Enhancement of nucleoside cytotoxicity through nucleotide prodrugs

Article abstract of DOI:10.1021/jm020107s

A common reason for the lack of cytotoxicity of certain nucleosides is thought to be their inability to be initially activated to the monophosphate level by a nucleoside kinase or other activating enzyme. In a search for other nucleosides that might be wo

Full text of DOI:10.1021/jm020107s

J . Med. Chem. 2002, 45, 4505-4512
4505
En h a n cem en t of Nu cleosid e Cytotoxicity th r ou gh Nu cleotid e P r od r u gs
J erry D. Rose, William B. Parker, Hitoshi Someya, Sue C. Shaddix, J ohn A. Montgomery, and
J ohn A. Secrist III*
Southern Research Institute, P.O. Box 55305, Birmingham, Alabama 35255-5305
Received March 7, 2002
A common reason for the lack of cytotoxicity of certain nucleosides is thought to be their inability
to be initially activated to the monophosphate level by a nucleoside kinase or other activating
enzyme. In a search for other nucleosides that might be worthwhile anticancer agents, we
have begun to examine the utilization of monophosphate prodrugs in order to explore whether
any enhanced cytotoxicity might be found for the prodrugs of candidate nucleosides that have
little or no cytotoxicity. To that end, 5′-bis(pivaloyloxymethyl) phosphate prodrugs of two weakly
cytotoxic compounds, 8-aza-2′-deoxyadenosine (5) and 8-bromo-2′-deoxyadenosine (9), have been
prepared. These prodrugs (8 and 12) were examined for their cytotoxicity in CEM cells and
were found to possess significantly enhanced cytotoxicity when compared with the corresponding
parent nucleosides. Further cell culture experiments were conducted to gain insight into the
mechanisms of cytotoxicity of these two prodrugs, and those data are reported.
Sch em e 1
In tr od u ction
Over the years, we have prepared many nucleosides
for evaluation as potential anticancer agents. While
many of these compounds proved to have in vitro and/
or in vivo activity, a significant number of compounds
had little or no cytotoxicity when examined in our in
vitro tumor panel. While a number of reasons might be
the cause of this lack of activity, a common one is that
the nucleoside is not converted to its monophosphate
derivative by the initial activating enzyme. In most
cases, though not all, the steps from the monophosphate
to the triphosphate derivatives are more easily ac-
complished by the relevant enzymes than is the first
step. It has been clear to us for some years that we
might have prepared a number of potentially valuable
nucleosides if we were able to find a way to get them to
the triphosphate level in cells. To that end, the strategy
that has been utilized in the antiviral area for the same
purpose, that is, to synthesize a monophosphate pro-
drug, appeared to have promise for our purposes. To
evaluate the concept, we selected two weakly cytotoxic
compounds, 7-amino-3-(2-deoxy-â-D-erythro-pentofura-
nosyl)-3H-1,2,3-triazolo[4,5-d]pyrimidine (8-aza-2′-deoxy-
adenosine, 5, Scheme 1) and 8-bromo-2′-deoxyadenosine
(9, Scheme 2), for conversion to an appropriate mono-
phosphate prodrug. Prior studies in our laboratories had
determined that these two nucleoside analogues were
poor substrates for mammalian kinases (unpublished
observation) and were therefore good candidates to
evaluate this strategy. From the variety of possible
monophosphate prodrugs that were known when we
began this work,1 we selected the bis(pivaloyloxymethyl)
group for the initial test of the concept. We describe
herein the synthesis and biological evaluation in cell
culture of bis(pivaloyloxymethyl) 8-aza-2′-deoxyadenos-
ine-5′-monophosphate 8 and bis(pivaloyloxymethyl) 8-bro-
mo-2′-deoxyadenosine-5′-monophosphate 12.
Ch em istr y
The two nucleoside starting materials are both re-
ported in the literature. Nucleoside 9 is readily prepared
by bromination of 2′-deoxyadenosine.2 The synthesis of
5 has been reported several times,3-5 including once
from our laboratories.3 None of these routes provides
good yields of the desired product, and all of them
provide multiple products relative to anomeric mixture
and glycosylation site; therefore, we chose to pursue a
modified approach. The anion of N6-nonanoyl-8-azaad-
enine (1),4 generated by treatment of 1 with 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU),6 was coupled with
* To whom correspondence should be addressed. Tel: 205-581-2442.
Fax: 205-581-2870. E-mail: secrist@sri.org.
10.1021/jm020107s CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/24/2002

4506 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Rose et al.
Sch em e 2
showed 4.9% of a barely resolved second component that
was tentatively identified as the tetrakis(pivaloyloxy-
methyl) ester of the 3′,5′-diphosphate. This sample,
containing the small diphosphate impurity, was utilized
for the biological experiments.
Bioch em ica l Stu d ies
Monophosphate prodrug 12 was 10-fold more potent
as an inhibitor of CEM cell growth than parent nucleo-
side 9. The concentration of compound that was required
to inhibit CEM cell growth by 50% during a 96 h
incubation period was 28 ( 16 vs 275 ( 72 µM
(respectively, mean and standard deviation of at least
three separate experiments). Nucleoside 5 was known
to be toxic to human cells by virtue of its conversion to
8-aza-hypoxanthine through the sequential action of
adenosine (Ado) deaminase and purine nucleoside phos-
phorylase (unpublished observation). Its toxicity can be
prevented by addition to the culture medium of deoxy-
coformycin (dCF), which is an inhibitor of Ado and
adenosine 5′-monophosphate (AMP) deaminase. In the
presence of dCF, 8 was also more cytotoxic than 5 (IC50
of 15 ( 7 vs 190 ( 17 µM, respectively, mean ( standard
deviation of at least four separate experiments). The
increased potency of 8 and 12 with respect to the parent
nucleosides suggested that this strategy may be useful
in identifying novel cytotoxic nucleosides that could
have antitumor activity.
As expected for these types of compounds,11,12 they
rapidly disappeared from the culture medium with a
half-life of approximately 7 h (data not shown). The half-
lives were similar in culture medium without serum,
culture medium with serum, and culture medium with
serum plus cells, which indicated that the degradation
of the parent compounds in CEM cell culture was
primarily due to a nonenzymatic process. Because the
half-life of these two compounds in culture medium was
much shorter than the duration of the cytotoxicity
experiments, it was possible that the potency of these
agents would be enhanced if they were repeatedly
administered during the cytotoxicity assay. However,
repeated administration of compound (dosing on days
0, 1, 2, and 3) only increased the potency of these
compounds by about 2-fold (data not shown). The
products of the degradation reaction were not deter-
mined in these experiments. If the mono-PIV phos-
phodiester was the major product, it is possible that it
could also penetrate cell membranes and result in
cytotoxicity to the cells.
