Synthesis, in vitro characterization, and radiolabeling of reboxetine analogs as potential PET radioligands for imaging the norepinephrine transporter

Article abstract of DOI:10.1016/j.bmc.2007.10.025

Six new (S, S)-enantiomers of reboxetine derivatives were synthesized and their binding affinities were determined via competition binding assays in cells expressing the human norepinephrine transporter (NET), serotonin transporter (SERT) or dopamine tran

Full text of DOI:10.1016/j.bmc.2007.10.025

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 783–793
Synthesis, in vitro characterization, and radiolabeling
of reboxetine analogs as potential PET radioligands
for imaging the norepinephrine transporter
Fanxing Zeng,a Nachwa Jarkas,a Jeffrey S. Stehouwer,a Ronald J. Voll,a
Michael J. Owens,b Clinton D. Kilts,b Charles B. Nemeroffb and Mark M. Goodmana,*
aDepartment of Radiology, Division of Radiological Sciences, Emory University, Atlanta, GA 30322, USA
bDepartment of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, GA 30322, USA
Received 26 July 2007; revised 2 October 2007; accepted 10 October 2007
Available online 13 October 2007
Abstract—Six new (S,S)-enantiomers of reboxetine derivatives were synthesized and their binding affinities were determined via
competition binding assays in cells expressing the human norepinephrine transporter (NET), serotonin transporter (SERT) or dopa-
mine transporter (DAT). All six compounds prepared exhibit high affinity for the NET (Ki 6 2 nM) and selectivity versus the SERT
and DAT. Radiolabeling methods were also developed to prepare these ligands in moderate to high radiochemical yield with high
radiochemical purity via O- or S-methylation with [11C]CH3I, or O-alkylation with [18F]fluoroethyl brosylate or [18F]fluoropropyl
brosylate, and their logP7.4 was measured. These new C-11- and F-18-labeled tracers will be utilized in comparative microPET stud-
ies to evaluate their potential as PET radioligands for imaging brain NET in nonhuman primates.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
namic properties of potential medications with affinity
for the NET.
The norepinephrine transporter (NET) is a 12-mem-
brane-spanning protein located on the nerve terminals
as well as the cell bodies of noradrenergic neurons.1,2
The NET plays a critical role in regulating norepineph-
rine (NE) concentration at noradrenergic synapses by
reclaiming uptake of NE from the extracellular space
into presynaptic terminals.1,3 Evidence from animal
model, postmortem human, and molecular neuroscience
studies support a brain regional and complex dysregula-
tion of noradrenergic function in mood and anxiety dis-
orders.4,5 The NET is also an established molecular
target for the treatment of attention-deficit/hyperactivity
disorder (ADHD), substance abuse, neurodegenerative
disorders, and depression. Therefore, A NET imaging
agent would be a unique tool to determine the role of
noradrenergic mechanisms in neuropsychiatric disorders
as well as the in vivo pharmacokinetic and pharmacody-
Several potent NET-selective antidepressants, such
as desipramine,6–8 nisoxetine,9–12 oxaprotiline,11 lortal-
amine,11 and analogs of tomoxetine13 have been labeled
for in vitro or in vivo mapping of brain NET. Unfortu-
nately, the high in vivo nonspecific binding of these
radioligands excluded their utility as NET imaging
agents.
Reboxetine (2-[a-(2-ethoxyphenoxy)benzyl]morpholine),
a racemic mixture of (R,R) and (S,S) enantiomers, is a po-
tent and relatively selective NE reuptake inhibitor and
commercially available in several countries outside the
United States as an antidepressant. Reboxetine has re-
cently been widely chosen as a molecular model for the de-
123
sign of NET radioligands and several 11C-, 18F- or
I-labeled derivatives of reboxetine have been developed
and evaluated as putative PET (positron emission
tomography) or SPECT (single photon emission com-
puted tomography) imaging agents for the NET.11,14–21
These radioligands, including O-methyl ((S,S)-[11C]-
MeNER),14–16 O-fluoromethyl ((S,S)-[18F]FMeNER-
D2),17,18 and 2-iodo ((S,S)-[123I]IPBM)19,20 analogs,
showed many desired in vivo properties for imaging non-
Keywords: Norepinephrine transporter; Positron Emission Tomogra-
phy; Carbon-11; Fluorine-18; (S,S)Reboxetine analogs.
