro evaluation of 5-fluoro-2′-deoxyuri-
dine phosphoramidate prodrugs 2a and 3b against L1210 mouse leukemia cells show potent
inhibition of cell growth (IC50 0.5-3 nM). Cell-based thymidylate synthase inhibition studies
show that, in contrast to FUdR, the nitrofuran compound 2a is of comparable potency in wild
type vs thymidine kinase deficient LM cells. This result indicates that the activation of this
novel prodrug occurs via the proposed mechanism of intracellular delivery. However, naph-
thoquinone 3b has an IC50 value for thymidylate synthase inhibition that is comparable to
FUdR in thymidine kinase deficient cells. Further studies revealed that 3b rapidly decomposes
to the nucleotide in cell culture medium, suggesting that the naphthoquinone analogue is not
sufficiently stable to function as a nucleotide prodrug.
In tr od u ction
drugs were found to be stable in media and human
serum, they were ultimately converted to FUdR and not
FdUMP intracellularly. It was found that these pro-
Purine and pyrimidine nucleoside analogues have
shown promise as anticancer agents. Several modified
9
1
drugs were ∼30 000-fold less cytotoxic in thymidine
kinase deficient cells compared to the wild-type cells.
The cyclosaligenyl prodrugs are proposed to hydrolyze
by cleavage of the phenolic ester bond followed by
spontaneous decomposition to yield FdUMP.8 These
prodrugs were found to degrade to FdUMP in medium
and, although they showed inhibitory activity compa-
rable to FUdR in wild-type cell lines, they were ∼200-
fold less potent in thymidine kinase deficient cells.
Hence the inability of the aromatic amino acid phos-
phoramidate prodrugs and cyclosaligenyl prodrugs to
exhibit inhibition of TS activity in intact thymidine
kinase deficient cells suggests that FdUMP is not
delivered intracellularly. Our approach involves the
development of nucleoside phosphoramidate prodrugs
that undergo enzymatic reduction followed by P-N bond
nucleoside inhibitors of thymidylate synthase (TS; EC
2
.1.1.45), a target enzyme for the control of cell growth,
have been developed. However, these compounds re-
quire intracellular metabolism to the nucleotide in order
to exert inhibitory activity.2 Although the nucleoside
may be membrane permeable, it relies on intracellular
phosphorylation to generate the active compound. Vari-
ous mechanisms of resistance to such analogues have
,3
4
developed in in vitro and in vivo experimental systems.
Among these are the increased activity of catabolic
enzymes such as phosphatases and the decreased levels
of anabolic enzymes such as kinases that lead to the
5
decreased accumulation of the active metabolite. These
problems have encouraged the development of prodrugs
that can release the nucleotide inside tumor cells in
order to bypass the need for intracellular phosphor-
ylation.6
-9
1
0,11
cleavage to release the active nucleotide.
These
FdUMP is an anticancer agent that inhibits TS and
is a candidate for the prodrug approach. TS is an
enzyme that catalyzes the formation of thymidine-5′-
monophosphate (TMP) from 2′-deoxyuridine-5′-mono-
phosphoramidate prodrugs contain a delivery group and
a masking group. The delivery group functions as a
substrate for intracellular enzymatic reduction which
2
12
phosphate (dUMP). Aromatic amino acid phosphora-
leads to expulsion of phosphoramidate anion, and the
midate prodrugs and cyclosaligenyl prodrugs of FdUMP
masking group consists of a haloalkylamine group that
undergoes spontaneous cyclization and P-N bond cleav-
age following activation. Previously we reported the
synthesis and biological activity of a series of haloethyl
nucleoside phosphoramidate prodrugs bearing a nitro-
furyl ester, which showed growth inhibition of L1210
8
,9
have been synthesized and studied in vitro.
The
aromatic amino acid nucleotide prodrugs described by
Wagner et al. are converted to the phosphoramidate;
this intermediate undergoes slow conversion to the
nucleotide via P-N bond cleavage by an unknown
phosphoramidase to yield FdUMP. Although the pro-
1
1
cell proliferation in the nanomolar range. However,
further mechanistic studies on model thymidine phos-
phoramidate compounds revealed that, after cyclization
*
To whom correspondence should be addressed. Tel: (765)494-1403.
Fax: (765)494-1414. E-mail: rickb@pharmacy.purdue.edu.
1
0.1021/jm010337r CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/02/2001

4
476 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 25
Tobias and Borch
Sch em e 1
Sch em e 2a
of the haloethyl phosphoramidate anion, nonselective
nucleophilic attack of water at carbon and phosphorus
1
3
of the aziridinium ion intermediate was observed.
