Novel halogenated nitrobenzylthioinosine analogs as es nucleoside transporter inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.02.016

Nucleoside transporter inhibitors have potential therapeutic applications as anticancer, antiviral, cardioprotective, and neuroprotective agents. We have synthesized and flow cytometrically evaluated the binding affinity of a series of novel halogenated nitrobenzylthioinosine analogs at the human es nucleoside transporter. Structure-activity relationships indicate the importance of hydrophobicity and electron withdrawing capacity of substituents at the para-position of the 6-position benzyl substituent. All of the compounds showed high binding affinity as shown by their ability to displace the fluorescent es transporter ligand, SAENTA-X8-fluorescein. Compound 16 (6-S-(para-iodobenzyl)-6- thioinosine) was the most tightly bound within the series with a K_i of 3.88nM (NBMPR exhibited a K_i of 0.70nM). This compound has higher affinity than the widely used nonnucleoside, nucleoside transport inhibitor, dipyridamole (K_i=8.79nM), and may serve as a new lead compound.

Full text of DOI:10.1016/j.bmcl.2004.02.016

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2257–2260
Novel halogenated nitrobenzylthioinosine analogs as es nucleoside
transporter inhibitors
Amol Gupte and John K. Buolamwini*
Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Sciences Center,
847 Monroe Avenue Suite 327, Memphis, TN 38163, USA
Received 22 August 2003; accepted 3 February 2004
Abstract—Nucleoside transporter inhibitors have potential therapeutic applications as anticancer, antiviral, cardioprotective, and
neuroprotective agents. We have synthesized and flow cytometrically evaluated the binding affinity of a series of novel halogenated
nitrobenzylthioinosine analogs at the human es nucleoside transporter. Structure–activity relationships indicate the importance of
hydrophobicity and electron withdrawing capacity of substituents at the para-position of the 6-position benzyl substituent. All of the
compounds showed high binding affinity as shown by theirability to displace the fluorescent es transporter ligand, SAENTA-X8-
fluorescein. Compound 16 (6-S-(para-iodobenzyl)-6-thioinosine) was the most tightly bound within the series with a Ki of 3.88 nM
(NBMPR exhibited a Ki of 0.70 nM). This compound has higher affinity than the widely used nonnucleoside, nucleoside transport
inhibitor, dipyridamole (Ki ¼ 8:79 nM), and may serve as a new lead compound.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Nucleoside transporters are integral membrane glyco-
proteins that regulate the uptake of physiological
NO2
S
CH2CH2OH
CH2CH2OH
N
N
N
N
1;2
N
N
N
N
nucleosides and theirsynthetic analogs in cells.
The
N
HOH2CH2C
HOH2CH2C
HO
N
two major families of nucleoside transporters are the
equilibrative transporters (ENTs) that are ubiquitously
distributed in mammalian tissues, and the concentrative
sodium-ion linked transporters (CNTs), which are
especially abundant in the intestine, kidney, liver, and
certain cultured cell lines. Two main equilibrative
transporters have been characterized, the es (equilibra-
tive sensitive) nucleoside transporter, also known as
ENT1 and the ei (equilibrative insensitive) nucleoside
transporter, also known as ENT2. These two
transporters differ on the basis of their sensitivity or
insensitivity to inhibition by S6-(4-nitrobenzyl)-mer-
captopurine riboside (NBMPR, 1), respectively. How-
ever, both of them are inhibited by the nonnucleoside,
dipyridamole (2). Both NBMPR and dipyridamole are
N
O
N
OH OH
Dipyridamole, 2
NBMPR, 1
Nucleoside transporter inhibitors have been shown to
have potential therapeutic applications in antimetabolite
chemotherapy in cancer, viral infections, and AIDS-
related opportunistic infections such as Toxoplasma
gondii,3 in inflammatory disease6 and as cardioprotective
and neuroprotective agents.7–11 Studies involving cellular
uptake and efficacy of anticancerand antiHIV nucleo-
side analogs has also generated interest in nucleoside
transporters. The ubiquitously distributed es transporter
appears as the most important nucleoside transporter of
most mammalian tissues and hence it might be the most
relevant therapeutic target. Toxicity, lack of in vivo
efficacy and/orlack of selectivity have hampeerd the
therapeutic application of the currently available es
nucleoside transporter inhibitors.3 Hence, there is a need
for developing novel nucleoside transporter inhibitors.
known to inhibit the es nucleoside transporter at low
3
nanomolarconcenatrtions.