*
Corresponding author. Tel.: +1 404 727 9366; fax: +1 404 727
3488; e-mail: mgoodma@emory.edu
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.10.025

784
F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
human primates that have not been observed with any
other NET ligands. (S,S)-[11C]MeNER and (S,S)-
[18F]FMeNER-D2 are probably the most promising can-
didates reported so far for imaging NET by PET with the
highest thalamus/striatum ratio of ꢀ1.5 in cynomolgus
monkey. However, these two radioligands do not appear
to be ideal. Previous evaluation of (S,S)-[11C]MeNER
has shown that specific binding toNET did not reach peak
equilibrium during a 90-min PET measurement. This, to-
gether with its relatively noisy signal at later time points,
might hamper its utility in the quantitative assessment
of NET binding occupancy in brain. (S,S)-[18F]FMeN-
ER-D2 undergoes a moderate degree of defluorination
resulting in skull uptake of radioactivity which was found
to contaminate images in cortical areas in PET studies.
ylthio)phenol and methyl salicylate under Mitsunobu
conditions24 followed by standard deprotection of N-
Boc group with excess trifluoroacetic acid, respectively.
The (R) configuration of the benzylic carbon in com-
pound 8 was inverted after Mitsunobu coupling reaction
to furnish the (S) form of aryl benzyl ethers, as demon-
strated by Tamagnan and colleagues.23 The structures of
target compounds (3–6) were successfully built by a dou-
ble inversion strategy25: conversion of alcohol 7 to the
corresponding benzyl bromide 11 using carbon tetrabro-
mide and triphenylphosphine with inversion of configu-
ration at the benzylic position from (S) to (R); reaction
of benzyl bromide 11 with thiophenols under nucleo-
philic substitution SN2 mechanism to provide arylthiom-
ethyl morpholines (12, 13, and 16) with inversion of (R)
configuration of benzylic carbon to (S). Therefore, 3 and
6 were prepared in 42% and 65% two-step yields by
In an effort to develop NET-specific PET imaging agents,
the promising results from reboxetine motif served as
the impetus for our group to continuously explore
reboxetine analogs to further optimize this ligand by,
for example, significantly decreasing the time to reach a
specific binding peak equilibrium while maintaining the
favorable specific-to-nonspecific binding ratios in NET-
rich tissues, and reducing the defluorination to an unob-
servable level for fluorine-18 radioligands. Because the
most active reboxetine isomer always corresponds to
the S,S isomer, we decided to focus our efforts on the
S,S isomer. We report here the synthesis, in vitro binding
affinities, and the radiolabeling with carbon-11 or fluo-
rine-18 of new reboxetine analogs (1–6, Fig. 1).22 Results
of comparative microPET studies of these tracers in non-
human primate brain are currently in preparation and
will be published in a separate report.
reacting
(2S,3R)-N-Boc-2-[a-bromo(phenyl)methyl]-
morpholine (11) with 2-methoxybenzenethiol and
methyl thiosalicylate followed by deprotection of N-
Boc group with excess trifluoroacetic acid, respectively.
The reference compounds 4 and 5 were prepared in
the following two steps from (2S,3S)-N-Boc-2-[a-(2-
hydroxyphenylthio)benzyl]morpholine (13) which was
obtained in 91% yield by the reaction of 11 with 2-
hydroxythiophenol. First, 13 was fluoroethylated and
fluoropropylated at the phenolic oxygen using 2-fluoro-
ethyl tosylate26 and 3-fluoropropyl tosylate27 to give 14
and 15, respectively. The subsequent Boc removal of 14
and 15 with CF3COOH yielded 4 and 5 in 44% and 57%
(two-step) yields, respectively.