2-deoxy-3,5-di-O-p-toluoyl-R-D-erythro-pentofuranosyl
chloride7 in acetonitrile. This procedure produced both
the R and the â anomers of nucleosides attached through
the 8- and 9-positions (purine numbering). Separation
of the desired 9â anomer, which had already lost its
nonanoyl group, was followed by conventional depro-
tection with NH3/EtOH at 75 °C to afford 5, with
properties in agreement with published data.3-5
Conversion of the two nucleosides into the desired
prodrugs involved initial phosphorylation followed by
introduction of the bis(pivaloyloxymethyl) group. We
chose to utilize standard phosphorylation methodology
for both nucleosides, involving treatment with 2 equiv
of phosphorus oxychloride in trimethyl phosphate at
reduced temperature. Phosphorylation of 5 under these
conditions at -10 °C produced a mixture of the desired
monophosphate 6 and the corresponding 3′,5′-diphos-
phate 7, unexpectedly in about equal amounts. Partial
separation was possible through chromatographic means,
and a sample with about 69% 6 was carried forward.
Various methods have been utilized for the introduc-
tion of pivaloyloxymethyl groups into nucleoside phos-
phate salts,8,9 and they are generally characterized by
low to moderate yields. We chose direct esterification
of 6 with commercially available chloromethyl pivalate
in Et3N/1-methyl-2-pyrrolidinone10 because it satisfied
our goal of rapidly obtaining sufficient 8 for biological
evaluation without additional synthetic steps. Separa-
tion of 8 from the corresponding 3′,5′-diphosphate
tetrakis(pivaloyloxymethyl) and tris(pivaloyloxymethyl)
esters was accomplished using a silica gel column.
Target compound 8 was obtained as a low melting, but
tractable, brittle glass.
Scheme 2 outlines the synthesis of target compound
12 from nucleoside 9. Phosphorylation was accom-
plished by the method described above except that the
temperature was kept in the -14 to -12 °C range. After
some chromatographic purification, a mixture comprised
of mainly monophosphate 10 (89%) but also containing
a small amount of starting nucleoside 9 (9%) and a
lesser quantity of diphosphate 11 was carried forward.
Reaction of the mixture with chloromethyl pivalate as
described above, followed by column purification, gave
the desired product 12. This material appeared to be of
high quality by thin-layer chromatography (TLC), but
high-pressure liquid chromatography (HPLC) analysis
Treatment of CEM cells with either compound re-
sulted in the inhibition of the incorporation of [3H]-
uridine (Urd) into RNA and [3H]thymidine into DNA
but did not affect the incorporation of [3H]leucine into
protein (Table 1). These results, suggesting that both
RNA and DNA synthesis were inhibited by these
compounds, are quite different from those normally
observed with other deoxynucleoside analogues. Inhibi-
tion of the incorporation of [3H]Urd into RNA could
result from the inhibition of Urd transport, RNA poly-
merases, or the enzymes involved in Urd metabolism.
Although treatment with these two compounds inhibited
the utilization of Urd, they did not have any effect on
uridine 5′-triphosphate (UTP) levels in CEM cells (data
not shown). Therefore, it is not likely that these agents

Enhancement of Nucleoside Cytotoxicity
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4507
Ta ble 1. Effect of 8 or 12 on RNA, DNA, and Protein
Synthesesa
percent of control
DNA
RNA
protein
treatment
synthesis
synthesis
synthesis
12
8
8 + dCF
53
11
11
14
3
3
94
95
98
a
CEM cells were incubated with either 29 µM 12 or 35 µM 8
(plus or minus 10 µM dCF). [3H]thymidine, [3H]Urd, or [3H]leucine
were added 30 min after the initiation of the incubation; samples
were taken 1, 2, and 4 h after the addition of radioactive labels;
and the incorporation of radiolabeled precursors into DNA, RNA,
or protein (respectively) was determined as described.12,13 This
experiment has been repeated twice with similar results.
F igu r e 1. Effect of 12 or 8 on the utilization of [3H]Urd in
CEM cells. CEM cells (300 000 cells/mL) were incubated at
37 °C with [3H]Urd (2 µC/mL) plus 29 µM 12, 35 µM 8, or no
drugs for 2 h. Cells (8 mL) were collected and washed, and an
acid soluble extract was prepared and injected onto SAX HPLC
as described.14 One minute fractions were collected as they
eluted from the column and were counted for radioactivity.
This experiment has been repeated two times with similar
results.
F igu r e 2. Effect of 12 or 8 on the incorporation of [3H]adenine
into RNA and DNA. CEM cells (600 000 cells/mL) were
incubated with 1 µM [3H]adenine (1 µCi/mL) plus 29 µM 12,
35 µM 8 plus 10 µM dCF, or no drugs. The compounds were
added 30 min prior to the addition of [3H]adenine. One, two,
and four hours after the addition of [3H]adenine, a 4 mL cell
sample was taken from each incubation and the amount of
radioactivity in RNA (A) and DNA (B) was determined as
described.12,13 This experiment has been repeated with similar
results.
inhibited uridylate kinase or nucleotide diphosphate
kinase. However, inhibition of Urd transport or Urd
kinase activity by these agents could inhibit the utiliza-
tion of [3H]Urd without affecting pyrimidine nucleotide
pools. To test for this possibility, the effect of both
compounds on the phosphorylation of [3H]Urd to UTP
was determined (Figure 1). Both compounds decreased
the metabolism of [3H]Urd, indicating that these agents
inhibited either the transport of Urd or the Urd kinase
activity and that RNA polymerase was not a target of
these compounds. To clarify the effect of these com-
pounds on RNA and DNA synthesis, the effect of both
compounds on the incorporation of [3H]adenine into
RNA and DNA was determined. Neither compound
inhibited the incorporation of [3H]adenine into RNA,
which indicated that RNA synthesis was not inhibited
by these agents (Figure 2A). However, both compounds
inhibited the incorporation of [3H]adenine into DNA
(Figure 2B).
20 µg/mL) inhibited Urd kinase activity in cell-free
extracts of CEM cells (data not shown), however, which
indicated that inhibition of this activity by the parent
compounds was not the reason for the decrease in Urd
metabolism in these cells. The parent compound was
stable for the duration of the experiment. Although
inhibition of Urd utilization by these agents is an
interesting observation, this effect of these agents would
not result in toxicity since CEM cells are not dependent
on pyrimidine salvage for generation of UTP; i.e., little
or no Urd is provided in the culture medium for the
growth of these cells, so inhibition of its utilization
would not affect growth of these cells.
