A novel noninvasive method for assessing glutathione-conjugate efflux systems in the brain

Article abstract of DOI:10.1016/j.bmc.2007.02.045

Brain efflux systems export such conjugated metabolites as glutathione (GSH) and glucuronate conjugates, generated by the detoxification process, from the brain and serve to protect the brain from harmful metabolites. The intracerebral injection of a radiolabeled conjugate is a useful technique to assess brain efflux systems; however, this technique is not applicable to humans. Hence, we devised a novel noninvasive approach for assessing GSH-conjugate efflux systems using positron emission tomography. Here, we investigated whether or not a designed proprobe can deliver its GSH conjugate into the brain. Radiolabeled 6-chloro-7-methylpurine (7m6CP) was designed as the proprobe, and [¹⁴C]7m6CP was prepared by the reaction of 6-chloropurine with [¹⁴C]CH₃I as a model of [¹¹C]CH₃I. The radiochemical yield and purity of [¹⁴C]7m6CP were 10-20% and greater than 99%, respectively. High brain uptake (0.8% ID/g) at 1 min was observed, followed by gradual radioactivity clearance from the brain for 5-60 min after the injection of [¹⁴C]7m6CP into rats. Analysis of metabolites confirmed that the presence of [¹⁴C]7m6CP was hardly observed, and 80% of the radioactivity was identical to its GSH conjugate for 15-60 min. The brain radioactivity was single-exponentially decreased during the period of 15-60 min post-injection of [¹⁴C]7m6CP, and the first-order efflux rate constant of the conjugate, estimated from the slope, was 0.0253 min^-1. These results showed that (1) [¹⁴C]7m6CP readily entered the brain, (2) it efficiently and specifically transformed to the GSH conjugate within the brain, and (3) after [¹⁴C]7m6CP disappearance, the clearance of radioactivity represented the only efflux of GSH conjugate. We conclude that 7m6CP can deliver the GSH conjugate into the brain and would be useful for assessing GSH-conjugate efflux systems noninvasively.

Full text of DOI:10.1016/j.bmc.2007.02.045

Bioorganic & Medicinal Chemistry 15 (2007) 3127–3133
A novel noninvasive method for assessing glutathione-conjugate
efflux systems in the brain
a
a
a
b
Toshimitsu Okamura, Tatsuya Kikuchi, Kiyoshi Fukushi, Yasushi Arano
a,
*
and Toshiaki Irie
a
Probe Research Section, Department of Molecular Probe, Molecular Imaging Center, National Institute of Radiological Sciences,
-9-1 Anagawa, Inage-ku, Chiba-shi, Chiba 263-8555, Japan
4
Graduate School of Pharmaceutical Sciences, Chiba University, Chiba 263 8522, Japan
b
Received 11 December 2006; revised 20 February 2007; accepted 21 February 2007
Available online 23 February 2007
Abstract—Brain efflux systems export such conjugated metabolites as glutathione (GSH) and glucuronate conjugates, generated by
the detoxification process, from the brain and serve to protect the brain from harmful metabolites. The intracerebral injection of a
radiolabeled conjugate is a useful technique to assess brain efflux systems; however, this technique is not applicable to humans.
Hence, we devised a novel noninvasive approach for assessing GSH-conjugate efflux systems using positron emission tomography.
Here, we investigated whether or not a designed proprobe can deliver its GSH conjugate into the brain. Radiolabeled 6-chloro-7-
1
4
methylpurine (7m6CP) was designed as the proprobe, and [ C]7m6CP was prepared by the reaction of 6-chloropurine with
1
4
11
14
[
tively. High brain uptake (0.8% ID/g) at 1 min was observed, followed by gradual radioactivity clearance from the brain for
C]CH
3 3
I as a model of [ C]CH I. The radiochemical yield and purity of [ C]7m6CP were 10–20% and greater than 99%, respec-
1
4
–60 min after the injection of [ C]7m6CP into rats. Analysis of metabolites confirmed confirmed that the presence of
5
[
1
4
C]7m6CP was hardly observed, and 80% of the radioactivity was identical to its GSH conjugate for 15–60 min. The brain radio-
1
4
activity was single-exponentially decreased during the period of 15–60 min post-injection of [ C]7m6CP, and the first-order efflux
rate constant of the conjugate, estimated from the slope, was 0.0253 min . These results showed that (1) [ C]7m6CP readily
entered the brain, (2) it efficiently and specifically transformed to the GSH conjugate within the brain, and (3) after [ C]7m6CP
disappearance, the clearance of radioactivity represented the only efflux of GSH conjugate. We conclude that 7m6CP can deliver
the GSH conjugate into the brain and would be useful for assessing GSH-conjugate efflux systems noninvasively.