The synthesis of precursors 18–20 for radiolabeling is
shown in Scheme 2. The preparation of [11C]1 through
an S-methylation reaction required preparation of the
corresponding thiol. Considering the fact that thiol is
generally unstable, we need to prepare S-protected pre-
cursor that would be stable to storage and directly used
for radiolabeling. In view of a previous paper, in which
an S-acetyl ester was used as a precursor for S-methyla-
tion,28 we attempted to protect the thiol group of the
starting material, 2-hydroxy-thiophenol, as a thioester
by treatment with acetyl chloride or acetic anhydride.
Unfortunately, very low regioselectivity was observed
under all attempted conditions. While we were in
the process of exploring synthetic methods of the pre-
cursor, Schou et al. reported an effective precursor con-
taining sulfanyl c-propionic acid methyl ester protecting
group for 11C-methylation of (R)-thionisoxetine.29 They
claimed that this protective group could be removed
in situ under mild basic condition to generate free thio-
late ion for radiolabeling. We therefore decided to use
this strategy. First, the thiol moiety of 2-hydroxy-thio-
phenol was regioselectively protected by reaction with
methyl acrylate to give S-alkylation compound 17.29
Then, 8 was coupled with 17 under Mitsunobu condi-
tions to give 18 as the labeling precursor for [11C]1 with
complete inversion of configuration. Esters 10 and 16
were hydrolyzed under basic conditions to afford corre-
sponding carboxylic acids 19 and 20 which were used as
radiolabeling precursors for [11C]2 and [11C]6, respec-
tively. Compound 13 was used as the precursor in the
radiosynthesis of [11C]3, [18F]4, and [18F]5.
2. Results and discussion
2.1. Synthesis of reference compounds and precursors
The stereoselective synthesis of new ligands 1–6 is shown
in Scheme 1. The key intermediates (2S,3S)-N-Boc-2-[a-
hydroxy(phenyl)methyl]morpholine (7) and (2S,3R)-N-
Boc-2-[a-hydroxy(phenyl)methyl]morpholine (8) were
synthesized in good yields and 99% enatiomeric excess
in six linear steps starting from commercially available
(S)-3-amino-1,2-propanediol according to a recently re-
ported procedure.23 7 was the major isomer and 8 was
the minor isomer. Ligands 1 and 2 were prepared in
29% and 44% two-step yields by reacting 8 with 2-(meth-
R
S
R
O
R
O
O
O
O
N
H
N
H
N
H
1:R = SCH3
3:R = OCH3
Reboxetine: R =OCH2CH3
MeNER: R = OCH3
4:R = OCH2CH2F
2:R = COOCH3
5
:R = OCH2CH2CH2F
6:R = COOCH3
FMeNER-D2: R = OCD2F
FERB-D2: R = OCD2CD2F
IPBM: R = I
Figure 1. (2S,3S) Enantiomers of reboxetine derivatives.

F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
785
HO
Ph
O
HO
Ph
O
HO
OH
6 steps
+
(S)
N
N
NH
2
(2S,3R)
(2S,3S)
Boc
b
ref. 23
Boc
7
8
SCH
O
3
SCH
O
3
HO
Ph
O
SCH
3
a
+
+
O
N
O
OH
(2S,3R)
Boc
Boc
N
8
N
H
Boc
1
9
COOCH
3
COOCH
3
HO
Ph
O
COOCH
3
a
O
O
b
O
O
N
OH
(2S,3R)
N
8
N
Boc
H
2
10
HO
Ph
O
Br
Ph
O
c
N
N
N
(2S,3S)
Boc
Boc
d
7
11
OCH
S
OCH
3
3
Br
O
OCH
3
S
b
+
O
N
O
Ph
SH
Boc
N
H
Boc
11
12
3
OCH CH F
2
2
OCH CH F
2 2
S
b
S
e
O
O
OH
Br
O
N
Boc
OH
SH
N
H
S
d
+
14
4
O
Ph
N
Boc
OCH CH CH F
OCH CH CH F
2
2
2
2
2
2
N
Boc
f
b
11
13
S
S
O
O
N
H
N
Boc
5
15
COOCH
COOCH
3
3
Br
O
COOCH
SH
3
S
S
b
d
+
O
O
Ph
N
Boc
N
Boc
N
H
11
6
16
Scheme 1. Reagents and conditions: (a) DIAD, PPh3, THF, 0 ꢁC–rt, 48 h; (b) TFA, CH2Cl2, 0 ꢁC–rt, 2 h; (c) CBr4, PPh3, CH2Cl2, 0 ꢁC–rt, 30 min;
(d) Cs2CO3, DMF, rt, 18 h; (e) FCH2CH2OTs, Bu4NOH, DMF; (f) FCH2CH2CH2OTs, Bu4NOH, DMF.