Acid soluble extracts were obtained from CEM cells
treated with 145 µM 12 and analyzed by strong anion
exchange HPLC (SAX) to determine whether 8-Br-dATP
was formed in the cells. In our standard SAX HPLC
system, 8-Br-dATP (chemically synthesized standard)
eluted between adenosine 5′-triphosphate (ATP) and
guanosine 5′-triphosphate (GTP) with a retention time
of approximately 38 min. Although there was evidence
of new peaks in the monophosphate region of the
chromatogram, there was no peak that eluted with a
Both compounds immediately inhibited the uptake
and utilization of the [3H]Urd in CEM cells (data not
shown). Unfortunately, because uptake was rate-limit-
ing in the conversion of Urd to UTP in these cells, we
could not determine which process was being inhibited,
Urd uptake or Urd kinase activity. Neither 12 nor 8 (at

4508 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Rose et al.
F igu r e 3. Metabolism of 12 in CEM cells. CEM cells (700 000
cells/mL) were treated with 145 µM 12 (B) or no compounds
(A) for 7 h. An acid soluble extract from 50 × 106 cells was
obtained and analyzed by SAX HPLC as described.4 The
natural nucleotides and 8-Br-dAMP metabolites were detected
by their absorbance at 260 nm as they eluted from the column.
The experiment has been repeated with similar results.
retention time equal to that of 8-Br-dATP (Figure 3).
In contrast to the results with 12, two new peaks were
detected in the triphosphate region of SAX HPLC
separation of acid soluble extracts created from cells
that were treated with 70 µM 8 (Figure 4). One of the
peaks eluted between ATP and GTP with a retention
time that was identical to that of an authentic standard
of 8-aza-dATP. The other peak eluted just after GTP
and had a UV spectrum that was similar to an authentic
standard of 8-aza-GMP, which was the standard avail-
able in our laboratories. The addition of either hypo-
xanthine or dCF prevented the synthesis of the peak
that eluted just after GTP but had no effect on the
generation of 8-aza-dATP from 8. The fact that hypox-
anthine inhibited the synthesis of this peak suggested
that the compound must first be cleaved to 8-aza-
hypoxanthine prior to conversion to triphosphate. There-
fore, the metabolite that eluted just after GTP is likely
to be mostly 8-aza-GTP. Hypoxanthine would not be
expected to affect the synthesis of this metabolite, if it
were 8-aza-dGTP. Once 8-aza-dAMP is formed from the
parent compound, however, it is possible that it could
be deaminated to 8-aza-dIMP, which could be converted
to 8-aza-dGTP. The peak that eluted after GTP but not
the one that eluted prior to GTP was also detected in
cells treated with 5 (data not shown). These studies are
direct evidence that 8 entered the CEM cells and that
the bis(pivaloyloxymethyl) groups came off the molecule
in the cells liberating 8-aza-dAMP, which was utilized
by the cellular kinases to create 8-aza-dATP.
F igu r e 4. Metabolism of 8 in CEM cells. CEM cells (700 000
cells/mL) were treated with 70 µM 8 (B), 70 µM 8 plus 735
µM hypoxanthine (C), 70 µM 8 plus 10 µM deoxycoformycin
(D), or no compounds (A) for 4 h. An acid soluble extract from
21 × 106 cells was obtained and analyzed by SAX HPLC as
described in the legend to Figure 3 except that no gradient
was used to separate the acid soluble metabolites. The samples
were eluted with 500 mM potassium phosphate buffer. The
natural nucleotides and 8-aza-dAMP metabolites were de-
tected by their absorbance at 260 nm as they eluted from the
column. The experiment has been repeated with similar
results.
µM, respectively, N ) 3). The Km of dATP with DNA
polymerase R was 4 ( 1 µM, which indicated that these
compounds did not potently interact with DNA poly-
merase R. 8-Aza-dATP was able to substitute for dATP
in reactions that utilized [3H]dGTP instead of [3H]dATP
to label the newly synthesized DNA (Figure 5), which
indicated that 8-aza-dATP was not an inhibitor of the
DNA polymerase reaction but was instead a good
alternative substrate for DNA synthesis. In contrast,
8-Br-dATP was not utilized as efficiently as dATP or
8-aza-dATP for DNA synthesis (Figure 5). However, in
this experiment, 8-Br-dATP did not inhibit the incor-
Both 8-Br-dATP and 8-aza-dATP were competitive
inhibitors of the incorporation of dATP into DNA by
DNA polymerase R (Ki values of 548 ( 93 and 156 ( 36

Enhancement of Nucleoside Cytotoxicity
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4509
of DNA polymerase R cannot account for the inhibition
of DNA synthesis. However, numerous DNA poly-
merases are involved in the synthesis and repair of
chromosomal DNA, and it is possible that one or more
of these enzymes could be inhibited by 8-aza-dATP. Our
results indicated that 8-aza-dAMP would readily be
incorporated into chromosomal DNA in place of dAMP.
The incorporation of 8-aza-dAMP into DNA could dis-
rupt the function of DNA and result in the inhibition of
DNA synthesis. This possibility seems unlikely, how-
ever, because DNA synthesis was immediately inhibited
by 8. The 5′-triphosphates of numerous deoxyadenosine
analogues are potent inhibitors of ribonucleotide reduc-
tase (6), and inhibition of this enzyme by 8-aza-dATP
would result in the inhibition of DNA synthesis. Finally,
it should be noted that although a considerable amount
of 8-aza-dATP was generated in CEM cells treated with
8 and triphosphate metabolites of nucleoside analogues
are often responsible for the toxicity of these types of
agents, it is possible that another metabolite, or the
parent prodrug itself, is responsible for the inhibition
of DNA synthesis and toxicity of this agent.
Our results indicated that the generation of 8-Br-
dATP was not likely to be involved in the mechanism
of action of 12. Although it is possible that small
amounts of 8-Br-dATP were formed in CEM cells that
were below the levels of detection used in these experi-
ments, it is not likely that generation of small amounts
of 8-Br-dATP would have much effect on DNA synthesis
or DNA function given the high concentrations needed
to inhibit DNA polymerase R.
These biological evaluations indicated that the syn-
thesis of 5′-bis(pivaloyloxymethyl) derivatives of non-
toxic nucleoside analogues can lead to toxic molecules
that do not need to be activated by nucleoside kinases.
It is likely that other nucleotide prodrug systems that
allow the by-passing of the kinase activation step will
also have utility in this regard.1 Nucleoside kinases are
often the rate-limiting step in the activation of nucleo-
side analogues used in the treatment of cancer. Thus,
the generation of these types of compounds should
expand the structural variety of nucleoside analogues
that are currently available for evaluation as antitumor
agents and could lead to novel classes of antitumor
drugs. The concept is a reasonable one to develop novel
cytotoxic agents, and we are pursuing its application
with other prodrug systems and with other nontoxic
nucleosides.