Two new equilibrative
transporters, known as ENT3 and ENT4 has been
recently identified but not fully characterized.4;5
Quite a few substituents on the S6-benzyl position of
NBMPR have been studied.12 However, largely missing
* Corresponding author. Tel.: +1-901-448-7533; fax: +1-901-448-6828;
e-mail: jbuolamwini@utmem.edu
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.016

2258
A. Gupte, J. K. Buolamwini / Bioorg. Med. Chem. Lett. 14 (2004) 2257–2260
Table 1. Flow cytometrically-determined Ki values
from this group of substituents, are the halogens. This
paper reports the synthesis and structure–activity rela-
tionship (SAR) studies of new halogen-substituted
NBMPR analogs as es transporter ligands. The experi-
ments were carried out in order to study the es nucleo-
side transporter binding affinity of the 6-S-(ortho, meta,
or para-position halogen substituted benzyl) thioinosine.
A systematic substitution involving the fourhalogens,
fluorine, chlorine, bromine, and iodine was carried out.
A comparison was made between these new halogen
analogs and the es nucleoside transporter inhibitors,
NBMPR, and dipyridamole.
Compound
X (Halogen substituent position) Ki (nM)
1
0.70
8.79
2
5
2-Bromo
3-Bromo
4-Bromo
2-Chloro
3-Chloro
4-Chloro
2-Fluoro
3-Fluoro
4-Fluoro
2-Iodo
134.23
20.54
9.98
6
7
8
99.49
15.96
14.23
19.87
11.21
9.02
9
10
11
12
13
14
15
16
16.13
13.79
3.88
Although none of the compounds appears to be as
active as NBMPR (Kd ¼ 0:1–1:0 nM), quite a few of
them exhibit substantial affinity, with Ki values in the
low nanomolarconcentrations. The most active among
the twelve was the para-iodo substituent compound (16),
which has a greater binding affinity than the nonnu-
cleoside analog, dipyridamole. We present the synthesis
and flow cytometry investigation of the SAR of these
new halogenated es transporter inhibitors.
3-Iodo
4-Iodo
A flow cytometric competitive binding assay was em-
ployed to study the es nucleoside transporter binding
affinity of the synthesized compounds. The advantages
of flow cytometry over conventional radioligand binding
assays are that, it eliminates the hazard of radiation
exposure and disposal problems associated with radio-
active waste, as well as allows the assay to be performed
using much fewercells. With flow cytometry we can use
as few as 5000 cells as compared to 2 million cells for
comparable radioligand assay. The cell line used to carry
out the assay was the human chronic myelogenous leu-
kemia, K562 cell line. A fluorescent ligand, 5-(SA-
ENTA)-X8-fluorescein, which has high affinity for the es
transporter,13 was used in competition with the synthe-
sized compounds14 to carry out the competitive binding
assay.15
Structures and synthetic approach are illustrated in
Scheme 1. Alkylation of commercially available 6-mer-
captopurine riboside (3) with substituted benzyl halides
(4) in dimethylformamide (DMF), in the presence of
potassium carbonate, gave S-benzylated derivatives of
6-mercaptopurine riboside (5–16).12 The structures and
purities of the compounds were confirmed using NMR,
mass spectroscopy, and elemental analysis.
Compounds synthesized forthe study aer as shown
below.
Within this new series of 6-S-(halogen substituted benz-
yl) thioinosine analogs, compounds 5–16 exhibited a
wide range of binding affinities toward the es nucleoside
transporter. The same was measured by their ability to
displace the es nucleoside transporter specific ligand,
shown by the Ki values in Table 1.