Herein we describe the design, synthesis, and in vitro
studies of a modified halobutyl phosphoramidate to
provide novel prodrugs that undergo rapid and quan-
titative conversion to the active nucleotide.
a
Reagents and conditions: (a) 5-nitrofurfuryl alcohol, 1,4-
dimethoxy-2-hydroxymethylnaphthalene or benzyl alcohol, DIEA,
CH2Cl2, -70 °C, 20 min; (b) N-methyl-N-(4-chlorobutyl)amine
hydrochloride, DIEA, CH2Cl2, -70 °C, 20 min; (c) thymidine or
FUdR, pyridine, -45 °C; (d) tert-butyl hydroperoxide, -45 °C to 0
Resu lts a n d Discu ssion
Ch em istr y. The proposed mechanism for the activa-
tion of these new compounds is shown in Scheme 1. It
was anticipated that prodrug A, after undergoing
enzymatic reduction and spontaneous expulsion to give
B, would cyclize to generate intermediate C. Although
by analogy to the haloethyl analogue this intermediate
might undergo attack at carbon or phosphorus by water,
it was hypothesized that nucleophilic attack at carbon
of the pyrrolidinium ring would be disfavored compared
to attack at phosphorus, and formation of the desired
nucleotide E would predominate. With this hypothesis
in mind, model nucleoside analogues 1a -c, 3a , and
FdUMP prodrugs 2a ,b and 3b were synthesized as
shown in Scheme 2. The synthesis was accomplished
in a one-pot method where a highly reactive phospho-
°
C, 30 min; (e) Ce(NH4)2(NO3)6, CH3CN/H2O.
release of the nucleotide after activation of the delivery
group. To determine whether these compounds are
releasing the active nucleotide, chemical activation of
model compounds 1a , 1c, and 3a was studied using 31
P
NMR. Reduction of the nitrofuryl prodrug 1a was
accomplished using sodium dithionite; the 31P NMR
stack plot of the reaction under model physiologic
conditions (0.4 M cacodylate buffer/CH3CN, pH ∼ 7.4,
37 °C) is shown in Figure 1. The spectrum at 0 min
represents 1a in CH3CN immediately prior to activation.
Five minutes after reduction, three new peaks are
present at -14.61, -21.03, and -23.45 ppm (reference
1% triphenyl phosphine oxide in benzene-d6). The peak
at -14.61 ppm corresponds to the phosphoramidate
anion B (Scheme 1). This intermediate presumably
cyclizes with a half-life of 2.0 min to form the pyrroli-
dinium zwitterion intermediate C (tentatively assigned
at -21.03 ppm) that hydrolyzes spontaneously to thy-
midine monophosphate (-23.45 ppm; E) as the exclusive
product. Attack by water at carbon of the pyrrolidinium
intermediate was not observed.
III
rus monochloro intermediate is generated and reacted
with the nucleoside. Phosphorus trichloride is reacted
with the corresponding alcohol in the presence of
diisopropylethylamine followed by reaction with N-meth-
yl-N-(4-chlorobutyl)amine hydrochloride to generate the
monochloro intermediate. This intermediate is reacted
with the nucleoside in situ and then oxidized with tert-
butyl hydroperoxide to yield 1a -c and 2a ,b in yields
ranging from 34 to 71%. Intermediates 1b and 2b are
then converted into the corresponding bioactive pro-
drugs 3a and 3b by oxidation with ceric ammonium
nitrate.
Biologica l Activity: Gr ow th In h ibtion . In vitro
growth inhibitory activity in L1210 mouse leukemia
cells was studied for compounds 2a and 3b. The growth
inhibitory activity of these compounds was compared
to previously synthesized compound 6, 5-fluorouracil
(5-FU), and FUdR (Chart 1). All compounds exhibited
31P NMR Stu d ies. The inhibitory activity of the
FdUMP prodrugs relies on the rapid and quantitative
11

Novel Nucleoside Phosphoramidate Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 25 4477
Ta ble 2. Enzyme Inhibition in L1210, LM, and LM(TK-) Cells
IC50, nMa
compound
L1210
LM
LM (TK-)
(TK-)/TK
2
3
a
b
65
560
230
260
1060
5900
5400
1.9
25
20
10
FUdR
7.9b
a
Cells were treated with drug for 2 h, then pulsed for 1 h with
b
labeled nucleoside. See experimental procedure for details. Report-
ed value ) 7.6 nM.16
En zym e In h ibition . Although our hypothesis of
reduction and P-N bond cleavage to release the active
3
1
nucleotide is supported by P NMR mechanistic studies,
the formation of 5-FU and FUdR following activation
is also possible in this cell-based system. In the case
where 5-FU is generated, it may be converted to FUdR
by pyrimidine nucleoside phosphorylase and then re-
phosphorylated to FdUMP by thymidine kinase.14 To
determine whether the prodrug might be acting via
formation of 5-FU and/or FUdR, a cell line that is
thymidine kinase deficient was used. A tritium release
F igu r e 1. Reaction of phosphoramidate 1a in cacodylate
buffer (ca. 100 mM, pH 7.4, 37 °C). B: phoshoramidate anion
intermediate. C: pyrrolidinium ion intermediate. E: thymi-
dine 5′-monophosphate. Chemical shifts are reported relative
to triphenylphosphine oxide reference. See Results and Dis-
cussion for details.