F igu r e 5. Utilization of 8-Br-dATP and 8-aza-dATP as
substrates for DNA synthesis. DNA polymerase R isolated from
CEM cells was incubated in 50 µL reactions with dATP (0,
0.5, 1, 10, or 100 µM), aza-dATP (0, 10, 50, 100, or 500 µM),
Br-dATP (0, 100, 200, 1000, or 3000 µM), or araATP (1, 10, or
100 µM). The numbers above the bars in the figure are the
concentration (expressed in mM) of the deoxyadenosine nucle-
otides that were used in the experiment. The reactions also
included 50 mM Tris (pH 8.0), 1 mg/mL bovine serum albumin,
10 mM MgCl2, 1 mM dithiothreitol, 100 µg/mL activated DNA,
100 µM each of dTTP and dCTP, and 10 µM [3H]dGTP (10 000
Ci/mol). After they were incubated at 37 °C for 1 h, the reaction
mixtures were spotted on glass fiber filters and washed with
5% trichloroacetic acid and ethanol, and the radioactivity on
each filter was determined.20 This experiment has been
repeated three times with similar results.
poration of [3H]dGTP into DNA at concentrations that
were above its Ki. The slight increase in DNA synthesis
activity suggested that the incorporation of 8-Br-dAMP
into the 3′ terminus also did not result in termination
of the DNA chain and that the Vmax of the DNA
synthesis reaction with 8-Br-dATP was less than that
for dATP and 8-aza-dATP. The effect of arabinofurano-
syl ATP (araATP) on the incorporation of [3H]dGTP into
DNA is shown in Figure 5 for comparison purposes.
AraATP is utilized by DNA polymerase R as a substrate
for DNA synthesis, and its incorporation into the 3′
terminus of a DNA chain results in the subsequent
inhibition of chain elongation by the polymerase.13
Discu ssion
Although both compounds selectively inhibited DNA
synthesis and were cytotoxic to CEM cells, our results
suggested that the mechanism of action of these two
compounds could be quite different. The major differ-
ence between these two compounds was that 8 was
readily converted to 8-aza-dATP in CEM cells, whereas
no 8-Br-dATP was detected in cells treated with 12. This
difference in the metabolism of these two compounds
indicated that one or both of the enzymes involved in
the generation of the triphosphate (AMP kinase and
nucleoside diphosphate kinase) were able to discrimi-
nate between 8-Br- and 8-aza-nucleotides. Given the
broad substrate specificity of nucleoside diphosphate
kinase, it is likely that 8-Br-dAMP is a poor substrate
for AMP kinase.
Exp er im en ta l Section
Ch em ica l Meth od ology. Melting points were determined
by the capillary method on a Mel-Temp apparatus and are
uncorrected. All evaporations were carried out in vacuo on a
rotary evaporator or under high vacuum by short-path distil-
lation into a glass trap cooled in dry ice/acetone. Samples were
dried under high vacuum over P2O5 at room temperature (rt)
unless an elevated temperature is given. Elemental analyses
were performed by the Spectroscopic and Analytical Labora-
tory of the Southern Research Institute or by Atlantic Micro-
lab, Inc., Atlanta, GA. Where solvents such as EtOH or H2O
are included in
a reported analysis, their presence was
confirmed by the 1H NMR spectrum. Elemental analyses
within (0.4 of the calculated values are reported only by the
symbols of the elements. Analtech precoated (250 µm) silica
gel G (F) plates were used for TLC, and spots were detected
by irradiation with a 254 nm UV light, by absorption of I2
Currently, it is not clear how 8 inhibits DNA synthe-
sis. Because 8-aza-dATP supported DNA synthesis with
DNA polymerase R at a rate that was similar to that
for dATP (although at higher concentrations), inhibition

4510 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Rose et al.
vapor, or by charring after a (NH4)2SO4/H2SO4 spray. Column
chromatography in either the flash column or the gravity mode
was performed with silica gel 60 (230-400 mesh) using the
slurry method of column packing. Mass spectra were recorded
on a Varian/MAT 311A double-focusing mass spectrometer in
the positive fast-atom bombardment (FAB) mode. A few
spectra of nucleoside phosphate salts were also obtained in
the negative electrospray mode from Mr. Marion Kirk of the
University of Alabama at Birmingham. Addition of a trace of
LiCl to some samples confirmed the identification of molecular
ions by giving (M + Li)+ cluster peaks that were stronger than
the (M + H)+ peaks. 1H NMR spectra were recorded on a
Nicolet/GE NT300NB spectrometer operating at 300.635 MHz.
Chemical shifts (δ) are reported in parts per million downfield
from internal tetramethylsilane. Chemical shifts of multiplets
are measured from the approximate center of the multiplet,
and coupling constants (J ) are reported in Hz. Ultraviolet
spectra were determined in 0.1 N HCl (pH 1), pH 7 buffer,
and 0.1 N NaOH (pH 13) with a Perkin-Elmer UV-visible
near-infrared model Lambda 9 spectrophotometer. Maxima
and extinction coefficients are reported in the format: λmax in
nanometers (ꢀ × 10-3). HPLC was performed on either
Hewlett-Packard or Perkin-Elmer equipment with variable
wavelength UV detection by the following methods. Method
A (reversed phase): Phenomenex Sphereclone 5 µm, 4.6 mm
× 250 mm column; H2O/MeCN (2:3); 1 mL per minute. Method
B (strong anion exchange): Keystone Scientific Partisil, 4.6
mm × 250 mm column; solution 1:5 mM NH4H2PO4 (ADHP)
adjusted to pH 2.8; solution 2: 750 mM ADHP adjusted to
pH 3.7; 50 min linear gradient from 100% 1 to 100% 2 at 2.0
mL/min. Method C (ion-pairing): same column as method A;
solution 1: 0.01 M ADHP + 0.005 M tetrabutylammonium
phosphate adjusted to pH 5.0 with H3PO4; solution 2: MeOH;
20 min linear gradient from 20 to 50% 2 at 1 mL/min. Method
D: same column as A; solution 1: 0.01 M ADHP; solution 2:
MeOH; 20 min linear gradient from 10 to 90% 2 at 1 mL/min.
of the previously unreported melting behavior, our data for
â4 are included for confirmation. MS: m/z 489 (M + H)+, 353
(sugar fragment)+, 137 (B + 2H)+. HPLC (method A): 99.7%;
retention time, 8.34 min (R4, 7.24 min). TLC (cycl/EtOAc
1:3): Rf 0.47 (R4, 0.42). UV(MeOH): 0.1 N HCl: 246 (36.9);
pH 7 buffer (slightly cloudy); 249(25.5), 286(17.8); 0.1 N NaOH:
1
238 (26.1), 277 (12.0). H NMR (DMSO-d6): δ 2.37, 2.41 (2 s,
6H, Tol CH3), 2.91, 3.55 (2 m, 1 each, H-2′s), 4.46, 4.57 (2 m,
2H, 5′-CH2), 4.65 (m, 1H, H-4′), 5.93 (m, 1H, H-3′), 6.83 (t, 1H,
H-1′, J ) 6.4), 7.32, 7.82, 7.96 (m, 8H, Tol CH), 8.21, 8.56 (2 s,
2H, 6-NH2), 8.31 (s, 1H, H-2′). Anal. Calcd for C25H24N6O5: C,
61.47; H, 4.95; N, 17.20. Found: C, 61.50; H, 5.15; N, 17.18.