À1
14
1
4
ꢀ
2007 Elsevier Ltd. All rights reserved.
1
. Introduction
MRP1 protein expression and adducts MRP1 bound
to 4-hydroxy-2-transnonenal (a lipid peroxidation prod-
uct) in frozen hippocampal samples from Alzheimer’s
There are active efflux transporters at the blood–brain
barrier (BBB) and/or blood–cerebrospinal fluid barrier,
and they can export potentially harmful metabolites
and xenobiotics from the brain. Hydrophilic conju-
gated metabolites such as glutathione (GSH) and glucu-
ronate conjugates, generated by detoxification of
xenobiotics, are also eliminated from the brain via efflux
systems. Currently, these conjugates are known to be
good substrates for the multidrug resistance-associated
protein (MRP) family involved in brain efflux systems.
Sultana and Butterfield observed a slight increase in
6
patients versus age-matched control subjects. Several
therapeutic drugs such as probenecid and indomethacin
1
,2
7,8
are known to be a MRP1 inhibitor. A noninvasive
assessment of brain efflux systems in human studies will
be useful for elucidating pathological conditions of neu-
rodegenerative disorders, and directly evaluating the
pharmacological effect of therapeutic drug on brain
efflux systems. The efflux rate constant, reflecting brain
efflux systems, can be calculated from a time course
measurement of brain concentration following the intra-
cerebral injection of a conjugate. Indeed, using the con-
jugate of radiolabeled 17b-estradiol-D-17b-glucuronide
3
4
,5
(
E 17bG), efflux systems involved in the transport of
2
2
Keywords: Active efflux transporter; Glutathione conjugate; Metabo-
lite Extrusion Method; Purine analogues.
9
E 17bG across the BBB were investigated. The intrace-
*
Corresponding author. Tel.: +81 43 206 3191; fax: +81 43 206
261; e-mail: t_irie@nirs.go.jp
rebral injection of a radiolabeled conjugate is a useful
technique to study mechanisms of brain-to-blood efflux;
3
0
968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.02.045

3128
T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
however, this invasive method is not applicable to
human study. We therefore devised a novel method,
referred to as the Metabolite Extrusion Method
provides an accurate keff because of the simple parame-
ter estimation of the single exponential curve, as cerebral
blood flow can be measured by the accurate estimation
of the clearance rate of brain radioactivity following
(
MEM), for the noninvasive assessment of GSH-conju-
1
0,11
gate efflux systems in living human brains using positron
emission tomography (PET). Positron emitters are not
convenient to perform basic experiments because of a
very short half-life. Here, we studied whether or not
proprobes designed according to MEM can noninva-
sively deliver their GSH conjugate into the brain using
the administration of radioactive xenon.
MEM has
the merit that keff can be estimated without the need
for arterial blood sampling, which requires considerable
technical expertise.
Our goal is to deliver a GSH conjugate into the brain
for the noninvasive assessment of GSH-conjugate efflux
systems. To achieve this, the proprobe must possess
three properties: high BBB permeability and efficient
conversion into the single probe (permeability, effi-
ciency, and specificity). The probe should not penetrate
the BBB by simple diffusion. Furthermore, it is desir-
able that the proprobe platform can be easily modified
for the introduction of a lipophilic group and radiola-
beling. 6-Chloropurine is reported to be converted to
S-(6-purinyl)glutathione by hepatic and renal glutathi-
1
4
11
rats and C nuclide (half-life: 5730 yr) in place of
nuclide for PET (half-life: 20.4 min).
C
2
. Results and discussion
2
.1. Design and chemistry
The rationale for MEM is illustrated in Figure 1a and b.
A radiolabeled lipophilic compound, a proprobe, is
administered intravenously. After injection, the prop-
robe enters the brain across the BBB by simple diffusion
1
2
one S-transferase (GST) in vivo. The rat brain has
a GSH concentration of approximately 2 mM, and
1
3,14
GST is markedly expressed in the brain.
The adap-
(
fuses back into the blood (k ), the rest is rapidly
K ). While a portion of the incorporated proprobe dif-
1
tation of purine as a proprobe platform allows easy
chemical modification. Given these aspects, we selected
2
converted to the probe, GSH conjugate (k ). If the con-
3
6
The methylation of 6-chloropurine with [ C]CH I as
-chloropurine as the basic skeleton of the proprobe.