2.2. Radiolabeling
rected radiochemical yield through S-alkylation of the
intermediate thiol 21 with [11C]CH3I in THF, followed
by deprotection with 6 M HCl and HPLC purification.
KOtBu-promoted in situ b-elimination of the sulfanyl c-
propionic acid methyl ester moiety in 18 generated the free
The radiosynthesis of [11C]1, [11C]2, [11C]6, [11C]3, [18F]4,
and [18F]5 is shown is Scheme 3, and the results are listed
in Table 1. [11C]1 was obtained in quantitative decay-cor-

786
F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
OH
O
S
OCH3
OCH3
SH
OH
S
OCH3
N
O
O
O
Boc
O
O
N
DMF, KHCO3
rt, 4h
DIAD, PPh3,THF
OH
17
Boc
18
COOH
O
COOCH3
NaOH, DMF, H2O
reflux
O
O
O
N
Boc
N
Boc
19
10
COOH
COOCH 3
1N NaOH, EtOH
reflux
S
S
O
O
N
N
Boc
Boc
16
20
Scheme 2. Synthesis of precursors for radiolabeling.
thiolate ion (21), which was then effective for accomplish-
ing 11C-methylation. The radiolabeling of [11C]2 and
[11C]6 was accomplished through O-methylation of N-
Boc carboxylic acid precursors 19 and 20 with
[11C]CH3I in the presence of 0.1 M Bu4NOH, followed
by removal of N-Boc group under acidic conditions.
The radiosynthesis of [11C]3 was conducted via O-methyl-
ation of deprotonated N-Boc phenol 13 with [11C]CH3I.
The preparation of [18F]4 and [18F]5 was accomplished
via a three-step procedure using 13 as the precursor. First,
2-[18F]fluoroethyl brosylate and 3-[18F]fluoropropyl
brosylate were prepared as a secondary radiolabeling
synthon by the nucleophilic substitution of 1,2-dibrosyle-
thane and 1,3-dibrosylpropane with [18F]FÀ in acetoni-
trile, respectively. The subsequent coupling of 13 with
2-[18F]fluoroethyl brosylate or 3-[18F]fluoropropyl brosy-
late was carried out in DMF at 100 ꢁC in the presence of
0.1 M Bu4NOH. Last, the Boc group was cleaved by treat-
ing N-Boc-[18F]4 or N-Boc-[18F]5 with 6 M HCl. After
HPLC purification, [18F]4 and [18F]5 were obtained in
6–8% decay-corrected radiochemical yields from
[18F]FÀ in a synthesis time of 120 min with a radiochem-
ical purity of >98% and a specific activity of 1.7–2.3 Ci/
lmol at time of injection. The lipophilicities of radioli-
gands 1–6 were measured by traditional extraction with
octanol and pH 7.4 phosphate buffer according to a previ-
ously reported procedure,30 and logP7.4 values are given
in Table 1. The logP7.4 values of all the radiotracers mea-
sured are in the optimal range (1.0–3.0) for compounds
expected to enter the brain readily.31
stably expressing the human NET (hNET), serotonin
transporter (hSERT), or dopamine transporter (hDAT)
according to a previously reported procedure.32,33 (S,S)-
MeNER was also screened under the same assay system
for comparison.