For comparison, our R4 melted 188-189 °C; literature5 187
°C; spectral data also agreed with published values.
8-Aza -2′-d eoxya d en osin e (5). A solution of â4 (3.70 g, 7.57
mmol) in EtOH saturated with NH3 (450 mL) was heated for
3 days in a glass-lined stainless steel bomb in a 75 °C oil bath.
The cooled solution was evaporated, and the residue was
washed twice by trituration and filtration with Et2O to remove
the p-toluamide and ethyl p-toluate. A solution of the solid
filter cake in hot MeOH was evaporated with silica gel (∼50
mL dry volume), and the powder was layered carefully onto a
SG column. Elution with CHCl3/MeOH (9:1) gave 5 in fractions
totaling 1.63 g (85%). The best of these (1.27 g) was recrystal-
lized from MeOH (200 mL) and dried in vacuo over P2O5 for
18 h at 100 °C; recovery, 0.96 g (75%); mp 201-202 °C,
resolidifies on continued heating and chars but does not melt
below 300 °C (literature4 200-201 °C). TLC (CHCl3/MeOH
4:1): Rf 0.53. 1H NMR5 and UV4 spectra agreed with the
literature. Anal. (C9H12N6O3) C, H, N.
Tr ieth yla m m on iu m 8-Aza -2′-d eoxya d en osin e-5′-m on o-
p h osp h a te (6) a n d Tr ieth yla m m on iu m 8-Aza -2′-d eoxy-
a d en osin e-3′,5′-d ip h osp h a te (7). Under an atmosphere of
dry N2, a stirred suspension of 5 (137 mg, 0.54 mmol) in a
mixture of trimethyl phosphate (2.5 mL) and Et3N (110 mg,
1.09 mmol, 2 equiv) was cooled to -10 °C in an ice/acetone
bath and treated dropwise over 5 min with POCl3 (0.10 mL,
167 mg, 1.09 mmol) added via hypodermic syringe. After 1 h,
the mixture was poured into 50 mL of ice/water, and the
solution was neutralized (meter) by dropwise addition of cold
Et3N. The pH drifted downward because of the slow hydrolysis
of cold POCl3, so it was monitored and adjusted for ∼40 min
until it stabilized at pH 7, and then the solution was frozen
and lyophilized overnight to give a paste of solid in PO(OMe)3.
Trituration with Et2O (50 mL) and filtration gave a hygroscopic
white solid: wt 816 mg (>100% because of extra triethylam-
monium salts). A mass spectrum in the negative mode showed
a strong 331 (M - 1)- (6), a medium 411 (M - 1)- (7), and a
weak 491 (M - 1)- for a triphosphate. An attempt was made
to separate the monophosphate and diphosphate components
on a column of washed and defined Darco G60 activated
charcoal, eluting first with H2O and then EtOH /H2O/
concentrated NH4OH (50:50:1). Some desalting was achieved,
but the nucleotides came off in a single broad band; wt 130
mg after lyophilization.
8-Aza -2′-d eoxy-3′5′-d i-O-(p -tolu oyl)a d en osin e (â4). Un-
der a N2 atmosphere, freshly opened DBU (0.550 g, 3.62 mmol)
was added to a stirred suspension of 14 (1.00 g, 3.62 mmol) in
dry acetonitrile (30 mL), forming a clear solution in 5 min.
After 15 min, 27 (1.62 g, 4.16 mmol, 1.15 eq) was added in one
portion, and the mixture became a clear solution in ∼10 min.
After 48 h, the solvent was evaporated, and the residual gum
was applied directly to a silica gel column. Elution with CHCl3/
MeOH (first 99:1 and then 4:1) separated the mixture into
three major zones. The first zone (0.7 g) was itself a complex
mixture of at least six mobile components. The main compo-
nents of the second zone (0.99 g, 56%) were an anomeric pair
that were shown subsequently to be â4 (9â isomer) followed
by R4. The third, least mobile zone (0.54 g, 31%) was likewise
an anomeric pair that was mostly R and â3. Recolumning of
the second zone using cyclohexane/EtOAc (2:3) gave fractions
totaling 423 mg (24%) of sufficient purity to allow positive
spectral identification as the desired â4 and trailing fractions
totaling 465 mg (26%) that were identified as R4.
The pilot reaction above was scaled up using the following
quantities of reagents: DBU (8.09 g, 53.2 mmol), 1 (14.69 g,
53.2 mmol), 2 (22.74 g, 58.5 mmol, 1.1 equiv). Evaporation of
the reaction mixture after 48 h gave a gum that was triturated
with Et2O (500 mL) and filtered under N2 pressure to remove
most of the insoluble DBU‚HCl. Evaporation of the filtrate
gave a soft foam (34.8 g) that was chromatographed on a SG
column with elution by cyclohexane/EtOAc 2:3 followed by 1:3
to give three major zones as described above. The middle zone
(23 g, crude) was recolumned repeatedly (cyclohexane/EtOAc
2:3), selecting the best fractions at each stage for further
purification. Finally, samples of the desired â4 of high isomeric
and anomeric purity were obtained totaling 4.58 g (18%). An
analytical sample was prepared by recrystallization of a
portion from cyclohexane/EtOAc (4:1) with filtration and
drying of the white crystalline solid in vacuo over P2O5 at 65
°C for 18 h; mp 93-94 °C; on continued heating, it solidifies
and melts again 163-164 °C. Spectral properties of â4, as well
as R4, R3, and â3, have all been published;5 however, because
With 925 mg of 5 as starting material, a 6.75-fold scale-up
of the phosphorylation reaction was performed as described
above except that reaction time was decreased to 45 min. After
the crude lyophilized mixture was washed with Et2O, the solid
(5.52 g) was washed by trituration and decantation with CHCl3
(3 × 50 mL) to remove most of the extraneous Et3N‚HCl. HPLC
(method B) on the washed solid (2.62 g) showed 46% 6 and
45% 7. This was combined with 120 mg of similar material
from the pilot run. A column (ca. 4 cm × 30 cm) was poured
from a slurry of “gravity grade” silica gel (70-230 mesh) in a
solvent mixture of MeOH/H2O/concentrated NH4OH (6:2:1),14
and the column was washed with the solvent until the initially
turbid effluent ran clear (ca. 250 mL). A solution of the
combined products in H2O (12 mL) was applied, and the
column was eluted at its maximum flow rate of ∼1.2 mL/min
with continuous UV monitoring at 275 nm. After TLC of
selected samples, the fractions containing most of the UV
active material were pooled into two samples, which were
evaporated, dissolved in H2O (10 mL), filtered to remove silica,

Enhancement of Nucleoside Cytotoxicity
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4511
frozen, and lyophilized overnight. The first of these, from the
leading edge of the nucleotide band, weighed 0.93 g and was
shown by HPLC (method C) to be 69% 6 and 22% 7 plus trace
constituents. The second sample (0.53 g), from the trailing
edge, was 9% 6 and 87% 7. For the 0.93 g sample: MS
(positive): m/z 333 (M + H)+ of 6, 413 (M + H)+ of 7, 137 (B
+ 2H)+; (negative): 331 (M - H)-, 411 (M - H)-. For the 0.53
g sample: MS (negative): very strong 411 (7), weak 331 (6).