version in the peripheral tissues is also rapid, the prop-
robe no longer exists in either the blood or the brain
with a short lag time after the injection, in which case
the situation of Figure 1a changes to that of Figure
14
3
11
a model for [ C]CH I was performed to increase
3
lipophilicityand introduceradioactivity. Asaconsequence,
two potential proprobes for the delivery of a GSH conju-
gate into the brain were generated: 6-chloro-9-[ C]meth-
1b. If passive transport of the GSH conjugate across
the BBB is negligible, the efflux rate of the GSH conju-
14
1
4
14
ylpurine ([ C]9m6CP) and 6-chloro-7-[ C]methylpurine
gate (k ) can be estimated noninvasively. MEM also
eff
a
BBB
K1
b
BBB
Blood
Brain
Blood
Brain
k2
carrier
carrier
k3
keff
keff
X
X
c
GS-
Cl
N
*
GS
*
CH3
N
CH3
N
GST
N
N
N
N
N
Radiolabeled Proprobe
Glutathione Conjugate
Figure 1. Diagrammatic representation of the Metabolite Extrusion Method (a and b). A proprobe (open circles) is intravenously administered. The
proprobe can penetrate the BBB and transform to a probe of GSH conjugate (circle-squares) in the brain. The probe is extruded by efflux systems
from the brain to the blood. After the proprobe is completely converted to the probe and no longer exists in either the brain or the blood, the
situation of Fig. 1a changes to that of Fig. 1b, similar to the state obtained by intracerebral injection. K
1 2
and k are the rate parameters illustrating
transport of the proprobe across the BBB by diffusion, while k and keff are the rate parameters indicating the GSH conjugation and transport of the
3
probe by efflux systems, respectively. A proprobe designed to deliver the radiolabeled GSH conjugate into the brain (c). Of two isomers generated by
À
the methylation of 6-chloropurine, 6-chloro-7-methylpurine is shown in the figure. GS denotes the thiolate anion of GSH.

T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
3129
1
4
(
[
[
[ C]7m6CP) (Fig. 1c). Radiochemical yield andpurity of
raphy (TLC). The radioactivity corresponding to the un-
changed form almost completely disappeared at 15 min
post-injection of [ C]7m6CP, while for [ C]9m6CP,
approximately 80% of radioactivity remained in the un-
changed form (Fig. 3b). The in vivo results could be rea-
1
4
4
C]9m6CP were 50% and 99%, respectively. Those of
C]7m6CP were 10–20% and 99%, respectively. The
1
14
14
production rate for isomers is similar to an early report
that the alkylation of 6-chloropurine with alkyl halide
formed major products (9-isomers) and minor products
sonably explained by differences in the in vitro k
3
1
5
14
14
14
(
and methyl group probably leads to the reduction of
7-isomers). The steric hindrance between chlorine atom
between [ C]7m6CP and [ C]9m6CP. [ C]9m6CP
would mostly diffuse back into the blood without reac-
tion with GSH because of its slower reaction rate. In
1
4
[
C]7m6CP formation.
1
4
contrast, [ C]7m6CP (high in vitro k ) appears to be
3
2.2. Verification of requirements (permeability, efficiency,
and specificity)
rapidly converted to a radioactive metabolite, which is
subsequently extruded from the brain to the blood. In
other words, the decrease in radioactivity following the
1
4
For reliable estimation of k , each proprobe should
eff
injection of [ C]9m6CP implies mainly the washout of
the intact proprobe from the brain, whereas that of
[ C]7m6CP means the washout of the radioactive
possess physiochemical properties required for MEM:
permeability, efficiency, and specificity. Figure 2a shows
brain uptake curves following the intravenous injection
1
4
1
4
metabolite. As [ C]9m6CP cannot sufficiently deliver
the GSH conjugate into the brain, it is abandoned as a
proprobe for MEM in this study.
1
4
14
of [ C]7m6CP and [ C]9m6CP into rats. They exhibit
a similar high uptake (0.7–0.8% ID/g) at 1 min after
injection, and then a gradual decrease in brain radioac-
tivity from 5 min. In general, the delivery of a low
molecular substance into the brain is not so difficult as
long as that substance is a lipophilic one. Indeed, these
proprobes showed a similar high uptake at 1 min after
injection, which is comparable to the 1 min uptake of
1
4
[ C]7m6CP readily entered the brain, and it was rapidly
metabolized within the brain. Further analysis of the
chemical form was performed by HPLC at 15 and
60 min after injection as shown in Figure 4. HPLC
analysis shows that approximately 80% of radioactivity
was identified as S-[6-(7-methylpurinyl)]glutathione
(7mPSG) (retention time of authentic sample,
15.3 min) both at 15 and 60 min post-injection. The
peak height corresponding to 7m6CP (17.3 min) was
very low at 15 min post-injection. Brain radioactivity
1
6
a clinically used radiopharmaceutical, and they had
the high BBB permeability. However, the two proprobes
showed quite different kinetics thereafter: while
1
4
[
C]9m6CP exhibited a gradual elimination of radioac-
1
4
tivity, [ C]7m6CP displayed clearance after temporary
accumulation. This is likely due to the enzymatic con-
version rates of two proprobes to GSH conjugates via
1
4
principally existed in the form of [ C]7mPSG during
1
4
the period of 15–60 min after [ C]7m6CP injection,
although some unknown peaks slightly increased with
time, which might be metabolites formed from the
GSH conjugate by metabolic enzymes such as c-glutam-
GST in the brain, that is, k in Figure 1a. In order to
3
examine this, the reaction rates were determined in rat
brain homogenates.