15
assay developed by Kalman et al. was used to measure
intracellular inhibition of TS activity. In this assay, TS
activity is measured by determining the tritiated water
3
3
that is released from [5- H] 2′-deoxycytidine ([5- H]-
dCyd).
Ch a r t 1
To compare the effects of thymidine kinase on drug
activity, wild type and thymidine kinase deficient LM
cells were used. L1210 cells were also used in order to
asseses the correlation between enzyme inhibition and
growth inhibition. Table 2 shows the results obtained
when compounds 2a , 3b, and FUdR were tested in this
assay. FUdR is ∼20-fold less potent in the thymidine
kinase deficient cells as expected. Prodrug 2a shows
comparable enzyme inhibition in both wild type and
thymidine kinase deficient LM cells, with the IC50 value
in LM (TK-) cells about 2-fold higher when compared
to the LM cells. Surprisingly, compound 3b was com-
parable to FUdR in both the thymidine kinase deficient
cells and the wild-type cells, suggesting that intracel-
lular delivery of nucleotide was not occurring. This led
us to study the stability of these prodrugs in cell culture
media. Stability tests in serum-free media (Fisher’s
media, pH ) 7.4, 37 °C) revealed that the naphtho-
quinone 3a analogue decomposes to the nucleotide with
a half-life of 3.5 min. In contrast, compound 1a is stable
in media with or without serum for >6 days. Presum-
ably, naphthoquinone 3a is rapidly converted to the
nucleotide in media and is dephosphorylated prior to
entry into the cell.
Ta ble 1. Growth Inhibition in L1210 Mouse Leukemia Cells
IC50, nM
a
treatment time , h
compound
2
8
24
7.2
3.1
56
360
4.1
48
2
3
6
5
a
116
44
1027
2500
45
47
2.3
0.63
16
200
0.64
b
16
347
1010
23
b
-FU
FUdR
a
Cells were treated with drug for 2, 8, 24, and 48 h then spun
down, resuspended in fresh drug-free medium, and returned to
the incubator. Cell counts were determined 48 h after the start of
drug treatment. b See ref 11.
potent inhibition of cell proliferation; the results are
summarized in Table 1. Addition of thymidine (5 µM)
resulted in reversal of growth inhibition when compared
to cells that were not treated with thymidine, suggesting
that the prodrugs are generating an inhibitor of thymidy-
late synthase. The halobutyl compounds 2a and 3b are
more potent than the corresponding haloethyl analogue
Con clu sion s
A novel prodrug approach for the intracellular deliv-
ery of nucleotides across the cell membrane has been
developed. The synthesis of nucleoside phosphorami-
dates bearing nitrofuryl and naphthoquinone phospho-
ramidates that would function as prodrugs of FdUMP
have been prepared. Modification of the previously
6
. The increase in potency of compounds 2a and 3b
coincides with the proposed mechanism where exclusive
formation of the nucleotide is observed by 3 P NMR
1
reported haloethyl moiety to a halobutyl moiety pro-
11
(
Figure 1). As expected, compounds 2a and 3b are
vides novel compounds that undergo rapid and quan-
titative conversion to the nucleotide. Compound 2a
shows excellent inhibition of cell proliferation and
thymidylate synthase inhibition, with IC50 values in the
nanomolar range. Most importantly, cell-based thymidy-
significantly more potent than 5-FU. Furthermore, it
is worth noting that the naphthoquinone analogue 3b
exhibited 2-3-fold greater potency than the correspond-
ing nitrofuryl analogue 2a .

4
478 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 25
Tobias and Borch
late synthase inhibition studies in thymidine kinase
deficient cells indicate that the activation of compound
mixture to eppendorf tubes containing 100 µL of 20% activated
charcoal suspended in 4% aqueous perchloric acid (w/v). The
tubes were vigorously vortexed and centrifuged using a min-
icentrifuge. The supernatant (100 µL) was transferred to 2 mL
of scintillation cocktail and counted for 1 min. The data were
analyzed by sigmoidal curve fit of counts per minute vs log-
2
a occurs via the proposed mechanism involving direct
intracellular conversion to FdUMP. However, the naph-
thoquinone analogue is rapidly degraded to the nucleo-
tide in media and is further metabolized presumably
to the nucleoside.