[3H]nisoxetine (NET ligand), [3H]citalopram (SERT li-
gand), and [125I]RTI-55 (DAT ligand) were used as
radiotracers during the in vitro displacement experi-
ments. Data in Table 2 indicate that all new reboxetine
derivatives displayed high affinities for the hNET
comparable to the value for (S,S)-MeNER. The rank
of order of hNET affinities (Ki in nM) was 3
(0.76) > (S,S)-MeNER (0.95) > 4 (1.0) > 5 (1.21) > 2
(1.32) P 6 (1.35) > 1 (2.06). The reboxetine derivatives
tested exhibited very low affinities for the DAT
(Ki > 3000) as well as moderate to low affinities for the
SERT (91 6 Ki 6 498), except for 3 with Ki = 9.69 nM
for the SERT. Replacing the methoxy group on the ben-
zene ring of MeNER with a methylthio group afforded
1, whose affinity for the NET decreased 2-fold, while
the affinity for the SERT increased 2-fold compared to
those affinities for MeNER. Substitution of the oxygen
atom with sulfur as the linker of the three rings in MeN-
ER slightly improved the affinity for the NET, while
increasing the affinity for the SERT 20-fold as seen in
compound 3. Consequently, 3 exhibited a lower selectiv-
ity for the hNET versus hSERT than MeNER, with a
SERT/NET ratio of 13. Incorporation of x-fluoroethyl
and x-fluoropropyl groups as in analogs 4 and 5 re-
tained the NET potency about the same compared to
3, while SERT affinity significantly decreased. Com-
pound 5 had the highest selectivity (ꢀ700-fold) for the
NET over the SERT. Two 2-(a-(2-carbomethoxy)) ana-
logs of reboxetine (2 and 6) have roughly the same affin-
ities toward hNET, hSERT, and hDAT, respectively.
2.3. Biological results: in vitro competition assays
New reboxetine derivatives 1–6 were screened for bind-
ing to human monoamine transporters using in vitro
competition binding assays in transfected HEK-293 cells

F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
787
11
SH
S
OCH
3
S CH
3
O
11
o
O
O
(2) [ C]CH I, 0 C;
3
t
O
(1) KO Bu, THF, rt
o
O
then 40 C, 2 min
O
N
O
o
(3) 6 M HCl, 100 C, 7 min
(4) 6 M NH OH, 0
o
N
C
4
N
H
Boc
Boc
11
21
18
[
C]1
11
COOH
COO CH
3
(1) 0.1 M Bu NOH , DMF
4
(aq)
11
o
(2) [ C]CH I, 0 C;
3
o
X
then 100 C, 10 min
X
O
o
O
(3) TFA, 100 C, 7 min
o
(4) 6 M NH OH, 0
C
4
N
Boc
N
H
11
[
C]2: X = O
19: X = O
11
20: X = S
[
C]6
: X = S
11
O
S
CH
3
(A)
O
N
H
(1) 0.1 M Bu NOH (aq), DMF
4
11
o
(A) [ C]CH I, 0 C;
3
(2)
or
11
o
[ C]3
then 100 C, 10 min
BsO
OBs
OBs
OH
S
18
BsO
OBs
18F-
OBs
OCH CH
2
F
2
(B)
or (C)
18F-
18F
18F
S
o
O
100 C, 10 min
o
(B)
O
(3) 6 M HCl, 100 C, 7 min
o
N
Boc
N
(4) 6M NH OH (aq), 0
C
4
H
18
[
F]4
13
18
OCH CH CH F
2
2
2
S
(C)
O
N
H
18
[
F]5
Scheme 3. Radiosynthesis of reboxetine analogs.