1
7
yltranspeptidase. In addition to the efficiency of rapid
1
4
Figure 3a shows the first-order rate constant (in vitro k3)
at a unit concentration of brain tissue. The rates of
conversion into the GSH conjugate, [ C]7m6CP pos-
sessed specificity that means the production of the single
probe within the brain.
1
4
14
C]7m6CP and [ C]9m6CP were estimated as
[
À1 À1 À1
mL ),
respectively. To confirm that proprobes are metabolized
0
.33 ± 0.01 and 0.0057 ± 0.0002 (min
g
The proprobe is essential for the delivery of radioactiv-
ity into the brain; however, it should be cleared from the
blood as shown in Figure 1b. HPLC and TLC analyses
of chemical form demonstrated that the amount of
in the brain according to the in vitro k , the fraction of
3
each unchanged form in the brain was analyzed 15 min
after intravenous injection using thin layer chromatog-
1
.8
.6
.4
1
1
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
1.2
1
0
.0
.8
0.6
0
0
.4
.2
0.0
0
10 20 30 40 50 60
Time after injection (min)
0
10
20
30
40
50
60
Time after injection (min)
1
Figure 2. Time–radioactivity curves in the rat brain (a) and blood (b) after intravenous injection of [ C]7m6CP (closed circles), [ C]9m6CP (open
4
14
3
circles), and [ S]7mPSG (closed squares). Data are given as means ± the standard deviation (n = 3 rats for each point), and brain uptakes are
5
corrected for the vascular volume of 3%, based on the radioactivity of the whole blood at each time-point. ID, injected dose.

3130
T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
0
0
.4
.3
0
.33
1
00
8
0
0
0
.2
.1
60
4
2
0
0
0
0
.0057
0
7
m6CP
9m6CP
7m6CP
9m6CP
Figure 3. (a) The reaction rate of proprobes with GSH in rat brain homogenate. Data are given as means ± standard deviation. (b) Analysis of the
1
4
14
chemical form in the rat brain by TLC at 15 min after the intravenous injection of [ C]7m6CP (left) and [ C]9m6CP (right). White and black areas
denote the fraction of each proprobe and metabolite, respectively. Data are shown from a single experiment on each compound.
8
6
4
2
0
0
0
0
containing human MRP1 and MRP1-defective vesicles
as controls. As shown in Figure 5, membrane vesicles
containing MRP1 showed a significant time- and ATP-
7
mPSG
1
5 min
60 min
3
5
dependent uptake of [ S]7mPSG; uptake of the
compound into control vesicles was very low in both
the absence and presence of ATP. This result suggested
that MRP1 participated in the 7mPSG efflux from the
brain. Although 7mPSG is a good substrate for
MRP1, we cannot exclude possibilities that other efflux
transporters such as MRP isoforms and P-glycoprotein
contribute to the transport of 7mPSG. Moreover, the
coordination of uptake transporters such as organic an-
ion transporters and efflux transporters seems to be
involved in the elimination of 7mPSG from the site with-
unknown
7
m6CP
0
0
5
10
15
20
Fraction number
Figure 4. Analysis of the chemical form in rat brain. HPLC analysis
shows that approximately 80% of radioactivity was identified as
7
mPSG at 15 (diamond) and 60 min (triangle) following the admin-
in the brain to the blood, similar to the efflux transport
9
1
4
istration of [ C]7m6CP. The peak corresponding to 7m6CP was very
low at 15 min. The retention times of authentic 7mPSG and 7m6CP
are 15.3 and 17.3 min, respectively. Data are shown from a single
experiment on each time-point.
of E 17bG (a glucuronate conjugate). Since MRP1 is
2
3
also expressed at the choroid plexus, 7mPSG is possibly
extruded via the blood–cerebrospinal fluid barrier. Here,
brain kinetics and chemical form analysis showed that
the GSH conjugate (7mPSG) was extruded from the
brain during the period of 15–60 min after the proprobe
injection. Uptake values during that period were plotted
in a semilog scale. As a result, the semilog plot exhibited
linearity (multiple correlation coefficient: R = 0.99),
1
4
[ C]7m6CP in the brain was negligible from 15 min
after the injection (Figs 3b and 4), implying that there
is no additional delivery of [ C]7m6CP into the brain
from blood. This rapid clearance would be attributed
to the metabolism by GSH/GST in the blood or periph-
eral tissues.