31P NMR (D2O, HEPES buffer, EDTA; referenced to external
H3PO4 as δ ) 0.0 ppm): δ 2.58 (d, J ) 7.7, 5′-PO4-2), 2.91 (t,
J ) 5.4, 3′-PO4-2).
Bis(p iva loyloxym et h yl) 8-Aza -2′-d eoxya d en osin e-5′-
m on op h osp h a te (8).10 The 0.93 g sample containing 69% 6
was refluxed with benzene (150 mL) using a Dean-Stark trap
to collect H2O. After 20 min, the mixture was cooled and
evaporated to give 0.900 g of amorphous powder. A stirred
suspension of the powder (69% as 6‚[Et3N]2 salt C21H43N8O6P,
621 mg, 1.2 mmol) in anhydrous N-methylpyrrolidinone (NMP,
13 mL) and Et3N (1.74 g, 17.2 mmol) was treated with
chloromethyl pivalate (3.33 g, 22.1 mmol) and heated for 5 h
in a 60 °C oil bath under a static N2 atmosphere. After it was
cooled, the mixture was diluted with EtOAc (200 mL), and the
soft, white precipitate (Et3N‚HCl and unreacted 6, TLC) was
collected under N2 pressure and washed liberally by trituration
on the funnel with CHCl3 (3 × 50 mL) to extract Et3N‚HCl.
The weight of recovered 6 was 372 mg (60%). The EtOAc
filtrate was washed by shaking with H2O (3 × 20 mL), and
the H2O wash was back-extracted with EtOAc. The dried (Na2-
SO4) organic layer was evaporated and then evacuated on the
oil pump overnight to give a viscous oil (0.69 g) that was shown
by MS and TLC to be a crude mixture containing 8 and the
corresponding 3′,5′-diphosphate tri- and tetra(pivaloyloxy-
methyl) esters.
The recovered 6 and various column fractions known to
contain 6 were combined and dried (536 mg). The reaction
above was repeated with this mixed starting material to give
an additional 0.21 g of material similar to the 0.69 g of product.
The combined sample was columned twice on SG with CHCl3/
MeOH 95:5 as eluent to give 8 as a brittle glass. A concentrated
solution in CH2Cl2 (2 mL) was evaporated under high vacuum
to give a tractable brittle foam, which was dried in vacuo over
P2O5 at room temperature for 24 h and at 56 °C for 1 h; yield
167 mg (∼24% from 1.2 mmol of 6); mp softens to a syrup 62-
64 °C. MS: m/z 561 (M + H)+. TLC (CHCl3/MeOH 95:5): Rf
0.27; diphosphate tetraester, Rf 0.56. UV (EtOH): 0.1 N HCl:
263 (11.8); pH 7 buffer; 279 (10.7); 0.1 N NaOH: 279 (10.7).
1H NMR (CDCl3): δ 1.23, 1.24 (2 s, 18H, pivaloyl CH3), 2.69,
3.22 (2 m, 1 each, H-2′), 3.39 (d, 1H, 3′-OH, J ) 3.7), 4.25
(complex m, 3H, H-4′ and H-5′), 5.09 (br s*, 1H, H-3′), 5.62
(m, 4H, O-CH2-O), 6.32 (br s, 2H, NH2), 6.78 (m, 1H, H-1′),
8.46 (s, 1H, H-2). *On expansion, this signal is a poorly
resolved multiplet. Anal. (C21H33N6O10P) C, H, N.
Tr iet h yla m m on iu m 8-Br om o-2′-d eoxya d en osin e-5′-
m on op h osp h a te (10) a n d Tr ieth yla m m on iu m 8-Br om o-
2′-d eoxya d en osin e-3′,5′-d ip h osp h a te (11). Under a N2 at-
mosphere, a stirred suspension of 92 (3.81 g, 11.4 mmol) in
PO(OMe)3 (50 mL) and Et3N (2.31 g, 22.8 mmol) was chilled
below -10 °C in an ice/acetone bath and treated dropwise over
5 min with POCl3 (3.50 g, 22.8 mmol). After 1 h at -14 to -12
°C, the mixture was poured into ice/H2O (250 mL); the yellow
solution was neutralized rapidly by addition of cold Et3N and
monitored (meter) for ∼45 min until it stabilized at pH 7.1.
The solution was lyophilized overnight to give a semisolid
paste, which was washed by trituration and decantation with
Et2O (2 × 200 mL) followed by CHCl3 (3 × 200 mL). The crude
solid residue weighed 7.1 g (102%). A partial solution of this
material in H2O was filtered to remove 202 mg (5%) of
unreacted 9, and the filtrate was applied to a SG column (70-
230 mesh, 600 mL dry volume), which was prepared and eluted
with MeOH/H2O/NH4OH (6:2:1) according to the method given
above for 6 to give 2.85 g (40%) of material that was mostly
the 5′-monophosphate 10, 1.49 g (15%) of the 3′,5′-diphosphate
11, 0.78 g of mixed 10 and 11, and 60 mg (1.6%) of 10. The
2.85 g sample of 10 was combined with similar material from
a previous pilot run, and the composite (3.7 g) was recolumned
as above. The best fractions (TLC) of 10 were combined,
concentrated to a small volume, filtered to remove silica,
frozen, and lyophilized to give a hygroscopic glass; wt 2.40 g.