(drug concentration) and the results expressed as the IC50
values. Background counts (<500 CPM) were obtained by
3
quenching a solution of cells and [5- H]dCyd immediately after
Exp er im en ta l Section
pulsing with the isotope. Negative controls were obtained by
treating cells with EtOH only and incubating after addition
of the isotope and then quenching.
Ma ter ia ls a n d Meth od s. NMR spectra were recorded
using a 250 MHz Bruker spectrometer equipped with a 5 mm
5
′-Th ym id yl 5-n it r ofu r fu r yl N-m et h yl-N-(4-ch lor o-
bu tyl) p h osp h or a m id a te (1a ): Phosphorus trichloride (6.19
mL, 2.0 M in CH Cl ) was cooled to -70 °C under argon.
-Nitrofurfuryl alcohol (1.77 g; 12.38 mmol) was dissolved in
anhydrous CH Cl (30 mL) and added slowly followed by the
1
multinuclear probe. H chemical shifts are reported in parts
per million using tetramethylsilane as an internal reference.
3
1
P NMR spectra were obtained using broadband 1H decou-
2
2
5
pling, and chemical shifts are reported in parts per million
2
2
6
using 1% triphenyl phosphine oxide in benzene-d as a coaxial
insert (25.17 ppm relative to 85% phosphoric acid). 3 P NMR
1
dropwise addition of diisopropylethylamine (3.24 mL, 18.58
mmol). The reaction mixture was allowed to stir at -70 °C
for 15 min. N-Methyl-N-(4-chlorobutyl)amine hydrochloride
kinetics carried out at 37 °C were conducted using a Bruker
variable temperature unit. Silica gel grade 60 was used to
carry out all chromatographic purifications. HPLC analysis
was done using a Beckman System Gold equipped with a 168
detector set to 250 nm, a 126 solvent module, and an econo-
sphere C18 column (5 µM, 4 × 250 mm) from Alltech Associ-
ates. Mass spectral analysis was obtained from the mass
spectrometry laboratory at Purdue University, West Lafayette,
IN. All anhydrous reactions were carried out under argon,
using flamed dried flasks, and all organic solvents were
distilled prior to use. Radioactivity was measured using a
Packard Tri-CARB liquid scintillation analyzer model 1900 TR.
EcoLite liquid scintillation cocktail was obtained from ICN.
2 2
(1.96 g, 12.38 mmol) was dissolved in anhydrous CH Cl (30
mL) and added dropwise to the reaction mixture. Diisopropy-
lethylamine (6.47 mL, 37.15 mmol) was added dropwise, and
the reaction mixture was stirred at -70 °C for 25 min.
Thymidine (1.0 g, 4.13 mmol) was coevaporated with anhy-
drous pyridine (5 × 30 mL), dissolved in anhydrous pyridine
2 2
(30 mL), and cooled to -45 °C. The CH Cl reaction mixture
was added dropwise to the solution of thymidine in pyridine
until thymidine disappeared by TLC (10% MeOH/CHCl ). After
5 min, the reaction mixture was oxidized with tert-butyl
hydroperoxide (2.4 mL, 4.6 M in decane) at -40 °C and
warmed to 0 °C over 30 min. Saturated NH Cl (15 mL) was
added, the layers were separated, and the aqueous layer was
extracted with CHCl
(5 × 20 mL). The organic layers were
combined, dried over Na SO , and concentrated under reduced
pressure. The crude reaction mixture was flushed through a
plug of silica gel (10% MeOH/CHCl ). Further purification by
chromatography on silica gel (5% MeOH:CHCl ) afforded 1a
1.34 g; 59%) as a light yellow foam. R ) 0.20 (5% MeOH/
): δ 8.54 (m, 1H); 7.36 (s, 1H); 7.29
d, 1H, J ) 3.66 Hz); 6.69 and 6.67 (d, 1H, J ) 3.66 Hz); 6.27
q, 1H); 5.03 (m, 2H); 4.53 (m, 1H); 4.22 (m, 2H); 4.05 (m, 1H);
3
3
3
1
4
P NMR Stu d ies. Kinetic experiments were carried out
1
1
as described previously. Briefly, the compound (15 mg) was
dissolved in acetonitrile (80 µL); to this was added a solution
of the activating agent (sodium dithionite; 3 equiv) in cacody-
late buffer (500 µL, 0.4 M, pH ) 7.4). The reaction mixture
was transferred to a 5 mm NMR tube, and data acquisition
was started with the probe maintained at 37 °C. Spectra were
acquired every 2.5 min for 30 min and when necessary every
3
2
4
3
3
(
f
1
3 3
CHCl ). H NMR (CDCl
1
0 min for an additional hour. Time points were assigned to
(
(
each data acquisition from the start of the reaction. The
integration of the peak areas was used to determine the
relative concentration of the intermediates and products.