1
4
1
,000
800
600
400
200
0
To confirm that 7mPSG cannot move into and out of
the brain by simple diffusion, brain uptake was mea-
3
5
sured following the intravenous injection of [ S]7mPSG
into rats (Fig. 2a). The brain radioactivity of
3
1
5
4
[
[
S]7mPSG was much lower than that of the proprobe,
C]7m6CP, during the period of 1–60 min. This result
verified that 7mPSG could not penetrate the BBB by
simple diffusion. Nevertheless, it was eliminated from
the brain, indicating that the carrier-mediated transport
is involved in the extrusion.
0
1
2
3
4
Time (min)
Various transporters probably contribute to the 7mPSG
efflux from the brain; however, GSH conjugates are well
known to be good substrates for MRP1 that is involved
in brain efflux systems. The possibility for the involve-
ment of MRP1 in the 7mPSG efflux was examined by a
rapid filtration technique using membrane vesicles
35
Figure 5. Transport of [ S]7mPSG by MRP1. Membrane vesicles
containing MRP1 (circles) or MRP1-defective vesicles (squares) were
4
,5
3
5
incubated at 37 ꢁC with 50 lM [ S]7mPSG in the presence of 4 mM
ATP (closed symbols) or AMP (open symbols). Each point and bar is
the means ± standard deviation of three experiments in duplicate.

T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
3131
indicating that 7mPSG was single-exponentially ex-
truded from the brain. The first-order efflux rate con-
stant (min ) of 7mPSG, estimated from the slope,
tic sample purchased from Toronto Research Chemicals
Inc. (North York, Ontario, Canada).
À1
1
was 0.025 ± 0.0013 (means ± standard error). Interest-
ingly, this efflux rate estimated by MEM using Wistar
3.2.2. 6-Chloro-9-methylpurine. H-NMR (d, CDCl3):
3.89 (3H, s, NCH ), 8.06 (1H, s, H-8), 8.68 (1H, s, H-
3
À1
+
2); FAB-MS, m/z: 169 (M+H) . Found: 169; mp: 138–
rats is similar to the E 17bG efflux rate (0.028 min
)
estimated by intracerebral injection using Sprague–
2
1
8
142 ꢁC (lit., 143–144 ꢁC).
9
Dawley rats, implying that the same efflux systems
1
3
.2.3. 6-Chloro-7-methylpurine. H-NMR (d, DMSO-
might contribute to the 7mPSG efflux at the BBB. To
investigate the involvement of efflux pathways and other
transporters, further studies will be needed.
d ): 4.07 (3H, s, NCH ), 8.73 (1H, s, H-8), 8.77 (1H, s,
6
3
+
H-2); FAB-MS, m/z: 169 (M+H) . Found: 169; mp:
1
9
1
96–197 ꢁC (lit., 198–199 ꢁC).
In conclusion, 7m6CP was newly designed as the
proprobe to deliver its GSH conjugate, based on the
concept of MEM. 7m6CP could successfully deliver
the conjugate into the brain and appears useful for the
noninvasive assessment of brain GSH-conjugate efflux
systems.
3.2.4. S-[6-(7-methylpurinyl)]glutathione (7mPSG). This
compound was prepared by slight modification of the
method as described previously. 6-Chloro-7-methylpu-
rine (26 mg) and GSH (105 mg) were dissolved in
2
0
760 lL each of 1 N NaOH solution and ethanol. After
the mixture was heated for 4 h at 50 ꢁC in a water bath,
the solvent was evaporated, and ethanol was completely
removed. Then, the concentrated residue was purified by
column chromatography (Cosmosil 75 C –OPN, 1%
3
. Material and methods
1
8
aqueous AcOH) to give 7mPSG as a colorless solid
3
.1. General
1
(
yield: 66%). H-NMR (d, D O): 1.93 (2H, m, Glu-c-
2
CH ), 2.27 (2H, m, Glu-b-CH ), 3.47 (1H, dd, J = 8.1,
All commercially available starting materials and sol-
vents were of reagent grade or better and were used
without further purification. [ C]Methyl iodide (specific
2
2
1
4.2 Hz, Cys-b-CH ), 3.54 (1H, d, J = 17.1 Hz, Gly-
2
1
4
CH ), 3.57 (1H, d, J = 6.7 Hz, Glu-a-CH ), 3.61 (1H,
2
2
3
5
d, J = 17.1 Hz, Gly-CH ), 3.82 (1H, dd, J = 4.6,
activity: 2.15 GBq/mmol) and [ S]GSH (35.4 TBq/
mmol) were purchased from Amersham Biosciences
Corp. (Piscataway, NJ, USA) and PerkinElmer Life Sci-
ences, Inc. (Boston, MA, USA), respectively. Thin layer
chromatography (TLC) plates were purchased from
Merck Ltd (Tokyo, Japan). Membrane vesicles contain-
ing human MRP1, or MRP1-defective vesicles as a con-
trol, were purchased from SOLVO Biotechnology, Inc.