Compound
X (Halogen substituent
position)
5
6
7
8
2-Bromo
3-Bromo
4-Bromo
2-Chloro
3-Chloro
4-Chloro
2-Fluoro
3-Fluoro
4-Fluoro
2-Iodo
9
Table 1 shows the inhibitory constant values of the
compounds ranging from a low nanomolar concentra-
tion forcompounds 13 and 16 to high nanomolar
concentrations for compounds 5 and 8 with at least, a
30-fold difference in binding affinity between the most
active and the least active compounds. A common trend
in binding affinity among the isomers in each set for all
the halogen substituents is readily apparent, that is
para>meta>ortho. This is consistent with the reported
10
11
12
13
14
15
16
3-Iodo
4-Iodo
3
SAR fornitro substitution as well. Of the new analogs,
compound 16, which has a para-iodo substituent,
proved to be better than, the widely used nonnucleoside
nucleoside transport inhibitor dipyridamole as the
ligand of the es transporter. Another analog, compound
13, which has a para-fluoro substituent, exhibited a
binding affinity comparable to that of dipyridamole.
X
S
SH
N
N
N
N
HO
N
X
N
a
O
HO
N
+
N
O
X'
It has been recognized that electron-withdrawing sub-
stituents are preferred in the para-position forhigh
affinity binding of NBMPR analogs at the es trans-
porter.3 However, the effect of hydrophobicity of the
OH OH
OH OH
3
4
5-16
Scheme 1. Reaction condition: (a) Potassium carbonate/ DMF.

A. Gupte, J. K. Buolamwini / Bioorg. Med. Chem. Lett. 14 (2004) 2257–2260
2259
3. Buolamwini, J. K. Curr. Med. Chem. 1997, 4, 35.
4. Hyde, R. J.; Cass, C. E.; Young, J. D.; Baldwin, S. A. Mol.
Membr. Biol. 2001, 18, 53.
5. Baldwin, S. A.; Beal, P. R.; Yao, S. Y.; King, A. E.; Cass,
C. E.; Young, J. D. Pflugers Arch. 2003, June 28 web
publication.
6. Ohta, A.; Stkovski, M. Nature 2001, 414, 916.
7. Abd-Elfattah, A. S.; Wechsler, A. S. J. Card. Surg. 1994,
9, 387.
8. Abd-Elfattah, A. S. A.; Hoehner, J.; Wechsler, A. S. Mol.
Cell. Biochem. 1998, 180, 105.
9. Van Belle, H. Cardiovasc. Res. 1993, 27, 68.
10. Deckert, J.; Morgan, P. F.; Marangos, P. J. Life Sci. 1988,
42, 1331.
11. Geiger, J. D.; Parkinson, F. E.; Kowaluk, E. A. In
Purinergic Approaches in Experimental Therapeutics; Jac-
obson, K. A., Jarvis, M. F., Eds.; Wiley–Liss: New York,
1997; pp 55–84.
12. Robins, M. J.; Asakura, J.; Kaneko, M.; Shibuya, S.;
Jakobs, E. S.; Agbanyo, F. R.; Cass, C. E.; Paterson, A. R.
P. Nucleos. Nucleot. 1994, 13, 1627.
13. Buolamwini, J. K.; Wiley, J. S.; Robins, M. J.; Craik, J.
D.; Cass, C. E.; Gati, W. P.; Paterson, A. R. P. Nucleos.
Nucleot. 1994, 13, 737.
NBMPR
p-iodo
100
50
0
p-bromo
p-chloro
p-fluoro
DPM
1
10
100
1000
10000
Inhibitor Concentration (nM)
Figure 1. Equilibrium displacement of SAENTA-fluorescein ligand by
new 6-S-(halogen substituted benzyl) thioinosine analogs of NBMPR
in K562 cells. Cells were incubated with 30 nM SAENTA-fluorescein
in the presence or absence of inhibitor for 45 min at room temperature
and analyzed by flow cytometry (FACSCaliburꢂ). Data was collected
on 5000 cells per sample, and mean channel numbers were used as a
measure of fluorescence output from ligands. DPM ¼ dipyridamole.