In Vitr o Gr ow th In h ibition Assa y. Stock solutions of
drugs were prepared in absolute ethanol, and serial dilutions
of drug were prepared. L1210 cells were suspended in Fisher’s
medium supplemented with 10% horse serum, 1% glutamine,
and 1% antibiotic-antimycotic solution to give 10 mL volumes
3
1
.57 (t, 2H); 3.09 (m, 2H); 2.67 (d, 3H, J ) 10.17 Hz); 2.61 (m,
31
H); 2.25 (m, 1H), 1.89 (d, 3H); 1.73 (m, 4H). P NMR
): δ -13.64, -13.75 (mixture of diastereomers). MS
(
CDCl
ESI): m/z 551 (M + H) .
′-Th ym idyl-2-(1,4-dim eth oxyn aph th yl)m eth yl N-m eth -
3
+
(
5
yl-N-(4-ch lor obu tyl) p h osp h or a m id a te (1b): Phosphora-
midate 1b was prepared from 1,4-dimethoxy-2-hydroxymeth-
ylnaphthalene (180 mg, 0.825 mmol), phosphorus trichloride
4
of cell suspension at a final density of 3-6 × 10 cells/mL.
Appropriate volumes of the drug solution were transferred to
the cell suspensions, and incubation was continued for 2, 8,
2 2
(0.41 mL, 2 M in CH Cl ), N-methyl-N-(4-chlorobutyl)amine
hydrochloride (130 mg, 0.825 mmol), thymidine (100 mg, 0.413
mmol), and tert-butyl hydroperoxide (0.24 mL, 4.6 M in decane)
as described for compound 1a . The reaction mixture was
filtered through a plug of Celite and concentrated under
reduced pressure. Column chromatography of the crude prod-
2
4, or 48 h. The cells were spun down, resuspended in fresh
drug-free medium and returned to the incubator. Final cell
counts were determined 48 h after the start of drug treatment.
The data were analyzed by sigmoidal curve fit of cell count vs
log(drug concentration) and the results expressed as the IC50
uct (1:1 CH
orange foam. R
CDCl ): δ 8.58 (s, 1H); 8.23 (d, 1H, J ) 7.50 Hz); 8.03 (m,
1H); 7.54 (m, 2H); 6.82 (s, 1H); 6.25 (m, 1H); 5.25 (m, 2H);
2 2
Cl /acetone) afforded 1b (156 mg; 61%) as a light
(the drug concentration that inhibits cell growth to 50% of
1
) 0.33 (1:1 CH
2
Cl
2
/acetone).
H NMR
control value).
f
(
3
In Vitr o En zym e In h ibition Assa y. Thymidylate syn-
thase inhibition was determined using the published assay.15
4
2
2
.49 (m, 1H); 4.18 (m, 2H); 3.98 (s, 3H); 3.92 (s, 3H); 3.52 (t,
Briefly, serial dilutions of drug were prepared in absolute
ethanol such that 5 µL of drug solution added to 445 µL of
cell suspension gave the desired final concentration. L1210
cells in exponential growth were suspended in Fischer’s
medium, and LM and LM (TK-) cells were suspended in
H); 3.08 (m, 2H); 2.65 (d, 3H, J ) 9.70 Hz); 2.37 (m, 1H);
3
1
.17 (s, 3H) 1.71 (m, 4H). P NMR (CDCl ): δ -12.93. MS
3
+
(ESI): m/z 626 (M + H) .
5′-Th ym id yl 2-(1,4-n a p h th oqu in on yl)m eth yl N-m eth yl-
Minimum Essential Medium to give a final density of 1.1 ×
N-(4-ch lor obu tyl) p h osp h or a m id a te (3a ): Ceric ammo-
nium nitrate (550 mg, 1.0 mmol) in water (7 mL) was added
dropwise over 15 min to a solution of 1b (250 mg, 0.4 mmol)
7
1
0
cells/mL. Aliquots of the cell suspension (445 µL) were
placed in eppendorf tubes, drug solution (5 µL) was added, and
the cells were incubated at 37 °C for 2 h. Measurement of
thymidylate synthase activity was initiated by the addition of
5
deoxycytidine (0.5 µM final concentration). The reaction
mixture was incubated for 1 h in a shaking water bath at 37
in CH
temperature for 15 min and extracted with CHCl
The combined organic layers were dried over Na
concentrated under reduced pressure. The crude product was
purified by silica gel chromatography (5% MeOH/CHCl ) to
give 3a (210 mg, 90%) as a yellow foam. R ) 0.50 (10% MeOH/
3
CN (7 mL). The reaction mixture was stirred at room
(3 × 10 mL).