2
1
4.2 Hz, Cys-b-CH ), 3.87 (3H, s, NCH ), 4.70 (1H,
2 3
dd, J = 4.6, 8.1 Hz, Cys-a-CH ), 8.09 (1H, s, H-8),
2
+
.46 (1H, s, H-2); FAB-MS, m/z: 440 (M+H) . Found:
8
440; mp: 157–165 ꢁC (dec).
1
.2.5. Radiochemical synthesis. 6-Chloro-9-[ C]methylp-
4
3
urine ([ C]9m6CP) and 6-chloro-7-[ C]methylpurine
1
4
14
14
[ C]7m6CP) were prepared by N-methylation using
(
(
Budapest, Hungary). Proton nuclear magnetic reso-
1
14
[ C]methyl iodide. To a mixture of 6-chloropurine
(9.3 mg) and potassium carbonate (9.7 mg), a solution
nance ( H-NMR) spectroscopy was performed on a
JEOL JNM-ALPHA 500 FT-NMR (JEOL Ltd, Tokyo,
Japan) with chemical shifts reported in units of parts per
million (ppm). Fast atom bombardment mass spectrom-
etry (FAB-MS) was performed using a JEOL JMS-HX-
1
4
of 9.25 MBq of [ C]methyl iodide in 1 mL of acetone
was added. In a sealed vial, the mixture was heated in
a water bath at 55 ꢁC for approximately 30 min, and
then acetone was removed in a stream of nitrogen.
The residue was dissolved in 200 lL acetone and applied
to TLC with a silica-gel plate. The developing solvents,
1
10A mass spectrometer (JEOL Ltd, Tokyo, Japan).
Radioactivity was measured using a liquid scintillation
counter (LS 6000; Beckman Instruments, Fullerton,
CA, USA), and the relative radioactivity of compounds
on TLC was quantified using an imaging phosphor
plates system (BAS 1800 system; Fuji Photo Film Co.,
Tokyo, Japan).
1
4
14
[ C]9m6CP and [ C]7m6CP, were a mixture of
chloroform/ethanol (9:1; v/v) and ethyl acetate/ethanol
(9:1; v/v), respectively. The radioactive zone with an R
f
value corresponding to the authentic compound was
scratched, collected, and extracted with acetone. Radio-
1
4
14
3
3
.2. Synthesis
chemical yields of [ C]9m6CP and [ C]7m6CP were
0% and 10–20%, respectively. Radiochemical purities
5
.2.1. 6-Chloro-9-methylpurine (9m6CP) and 6-chloro-7-
(99%) were determined by TLC using the solvent
systems as described above.
methylpurine (7m6CP). Methyl iodide (1.7 g) was added
to a solution of 6-chloropurine (930 mg) and potassium
carbonate (830 mg) dissolved in 30 mL of dimethylsulf-
oxide (DMSO). After the reaction mixture was stirred
at room temperature for 24 h, DMSO was evaporated
in vacuo, and then the residue was purified by column
chromatography (silica gel C-200, ethyl acetate/ethanol,
3
5
35
.2.6. [ S]S-[6-(7-methylpurinyl)]glutathione ([ S]7mPSG).
3
3
5
[ S]7mPSG was synthesized enzymatically by the conjuga-
tion of 7m6CP with [ S]GSH using glutathione S-trans-
ferase (GST) from equine liver (purchased from Sigma).
35
35
[
S]GSH (3.7 MBq) was dissolved in 150 lL of phosphate
buffer (0.1 M, pH 7.4), and the solution was extracted with
ethyl acetate to remove excess dithiothreitol. A mixture of
9
:1, v/v). In addition, products were recrystallized from
benzene to give 9m6CP (35%, white solid) and 7m6CP
6.2%, light yellow solid). Each compound was identified
by TLC and comparison of NMR data using an authen-
7m6CP (3 mg) and GST (0.3 mg) dissolved in 100 lL
(
of the phosphate buffer was added to the solution of

3132
T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
3
5
[
S]GSH, and the solution was incubated at 37 ꢁC for 1 h.