MS (negative): m/z 408 (M - H)- of 10 with a peak of nearly
equal strength at 410 because of the 81Br isotope peak. HPLC
(method B): 89.4% 11 (retention time, 5.7 min) and 8.5% 9
(retention time, 2.2 min). TLC (MeCN/1 N NH4OH 3:2): Rf
0.46. 1H NMR (D2O + sodium 3-(trimethylsilyl)propionate-
2,2,3,3-d4): δ 2.45, 3.39 (2 m, 1 each, H-2′s), 4.10 (complex m,
3H, 5′-CH2 and H-4′), 6.52 (t, 1H, H-1′, J ) 7.3), 8.17 (s, 1H,
H-2). δ ) 0.0. Likewise, the best column fractions containing
11 were combined, evaporated, and dried as above; wt 0.26 g.
MS (negative): m/z 488, 490 (M - H)- of 11. HPLC (method
B): 96% 11 (retention time, 20.0 min). TLC (MeCN/1 N NH4-
OH): Rf 0.07. 1H NMR (D2O): 2.63, 3.45 (2 m, 2H, H-2′), 4.08,
4.15 (2 m, 2H, 5′-CH2), 4.32 (m, 1H, H-4′), 5.01 (m, 1H, H-3′),
6.59 (t, 1H, H-1′, J ) 7.4), 8.22 (s, 1H, H-2).
Bis(p iva loyloxym et h yl) 8-Br om o-2′-d eoxya d en osin e-
5′-m on op h osp h a te (12). The remainder after analyses of the
2.4 g sample of 10 was dried by azeotropic distillation with
benzene (see 8) to give 2.12 g of amorphous powder. A stirred
suspension of the powder (3.08 mmol calculated as 89% 10‚
[Et3N]2) in a mixture of N-methylpyrrolidinone (27 mL, NMP),
Et3N (3.63 g, 35.9 mmol), and chloromethyl pivalate (7.30 g,
48.6 mmol) was heated in a 60 °C oil bath under a N2
atmosphere for 5 h. The cooled mixture was diluted with
EtOAc (400 mL), and the soft precipitate (mostly Et3N‚HCl,
TLC) was filtered off. The filtrate was washed by shaking with
H2O (3 × 50 mL), dried (Na2SO4), and evaporated to give a
mixture of crude 12 and NMP; wt 6.8 g. Column chromatog-
raphy on SG with CHCl3/MeOH 95:5 as eluent gave 0.82 g
(41%) of impure 12. The best fractions containing 12 were
combined (0.45 g) and recolumned. Again, the best fractions
(TLC) were pooled, evaporated, and reevaporated with CH2-
Cl2 to give a brittle foam, which was dried in vacuo over P2O5
at room temperature for 48 h; yield 379 mg (19%); mp softens
to a syrup ∼40 °C. This apparently homogeneous (TLC) sample
of 12 was shown by HPLC (method D) to contain 4.9% of
another component tentatively believed to be the tetrakis-
(pivaloyloxymethyl) ester of the diphosphate. Elemental analy-
ses and calculated extinction coefficients reflect the approxi-
mate composition C22H33BrN5O10P‚0.05C34H54BrN5O17P2. MS
(positive): m/z 638 (M + H)+ and 81Br isotope peak at 640;
214, 216 (B + 2H)+. TLC (CHCl3/MeOH 9:1): Rf 0.37. UV
(EtOH): 0.1 N HCl: 263 (18.2); pH 7 buffer; 212 (25.4), 265
(16.7); 0.1 N NaOH: 265 (17.1). 1H NMR (CDCl3): δ 1.22, 1.23
(2 s, 18.8 H*, pivaloyloxymethyl CH3), 2.43, 3.55 (2 m, 1 each,
H-2′ s), 3.17 (br s, 1H, 3′-OH), 4.19 (m, 1H, H-4′), 4.26, 4.44 (2
m, 1 each, 5′-CH2), 5.01 (br s, 1H, H-3′), 5.53 (br s, 2H, 6-NH2),
5.65 (complex m, 4.5H*, O-CH2-O), 6.44 (t, 1H, H-1′, J )
6.9), 8.28 (s, 1H, H-2). Anal. (C22H33BrN5O10P‚0.05C34H54
-
BrN5O17P2) C, H, N, Br. *Both of these signals include the
“extra” protons from the minor component.
Biologica l Meth od ology. CEM cells (American Type
Culture Collection, Rockville, MD) were grown in RPMI 1640
medium (Gibco-BRL, Gaithersburg, MD) containing 10% fetal
bovine serum (Atlanta Biologicals, Atlanta, GA), 1 mg/mL
sodium bicarbonate, 10 units/mL penicillin, 10 µg/mL strep-
tomycin, and 50 µg/mL gentamycin. Cell numbers were
determined with a Coulter Counter, and the concentration of
compound that resulted in 50% inhibition of cell growth over
a 96 h incubation period was determined (IC50) and used as a
measure of cytotoxicity.
The effect of the compounds on the incorporation of radio-
labeled precursors ([8-14C]adenine, [5-3H]Urd, [methyl-3H]-
thymidine, or [4,5-3H]leucine; Moravek Biochemicals, Brea,
CA) into RNA, DNA, or protein was determined as de-
scribed.15,16 The incorporation of Urd or adenine into RNA is
determined by subtracting the incorporation of radiolabel into
the alkali stable/acid precipitable fraction (DNA) from the total
acid precipitable fraction (DNA plus RNA). Urd and adenine
are primarily incorporated into RNA but can also be incorpo-
rated into DNA as 2′-deoxycytidine or 2′-deoxyadenosine,

4512 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Rose et al.
(8) Lefebvre, I.; Pe´rigaud, C.; Pompon, A.; Aubertin, A.-M.; Girardet,
J .-L.; Kirn, A.; Gosselin, G.; Imbach, J .-L. Mononucleoside
phosphotriester derivatives with S-acyl-2-thioethyl bioreversible
phosphate-protecting groups: Intracellular delivery of 3′-azido-
2′,3′-dideoxythymidine 5′-monophosphate. J . Med. Chem. 1995,
38, 3941-3950.
(9) Farquhar, D.; Khan, S.; Srivastva, D. N.; Saunders, P. P.
Synthesis and antitumor evaluation of bis[(pivaloyloxy)methyl]
2′-deoxy-5-fluorouridine 5′-monophosphate (FdUMP): A strategy
to introduce nucleotides into cells. J . Med. Chem. 1994, 37,
3902-3909.
(10) This nonstoichiometric method was based upon a private com-
munication with Dr. Ernest Prisbe, Gilead Sciences, Foster City,
California.
(11) Pompon, A.; Lefebrve, I.; Imbach, J .-L.; Kahn, S.; Farquhar, D.