SO and
3
3
0 µL of serum free media containing 1 µCi of [5- H]-2′-
2
4
3
°C, and then terminated by transferring 100 µL of the reaction
f

Novel Nucleoside Phosphoramidate Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 25 4479
1
CHCl
(
4
3
). H NMR (CDCl
3
): δ 8.08 (m, 2H); 7.79 (m, 2H); 7.05
was cooled to -78 °C and added to the solution. The pressure
tube was closed and warmed to room temperature, then heated
in an oil bath at 80 °C. The reaction mixture was allowed to
stir at this temperature for 16 h. The tube was cooled, the
pressure was released, and methanol was distilled off to give
a viscous orange oil. This mixture was washed with diethyl
ether (3 × 20 mL), then slowly added to a solution of KOH in
water (30 mL, 1.2 g/mL). The water layer was extracted with
toluene, the toluene was removed under reduced pressure, and
d, 1H, J ) 5.68 Hz); 6.21 (t, 1H); 5.03 (m, 2H); 4.59 (m, 1H);
.26 (m, 2H); 4.05 (m, 1H); 3.58 (t, 2H); 2.73 (d, 3H, J ) 10.07
Hz); 2.27 (m, 2H); 1.91 (d, diastereomers, 3H); 1.59 (m, 4H).
3
1
P NMR (CDCl
mers). MS (ESI): m/z 596 (M + H) .
′-Th ym id yl ben zyl N-m eth yl-N-(4-ch lor obu tyl) p h os-
p h or a m id a te (1c): Phosphoramidate 1c was prepared from
phosphorus trichloride (2.06 mL, 2 M in CH Cl ), benzyl
3
): δ -12.67, -12.99 (mixture of diastereo-
+
5
2
2
the product was purified by distillation to give a clear oil (10.14
alcohol (0.43 mL, 4.13 mmol), N-methyl-N-(4-chlorobutyl)-
amine hydrochloride (648 mg, 4.13 mmol), thymidine (500 mg,
1
g, 50%), bp 60 °C (2 mm). H NMR (CDCl
2
H) .
3
): δ 3.58 (t, 2H);
.6+4 (t, 2H); 2.44 (s, 3H); 1.65 (m, 4H). MS (CI): m/z 104 (M +
2
.06 mmol), and tert-butyl hydroperoxide (1.17 mL, 4.6 M in
decane) as described above for compound 1a . Column chro-
matography of the crude product (5% MeOH/CHCl ) afforded
c (676 mg; 64%) as a white foam. R ) 0.23 (10% MeOH/
): δ 8.41 (m, 1H); 7.40 (s, 5H); 6.29
m, 1H); 5.06 (m, 3H); 4.39 (m, 1H); 4.16 (m, 2H); 4.02 (m,
3
N-Meth yl-N-(4-ch lor obu tyl) a m in e h yd r och lor id e: HCl
gas was bubbled into a stirred solution of N-methyl-4-hydroxy-
butylamine (2.00 g, 19.38 mmol) in CH Cl (10 mL) until the
2 2
solution turned moist pH paper red (pH ) 2). The reaction
mixture was cooled to 0 °C, thionyl chloride (1.41 mL, 19.38
mmol) was added dropwise, and the reaction mixture was
stirred at room temperature overnight. The solvent was
1
f
1
CHCl
(
1
2
(
3
). H NMR (CDCl
3
H); 3.54 (t, 2H); 3.04 (m, 2H); 2.61 (d, 3H, J ) 10.07 Hz);
.39 (m, 1H); 2.12 (m, 1H); 1.84 (d, diastereomers, 3H); 1.60
3
1
m, 4H). P NMR (CDCl
diastereomers). MS (ESI): m/z 538 (M + Na) .
-F lu or o-2′-d eoxyu r id yl 5-n it r ofu r fu r yl N-m et h yl-N-
4-ch lor obu tyl) p h osp h or a m id a te (2a ): Phosphoramidate
a was prepared from 5-nitrofurfuryl alcohol (233 mg, 1.63
mmol), phosphorus trichloride (0.81 mL, 2 M in CH Cl ),
3
): δ -13.27, -13.33 (mixture of
+
removed under reduced pressure to give the product as a white
1
solid (2.90 g, 95%), mp 119-121 °C. H NMR (CDCl
3
): δ 8.79
5
(
2
(s, 1H); 3.59 (t, 2H, J ) 5.95); 2.97 (m, 2H); 2.70 (s, 3H); 2.06
(m, 2H); 1.95 (m, 2H); 1.63 (s, 1H). MS (CI): m/z 122 (M +
+
H) .