Netherlands) was added to blood samples, and then
hydrogen peroxide was added for bleaching. Cerebral
cortexes were dissected, weighed, and then Soluene-350
(Packard Bioscience) was added to dissolve the tissue
at 60 ꢁC, after which radioactivity was measured using
After incubation, the solution was concentrated to a final
volume of ca. 100 lL under a stream of nitrogen, and the
solution to which 50 lL of ethanol was further added was
applied to a TLC plate (RP-18 F254s). The TLC plate was
developed using water/acetonitrile/AcOH (9:1:0.1, v/v/v) as
an eluent. After visualization by autoradiography,
1
4
a liquid scintillation counter. The tissue uptake of
C
3
5
or S activity was expressed as the percentage of in-
jected dose per gram tissue (% ID/g tissue).
35
[
S]7mPSG was scraped from the plate and extracted
from C18 with 50% ethanol in water. The extracted solu-
tion was filtered through a 0.45 lm filter, and analyzed
by HPLC under the conditions described below. Radio-
chemical yield was 50%, and radiochemical purity (97%)
was calculated from the radioactivity in the eluate fraction
corresponding to the authentic sample and the total radio-
activity injected into HPLC.
3
.2.10. Analysis of the chemical form in rat brain. For
TLC analysis, rats were decapitated 15 min after
1
4
14
C]9m6CP or [ C]7m6CP (ca. 2.2 MBq) injection.
[
The head was quickly dipped in liquid nitrogen and
the brain was removed. The brain was homogenized in
5
0% ethanol in water and then centrifuged. The super-
natant was applied to a silica-gel TLC plate and devel-
oped using ethyl acetate/ethanol (9:1; v/v). Then, the
radioactivity on TLC was quantified using BAS. For
HPLC analysis, rats were decapitated at 15 and 60 min
1
4
14
Since these compounds ([ C]7m6CP, [ C]9m6CP, and
3
5
[ S]7mPSG) were prepared without adding carriers,
each specific activity may be considered to be equal to
1
that of the original radioactive compound, CH I or
4
1
4
3
after [ C]7m6CP injection. The brain was treated and
homogenized as described above, and one half-volume
of ethanol was added to the homogenate to be centri-
fuged. The supernatant was concentrated in a stream
of nitrogen, followed by centrifugation and filtration
through a 0.45 lm filter. The concentrated sample was
analyzed by HPLC under the conditions described
below. Both the recovery of radioactivity in HPLC
elutions and the extraction efficiency of that to the
supernatant were more than 95%.
3
5
[
S]GSH.
3
(
.2.7. Animal studies. Wistar rats weighing 185–205 g
male, 8 weeks old) were purchased from Japan SLC
Inc. (Shizuoka, Japan). In the present study, they were
treated and handled according to the ‘Recommenda-
tions for Handling of Laboratory Animals for Biomed-
ical Research’, compiled by the Committee on Safety
and Ethical Handling Regulations for Laboratory
Animal Experiment, NIRS.
3
.2.11. Vesicular transport assay. Transport studies were
3
.2.8. Reaction rates of purine derivatives with GSH in
performed using the rapid filtration technique based on
the assay protocol of SOLVO Biotechnology, Inc. with a
slight modification. Briefly, 73.5 lL of transport med-
ium [40 mM MOPS–Tris, 50 mM KCl, 6 mM MgCl2,
10 lL of membrane vesicle suspension (50 lg of pro-
tein), [ S]7mPSG (3.7 kBq), 50 lM 7mPSG, and
2 mM GSH] was preincubated at 37 ꢁC for 5 min, and
then a transport assay was initiated by the addition of
1.5 lL of Mg–ATP (0.2 M) or Mg–AMP. The transport
reaction was stopped by the addition of 1 mL ice-cold
buffer containing 40 mM MOPS–Tris and 70 mM KCl.
The stopped reaction mixture was filtered through a
GF/F glass fiber filter (Whatman, Maidstone, England)
and then washed twice with 5 mL of the ice-cold buffer.
Radioactivity retained on the filters was determined
using a liquid scintillation counter.
tissue homogenate. Rats were anesthetized with ether
and decapitated. The cortex was removed and homoge-
nized in phosphate buffer (0.1 M, pH 7.4). The homog-
enate (250 lL) supplemented with 2 mM GSH was
placed in tubes and preincubated at 37 ꢁC for 30 min.
3
5
1
4
Each [ C]-purine solution (111 kBq in 20 lL buffer)
was added to each tube to initiate the reaction. At des-
ignated intervals, 25 lL of the homogenate was immedi-
ately added to 50 lL of the stop solution, ethanol
containing excess diethyl maleate (ca. 50 mM), and then
centrifuged. The supernatant (5 lL) was applied to TLC
with a silica-gel plate, and developed with ethyl acetate/
ethanol (9:1; v/v). Then, the air-dried TLC plate was
covered with a 5 lm thick lumilar membrane and placed
in a cassette in contact with an imaging phosphor plate
for several hours. Radioactivity corresponding to the Rf
value of the unchanged purines was quantified using
BAS. The first-order rate constants were calculated by
plotting the change in the log of the radioactivity of
3.2.12. Estimation of efflux rate constant (keff). Rats were
decapitated 15, 30, and 60 min after the intravenous
1
4
injection of [ C]7m6CP, and then cerebral cortexes were
dissected and weighed. The radioactivity in the tissue
was measured as described above. Nine data corrected
for vascular volume (n = 3, each point) were fitted to
the exponential function by the least-square method to
1
4
[
C]-purine derivatives against time, and corrected for
the concentration of the homogenate to obtain the reac-
À1 À1 À1
mL ). All determinations were
tion rates (min
made in triplicate.
g
À1
obtain the efflux rate constant (min ).