14. To a solution of 6-mercaptopurine riboside (200 mg,
0.70 mmol) in DMF (2.5 mL) was added the correspond-
ing halogen substituted benzyl halides (0.84 mmol) and
potassium carbonate (1 g, 7.23 mmol). The mixture was
stirred overnight at ambient temperature. The mixture was
evaporated in vacuo at a temperature below 45 ꢁC to
remove the DMF. The residue was dissolved in water and
extracted with three 75 mL portions of ethyl acetate. The
combined organic layer was washed once with brine, dried
overmagnesium sulfate, and evapoarted in vacuo. The
crude product was purified using silica gel chromatogra-
phy followed by recrystallization from methanol, to afford
the pure compounds.
15. The es nucleoside transporter binding ability of the
compounds was evaluated using a flow cytometric assay.13
Human leukemia K562 cells, grown in RPMI 1640
medium were washed once and resuspended at
1.6 · 105 cells/mL in phosphate buffered saline at pH 7.4,
and incubated with 5-(SAENTA)-X8-fluorescein (30 nM)
in the presence or absence of varying concentrations of the
test compounds at room temperature for 45 min. Flow
cytometric measurements of cell associated fluorescence
were then performed with a FACSCalibur (Becton Dick-
inson, San Jose, CA) equipped with a 15 mW-argon laser
(Molecular Resources Flow Cytometry Facility, Univer-
sity of Tennessee Health Sciences Center). In each assay,
5000 cells were analyzed from suspensions of 5 · 105 cells/
mL. The units of fluorescence were arbitrary channel
numbers. Percentage (%) of control (i.e., es transporter-
specific fluorescence in the presence of SAENTA-fluores-
cein without test compounds) was calculated foreach
sample using Eq. 1.
analogs has not been well examined. The present results
indicate that both increasing hydrophobicity and elec-
tron withdrawing capacity favor es transporter binding.
The results obtained in the present study do support that
notion. The optimum combination of these properties
among the halogens in this regard appears to reside in
the para-iodo compound (16). The enhanced affinity of
the fluoro compound (13) relative to the bromo (7) and
chloro (10) compounds stems possibly from the higher
electron withdrawing ability of the fluorine substituent.
This information coupled with information regarding
the bioactive conformation of NBMPR that we are
gathering from conformationally restricted analogs16
should be useful in ourefforts at modeling es nucleoside
transporter inhibitors. The substantial binding affinity
of some of these new halogen-substituted NBMPR
analogs, especially compound 16 (Ki 3.88 nM), marks
them as new potential leads fordeveloping better es
transporter inhibitors for therapeutic applications.
Acknowledgements
Financial support from the American Heart Association
Southeast Affiliate Grant No 9950872V is hereby
acknowledged.
% control ¼ ðSFsÞ Â 100=ðSFf Þ
ð1Þ
where SFs is the es transporter-specific fluorescence of test
samples, and SFf , is the es transporter-specific fluorescence
of the SAENTA-fluorescein ligand standard in mean
channel numbers. The results obtained were entered in the
PRISM program (GraphPad, San Diego, CA) to derive the
concentration dependent curves (examples are shown in
Fig. 1). From these curves, the IC50 values were computed
and used to calculate inhibition constants (Ki) values from
Eq. 2.
References and notes
1. Cass, C. E. In Drug Transport in Antimicrobial and
Anticancer Chemotherapy; Georgopapadakou, N. H.,
Ed.; Marcel Dekker: New York, 1995; pp 403–451.
2. Young, J. D.; Cheeseman, C. I.; Mackey, J. R.; Cass, C.
E.; Baldwin, S. A. In Current Topics in Membranes;
Barrett, K. E., Donowitz, M., Eds.; Academic: San Diego,
CA, 2000; Vol. 50, pp 329–378.
Ki ¼ IC50=ð1 þ ½Lꢀ=KLÞ
ð2Þ

2260
A. Gupte, J. K. Buolamwini / Bioorg. Med. Chem. Lett. 14 (2004) 2257–2260
X8-fluorescein, and for that matter their affinity for the es
where, ½Lꢀ and KL are the concentration and the Kd value
transporter.
16. Zhu, Z.; Furr, J.; Buolamwini, J. K. J. Med. Chem. 2003,
46, 831.
of SAENTA-fluorescein, respectively. The Ki values were
used to compare the abilities of the new compounds to
displace the es transporter-specific ligand 5-(SAENTA)-