2
2
N-methyl-N-(4-chlorobutyl)amine hydrochloride (257 mg, 1.63
mmol), 5-fluoro-2′-deoxyuridine (200 mg; 0.812 mmol), and tert-
butyl hydroperoxide (0.44 mL, 4.6 M in decane) as described
above for compound 1b. Column chromatography of the crude
Ack n ow led gm en t. Support from the National Can-
cer Institute (Grants R01 CA34619 and T32 CA09634)
is gratefully acknowledged. The assistance of Dr. Mari-
etta Harrison and Chris Isaacson with the development
of the tritium assay is also acknowledged. We thank Dr.
Thomas Kalman for helpful discussions.
product (1:1 CH
yellow foam. R
CDCl
2
Cl
) 0.34 (1:1 CH
): δ 9.68 (m, 1H); 7.80 and 7.74 (d, 1H, J ) 6.22 and
.41 Hz); 7.30 (d, 1H, J ) 3.48 Hz); 6.71 (d, 1H, J ) 3.30);
.22 (m, 1H); 5.03 (m, 2H); 4.52 (m, 1H); 4.24 (m, 2H); 4.07
2
/acetone) afforded 2a (155 mg; 34%) as a
1
f
2
Cl
2
/acetone).
H NMR
(
3
6
6
(
m, 1H); 3.57 (m, 2H); 3.07 (m, 2H); 2.68 (d, 3H, J ) 10.25
Refer en ces
31
Hz); 2.49 (m, 1H); 2.18 (m, 1H), 1.74 (m, 4H). P NMR
(
1) Hatse, S.; De Clercq, E.; Balzarini, J . Role of Antimetabolites of
Purine and Pyrimidine Nucleotide Metabolism in Tumor Cell
Differentiation. Biochem. Pharmacol. 1999, 58, 539-555.
(
CDCl
25ClFN
-F lu or o-2′-d eoxyu r id yl 2-(1,4-d im et h oxyn a p h t h yl)-
m eth yl N-m eth yl-N-(4-ch lor obu tyl) p h osp h or a m id a te
2b): Phosphoramidate 2b was prepared from phosphorus
trichloride (0.41 mL, 2 M in CH Cl ), 1,4-dimethoxy-2-hy-
droxymethylnaphthalene (177 mg, 0.823 mmol), N-methyl-N-
4-chlorobutyl)amine hydrochloride (128 mg, 0.812 mmol),
-fluoro-2′-deoxyuridine (100 mg, 0.406 mmol), and tert-butyl
hydroperoxide (0.23 mL, 4.6 M in decane) as described for 1b.
3
): δ -15.59, -15.78 (mixture of diastereomers). Anal.
(C
19
H
4
O
10P) C, H, N.
5
(2) Danenberg, P. V. Thymidylate Synthetase-A Target Enzyme in
Cancer Chemotherapy. Biochim. Biophys. Acta 1977, 473, 73-
9
2.
(
(
3) Maggiora, L.; Chang, C. C. T. C.; Torrence, P. F.; Mertes, M. P.
2
2
5
-Nitro-2′-Deoxyuridine 5′-phosphate. A Mechanism-Based In-
hibitor of Thymidylate Synthetase. J . Am. Chem. Soc. 1981, 103,
3192-3198.
(
5
(4) Mader, R. M.; Muller, M.; Steger, G. Resistance to 5-Fluorouracil.
Gen. Pharmacol. 1998, 31, 661-666.
(
5) Copur, S.; Aiba, K.; Drake, J . C.; Allegra, C. J . Thymidylate
Synthase Gene Amplification in Human Colon Cancer Cell Lines
Resistant to 5-Fluorouracil. Biochem. Pharmacol. 1995, 49,
1419-1426.
Column chromatography of the crude product (1:1 CH
acetone) afforded 2b (181 mg; 71%) as a light orange foam. R
)
2
Cl
2
/
f
1
2 2 3
0.53 (1:1 CH Cl /acetone). H NMR (CDCl ): δ 8.22 (dd, 1H);
8
2
4
2
1
.04 (dd, 1H); 7.74 and 7.66 (d, 1H, J ) 6.23 & 6.39); 7.55 (m,
H); 6.83 (s, 1H); 6.16 (m, 1H); 5.26 (m, 2H); 4.51 (m, 1H);
.23 (m, 2H); 4.05 (m, 1H); 3.99 (s, 3H); 3.93 (s, 3H); 3.50 (m,
H); 3.08 (m, 2H); 2.65 (d, 3H, J ) 9.70 Hz); 2.42 (m, 2H);
(6) Farquhar, D.; Chen, R.; Khan, S. 5′-[4-(Pivaloyloxy)-1,3,2-
dioxaphosphorinan-2-yl]-2′-deoxy-5-fluorouridine: A membrane-
permeating prodrug of 5-fluoro-2′-deoxyuridylic acid (FdUMP).
J . Med. Chem. 1995, 38, 488-495.