Table 1. Overall synthesis times from EOB, average decay-corrected radiochemical yields (RCY), radiochemical purities, specific activities at time of
injection, and logP7.4 values of radioligands
Radioligands
Synthesis time (min)
RCY (%)
Radiochemical purity (%)
Specific activity (Ci/lmol)
LogP7.4
[
[
[
[
[
[
11C]1
11C]2
11C]6
11C]3
18F]4
18F]5
50
58
ꢀ100a
24a
23a
35a
8b
>98
>98
>98
>98
>98
>98
0.4–0.8
0.3–0.9
0.2–0.6
0.5–1.0
1.7–2.2
1.9–2.3
1.91
1.38
1.64
1.81
1.83
2.30
60
55
120
120
6b
a From end of [11C]CH3I synthesis.
b From [18F]FÀ produced.
3. Conclusions
pounds tested exhibited high affinity and selectivity for
the NET over the SERT and DAT. [11C]1–3, and 6,
and [18F]4–5 were prepared in moderate to high radio-
chemical yields with high radiochemical purity and spe-
cific activity, and their lipophilicity was measured. These
new C-11- and F-18-labeled tracers will permit compar-
A series of reboxetine analogs have been prepared and
their in vitro affinities to the human monoamine trans-
porters, NET, SERT, and DAT have been evaluated.
Competition binding assays showed that all of the com-

788
F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
Table 2. In vitro evaluation of candidate NET ligands in competition assays with human monoamine transportersa
Compound
Ki for hNETb
Ki for hSERTc
Ki for hDATd
NET selectivity
SERT/NET
DAT/NET
1
2.06 0.13
1.32 0.06
0.76 0.06
1.00 0.10
1.21 0.37
1.35 0.17
0.95 0.03
90.88 4.63
614
>6000
>3000
44
465
13
>3000
>2000
2
9
3
9.69 1.34
733 101
121 18
>4000
>10,000
>9000
>4000
>10,000
>7000
4
733
100
369
195
5
6
(S,S)-MeNER
498 49
185
>30,000
>10,000
>20,000
>10,000
3
a All Ki values are reported with nanomolar (nM) units. Data are expressed as geometric means standard deviation of at least three separate
experiments performed in triplicate.
b Competitive binding vs [3H]nisoxetine in HEK-293 cells transfected with human norepinephrine transporters.
c Competitive binding vs [3H]citalopram in HEK-293 cells transfected with human serotonin transporters.
d Competitive binding vs [125I]RTI-55 in canine kidney cells transfected with human dopamine transporters.
ative microPET studies to evaluate their potential as
PET radioligands for imaging NET in the central ner-
vous system of nonhuman primates.
EtOAc (80:20) to afford 9 as colorless oil (20 mg,
37%): 1H NMR (CDCl3, 400 MHz), d 7.25–7.39 (m,
5H), 7.09 (dd, J = 1.9, 7.3 Hz, 1H), 6.85–6.93 (m, 2H),
6.60–6.62 (m, 1H), 5.25 (br s, 1H), 3.92 (dd, J = 2.2,
11.4 Hz, 1H), 3.83–3.86 (m, 3H), 3.52–3.58 (m, 1H),
2.73–2.88 (m, 2H), 2.44 (s, 3H), 1.42 (s, 9H).