Decomposition pathways of the mono- and bis(pivaloyloxy-
methyl)esters of azidothymidine 5′-monophosphate in cell extract
and in tissue culture medium: an application of the ‘on-line
ISRP-cleaning’ HPLC technique. Antiviral Chem. Chemother.
1994, 5, 91-98.
respectively. The incorporation of thymidine and leucine into
acid precipitable material is a measure of their incorporation
into DNA and protein, respectively.
The acid soluble extracts were obtained from CEM cells
incubated with the nucleotide analogues and were analyzed
using strong anion exchange HPLC as described.17 8-Br-dATP
and 8-aza-dATP were made by Sierra Bioresearch (Tucson, AZ)
from the nucleoside that was provided by us. The structures
of both compounds were verified by phosphorus and hydrogen
NMR.
Urd kinase activity was measured in cell-free extracts of
CEM cells as described.18 Urd uptake was measured using the
oil-stop method described by Paterson et al.19 DNA polymerase
R was purified 52-fold from CEM cells to a specific activity of
120 nmol/mg/hr, and its activity was measured in 50 µL
reactions as described.20 The inhibition constants (Ki) of 8-aza-
dATP and 8-Br-dATP were determined from a plot of the slopes
of the Lineweaver-Burk lines vs the concentration of inhibitor
(three concentrations of each compound were used in the
determination of the Ki). Each line of the Lineweaver-Burk
plots was determined by linear regression from at least five
points. The experiment was done three times.
(12) Robbins, B. L.; Connelly, M. C.; Marshall, D. R.; Srinivas, R. V.;
Fridland, A. A human T lymphoid cell variant resistant to the
acyclic nucleoside phosphonate 9-(2-phosphonylmethoxyethyl)-
adenine shows a unique combination of a phosphorylation defect
and increased efflux of the agent. Mol. Pharmacol. 1995, 47,
391-397.
(13) Parker, W. B.; Bapat, A. R.; Shen, J . X.; Townsend, A. J .; Cheng,
Y. C. Interaction of 2-halogenated dATP analogues (F, Cl, and
Br) with human DNA polymerases, DNA primase, and ribo-
nucleotide reductase. Mol. Pharmacol. 1988, 34, 485-491.
(14) Herdewijn, P.; Balzarini, J .; DeClerq, E.; Vanderhaeghe, H.
Resolution of Aristeromycin Enantiomers. J . Med. Chem. 1985,
28, 1385-1386.
Ack n ow led gm en t. We thank the Analytical and
Spectroscopic Laboratory, including Dr. J ames M. Ri-
ordan, Mr. Mark D. Richardson, Ms. Sheila R. Camp-
bell, and Ms. J oan C. Bearden, for analytical data.
Support for this research was provided by the National
Cancer Institute, National Institutes of Health, through
Grant P01 CA34200.
(15) Hershko, A.; Mamont, P.; Shields, R.; Tompkins, G. M. Pleiotypic
response. Nature New Biol. 1971, 232, 206-211.
(16) Bennett, L. L., J r; Smithers, D.; Hill, D. L.; Rose, L. M.;
Alexander, J . A. Biochemical properties of the nucleoside of
3-amino-1,5-dihydro-5-methyl-1,4,5,6,8-pentaazaacenaphthyl-
ene (NSC-154020). Biochem. Pharmacol. 1978, 27, 233-241.
(17) Parker, W. B.; Shaddix, S. C.; Rose, L. M.; Shewach, D. S.; Hertel,
L. W.; Secrist, J . A., III; Montgomery, J . A.; Bennett, L. L., J r.
Comparison of the mechanism of cytotoxicity of 2-chloro-9-(2-
deoxy-2-fluoro-â-D-arabinofuranosyl)adenine, 2-chloro-9-(2-deoxy-
2-fluoro-â-D-ribofuranosyl)adenine, and 2-chloro-9-(2-deoxy-2,2-
difluoro-â-D-ribofuranosyl)adenine in CEM cells. Mol. Pharmacol.
1999, 55, 515-520.
Refer en ces
(1) For examples, see Wagner, C. R.; Iyer, V. V.; McIntee, E. J .
Pronucleotides: Toward the in vivo delivery of antiviral and
anticancer nucleotides. Med. Res. Rev. 2000, 20, 417-451.
(2) Ettner, N.; Haak, U.; Niederweis, M.; Hillen, W. Synthesis of
8-bromo and 8-azido-2′-deoxyadenosine-5′-O-(1-thiophosphate).
Nucleosides Nucleotides 1993, 12, 757-771.
(3) Montgomery, J . A.; Thomas, H. J . Nucleosides of 8-Azapurines
(ν-Triazolo[4,5-d]pyrimidines. J . Med. Chem. 1972, 15, 305-307.
(4) Tong, G. L.; Lee, W. W.; Goodman, L.; Frederiksen, S. Synthesis
of some 2′-deoxyribosides of 8-azaadenine. Arch. Biochem. Bio-
phys. 1965, 112, 76-81.
(5) Kazimierczuk, Z.; Binding, U.; Seela, F. Synthesis of 8-aza-2′-
deoxyadenosine and related 7-amino-3H-1,2,3-triazolo[4,5-d]-
pyrimidine 2′-deoxyribofuranosides: Stereoselective glycosyla-
tion via the nucleobase anion. Helv. Chim. Acta 1989, 72, 1527-
1536.
(6) J ungmann, O.; Pfleiderer, W. A new efficient method in nucleo-
side synthesis. Tetrahedron Lett. 1996, 37, 8355-8358.
(7) Bhat, C. C. In Synthetic Procedures in Nucleic Acid Chemistry;
Zorbach, W. W., Tipson, H. S., Eds.; J ohn Wiley and Sons: New
York, 1968; Vol. 1, pp 521-522.
(18) Shen, F.; Look, K. Y.; Yeh, Y. A.; Weber, G. Increased Urd kinase
(ATP: Urd 5′-phosphotransferase; EC 2.7.1.48) activity in hu-
man and rat tumors. Cancer Biochem. Biophys. 1998, 16, 1-15.
(19) Paterson, A. R. P.; Kolasssa, N.; Cass, C. E. Transport of
nucleoside drugs in animal cells. Pharmacol. Ther. 1981, 12,
515-536.
(20) Parker, W. B.; Shaddix, S. C.; Chang, C. H.; White, E. L.; Rose,
L. M.; Brockman, R. W.; Shortnacy, A. T.; Montgomery, J . A.;
Secrist, J . A., III; Bennett, L. L., J r. Effects of 2-chloro-9-(2-
deoxy-2-fluoro-â-D-arabinofuranosyl)adenine on K562 cellular
metabolism and the inhibition of human ribonucleotide reduc-
tase and DNA polymerases by its 5′-triphosphate. Cancer Res.
1991, 51, 2386-2394.
J M020107S