3
.2.9. Time–radioactivity curves. Each labeled com-
pound dissolved in saline (111 kBq/200 lL) was injected
via a lateral tail vein. Rats, anesthetized with ether
immediately before decapitation, were decapitated at
3.2.13. HPLC analysis. Unlabeled compounds were
analyzed using an HPLC system consisting of a model
L-7100 pump (Hitachi, Tokyo, Japan), a model
L-7450/L-7400 UV absorbance detector set at 288 nm
(Hitachi, Tokyo, Japan), and a cosmosil 5C18-AR-II
column (4.6 I.D. · 250 mm) (Nacalai Tesque, Kyoto,
1
, 5, 15, 30, and 60 min after the injection (n = 3 per
time). Blood samples were collected and weighed.
Protosol (1 mL) (Packard Bioscience, Groningen, The

T. Okamura et al. / Bioorg. Med. Chem. 15 (2007) 3127–3133
3133
Japan). The column was eluted with mobile phase (ace-
tonitrile and water containing 0.1% trifluoroacetic acid)
at a flow rate of 1.0 mL/min. The acetonitrile gradient
was changed from 0% to 10% over 10 min, kept
constant at 10% for 5 min, and increased to 20% over
8. Sun, H.; Johnson, D. R.; Finch, R. A.; Sartorelli, A. C.;
Miller, D. W.; Elmquist, W. F. Biochem. Biophys. Res.
Commun. 2001, 284, 863.
9
. Sugiyama, D.; Kusuhara, H.; Lee, Y. J.; Sugiyama, Y.
Pharm. Res. 2003, 20, 1394.
3
5
10. Risberg, J. Brain Lang. 1980, 9, 9.
1
5
min. [ S]7mPSG and brain chemical form were
1. Olesen, J.; Paulson, O. B.; Lassen, N. A. Stroke 1971, 2,
19.
analyzed under the same HPLC systems, and eluate
from the column was collected in 1 mL fractions.
Radioactivity in the fractions was measured by a liquid
scintillation counter.
5
1
2. Hwang, I. Y.; Elfarra, A. A. J. Pharmacol. Exp. Ther.
1993, 264, 41.
13. Gerard-Monnier, D.; Fougeat, S.; Chaudiere, J. Biochem.
Pharmacol. 1992, 43, 451.
1
4. Johnson, J. A.; el Barbary, A.; Kornguth, S. E.; Brugge, J.
F.; Siegel, F. L. J. Neurosci. 1993, 13, 2013.
5. Montgomery, J. A.; Temple, C., Jr. J. Am. Chem. Soc.
1961, 83, 630.
References and notes
1
1
2
. Kusuhara, H.; Sugiyama, Y. NeuroRx 2005, 2, 73.
. Lee, G.; Dallas, S.; Hong, M.; Bendayan, R. Pharmacol.
Rev. 2001, 53, 569.
16. Irie, T.; Fukushi, K.; Namba, H.; Iyo, M.; Tamagami, H.;
Nagatsuka, S.; Ikota, N. J. Nucl. Med. 1996, 37, 649.
3
. Ghersi-Egea, J. F.; Strazielle, N. Microsc. Res. Tech. 2001,
2, 83.
17. Hemmings, S. J.; Storey, K. B. Mol. Cell Biochem. 1999,
202, 119.
5
4
5
6
. Deeley, R. G.; Cole, S. P. FEBS Lett. 2006, 580, 1103.
. Loscher, W.; Potschka, H. Prog. Neurobiol. 2005, 76, 22.
. Sultana, R.; Butterfield, D. A. Neurochem. Res. 2004, 29,
18. Robins, R. K.; Lin, H. H. J. Am. Chem. Soc. 1957, 79,
490.
19. Prasad, R. N.; Robins, R. K. J. Am. Chem. Soc. 1957, 79,
6401.
20. Hwang, I. Y.; Elfarra, A. A. J. Pharmacol. Exp. Ther.
1991, 258, 171.
2
215.
. Sun, H.; Miller, D. W.; Elmquist, W. F. Pharm. Res. 2001,
8, 1542.
7
1