(
7) Farquhar, D.; Kuttesch, N. J .; Wilkerson, M. G.; Winkler, T.
Synthesis and biological evaluation of neutral derivatives of
5-fluoro-2′-deoxyuridine 5′-phosphate. J . Med. Chem. 1983, 26,
1153-1158.
8) Lorey, M.; Meier, C.; De Clercq, E.; Balzarini, J . New Synthesis
and Antitumor Activity of CycloSal-Derivatives of 5-Fluoro-2′-
deoxyuridinemonophosphate. Nucleosides Nucleotides 1997, 16,
789-792.
3
1
.66 (m, 4H). P NMR (CDCl
3
): δ -12.82, -12.90 (mixture of
diastereomers). MS (FAB) C27
H
34ClFN P calculated 630.1784
3 4
O
+
(M + H) , found 630.1760.
(
5
-F lu or o-2′-d eoxyu r id yl 2-(1,4-n a p h th oqu in on yl)m e-
th yl N-m eth yl-N-(4-ch lor obu tyl) p h osp h or a m id a te (3b)
was prepared from ceric ammonium nitrate (225 mg, 0.41
mmol) and 2b (100 mg, 0.16 mmol) as described for compound
(9) Abraham, T. W.; Kalman, T. I.; McIntee, E. J .; Wagner, C. R.
Synthesis and biological activity of aromatic amino acid phos-
phoramidates of 5-Fluoro-2′-deoxyuridine and 1-B-arabinofura-
nosylcytoxine: Evidence of phosphoramidase activity. J . Med.
Chem. 1996, 39, 4569-4575.
(10) Fries, K. M.; J oswig, C.; Borch, R. F. Synthesis and biological
evaluation of 5-fluoro-2′-deoxyuridine phosphoramidate ana-
logues. J . Med. Chem. 1995, 38, 2672-2680.
(11) Meyers, C. L. F.; Hong, L.; J oswig, C.; Borch, R. F. Synthesis
and biological activity of novel 5-fluoro-2′-deoxyuridine phos-
phoramidate prodrugs. J . Med. Chem. 2000, 43, 4313-4318.
(12) Borch, R. F.; Liu, J .; Schmidt, J . P.; Marakovits, J . T.; J oswig,
C.; Gipp, J . J .; Mulcahy, R. T. Synthesis and evaluation of
nitroheterocyclic phosphoramidates as hypoxia-selective alky-
lating agents. J . Med. Chem. 2000, 43, 2258-2265.
13) Meyers C. L. F.; Borch, R. F. Activation mechanisms of nucleo-
side phosphoramidate prodrugs. J . Med. Chem. 2000, 43, 4319-
4327.
3
a . The crude product was purified by silica gel chromatog-
raphy (10% MeOH/CHCl ) to give 3b (88 mg, 93%) as a yellow
foam. R
3
1
f
) 0.50 (10% MeOH/CHCl
3 3
). H NMR (CDCl ): δ 8.10
(m, 2H); 7.77 (m, 2H); 7.72 (m, 1H); 7.03 (s, 1H); 6.18 (m, 1H);
5
.02 (m, 2H); 4.58 (m, 1H); 4.29 (m, 2H); 4.05 (m, 1H); 3.57
(
1
m, 2H); 2.73 (d, 3H, J ) 10.07 Hz); 2.47 (m, 1H); 2.31 (m,
3
1
H); 1.75 (m, 4H). P NMR (CDCl
3
): δ -12.67, -12.99
28ClFN P‚1.5 H O)
(
mixture of diastereomers). Anal. (C25
C, H, N.
N-Meth yl-N-4-h yd r oxybu tyla m in e was prepared accord-
H
3
O
9
2
1
6
ing to the procedure of Kuznetsov et al. Methylamine (53
mL, 1.20 mol) gas was condensed in a 12 in. pressure tube
using a dry ice-acetone bath. 4-Chloro-1-butanol (24 mL, 0.24
mol) was slowly added to the methylamine. Methanol (23 mL)
(

4
480 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 25
Tobias and Borch
(
14) Thomas, D. M.; Zalcberg, J . R. 5-Fluoruracil: A Pharmacological
(16) Kuznetsov, S. G.; Ioffe, D. V. Investigation in the Field of
Formation of Polymethyleneammonium Rings. J . Gen. Chem.
USSR (Eng. Transl.) 1961, 31, 2133-2139.
Paradigm in the Use of Cytotoxics. Clin. Exp. Pharm. Phys.
1
998, 25, 887-895.
(
15) Yalowich, J . C.; Kalman, T. I. Rapid determination of thymidy-
late synthase activity and its inhibition in intact L1210 leukemia
cells in vitro. Biochem. Pharmacol. 1985, 34, 2319-2324.
J M010337R