4. Experimental
4.1. General methods
4.3. (2S, 3S)-2-[a-(2-(Methylthio)phenoxy)benzyl]mor-
pholine (1)
2-Fluoroethyl tosylate,26 3-fluoropropyl tosylate,27 and
methyl 3-(2-hydroxyphenylthio)propanoate (17)29 were
prepared according to the published procedures. All
other reagents used were obtained from commercially
available sources and were used without further purifi-
cation. All reactions were performed in oven glassware
Trifluoroacetic acid (0.07 mL, 0.91 mmol) was added
dropwise to a solution of 9 (20 mg, 0.048 mmol) in
CH2Cl2 (1.5 mL) at 0 ꢁC. The reaction mixture was al-
lowed to reach room temperature and stirred for an-
other 2 h. Two milliliters of 1 M NaOH solution was
then slowly added at 0 ꢁC and the mixture was extracted
with CH2Cl2. The extracts were combined, dried over
Na2SO4, and the solvent was evaporated under reduced
pressure. The crude product was purified by flash chro-
matography on silica eluted with MeOH/CH2Cl2 (10:90)
to afford 1 as colorless oil (12 mg, 79%): 1H NMR
(CDCl3, 400 MHz), d 7.25–7.38 (m, 5H), 7.08 (dd,
J = 1.9, 7.6 Hz, 1H), 6.83–6.93 (m, 2H), 6.63 (dd,
J = 1.2, 8.2 Hz, 1H), 5.20 (d, J = 5.7 Hz, 1H), 3.92–
3.96 (m, 2H), 3.65 (td, J = 3.5 10.8 Hz, 1H), 2.73–2.82
(m, 3H), 2.60–2.66 (m, 1H), 2.44 (s, 3H), 2.10 (s, br,
1H); 13C NMR (CDCl3, 100 MHz), d 154.5, 137.4,
128.6, 128.3, 127.3, 125.9, 125.6, 121.6, 113.3, 81.6,
79.1, 68.3, 47.3, 45.8, 14.8. HRMS [MH]+ Calcd for
C18H22NO2S: 316.1366, found: 316.1367.
1
fitted with rubber septa under argon. H NMR spectra
were recorded on a Varian spectrometer at 400 MHz
and referenced to the NMR solvent (chemical shifts in
ppm values, J values in Hz). High-resolution mass spec-
tra were acquired on a VG 70-S double focusing mass
spectrometer using electron ionization (EI). Thin-layer
chromatography (TLC) was performed using 250 lm
layers of F-254 silica on aluminum plates obtained from
Whatman (Clifton, NJ) and visualized by UV light.
Flash chromatography was carried out using Merck sil-
ica gel 60 (40–63 lm particle size). No carrier-added
[11C]CO2 was produced through the bombardment of
14N2 gas containing 1% 16O2 by a Siemens 11 MeV
RDS 112 cyclotron at Emory University Hospital
through the 14N[p,a]11C reaction. A GE MicroLab
methyl iodide system was employed for the conversion
of [11C]CO2 to [11C]CH3I.
4.4. (2S, 3S)-N-tert-Butoxycarbonyl-2-[a-(2-(carbome-
thoxy)phenoxy)benzyl]morpholine (10)
4.2. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-(methylthio)
phenoxy)benzyl]morpholine (9)
To a mixture of 8 (40 mg, 0.136 mmol), methyl salicylate
(35 lL, 0.27 mmol), and triphenylphosphine (71 mg,
0.27 mmol) in dry THF (3 mL) was added diisopropyl
azodicarboxylate (53 lL, 0.27 mmol) at 0 ꢁC. The reac-
tion mixture was allowed to reach room temperature
and stirred for 24 h. The crude product was concen-
trated under reduced pressure and was purified by flash
chromatography on silica eluted with hexane/EtOAc
To a mixture of (2S,3R)-N-Boc-2-[a-hydroxy(phenyl)
methyl]morpholine (8) (38 mg, 0.13 mmol), 2-(methyl-
thio)phenol (36 mg, 0.26 mmol), and triphenylphos-
phine (68 mg, 0.26 mmol) in dry THF (2 mL) was
added diisopropyl azodicarboxylate (52 lL, 0.26 mmol)
at 0 ꢁC. The reaction mixture was allowed to reach room
temperature and stirred for 24 h. The crude product was
concentrated under reduced pressure and was purified
by flash chromatography on silica eluted with hexane/
1
(80:20) to afford 10 as colorless oil (43 mg, 74%): H
NMR (CDCl3, 400 MHz), d 7.75 (d, J = 1.9, 7.6 Hz,
1 H), 7.21–7.39 (m, 6 H), 6.88 (t, J = 7.3 Hz, 1 H), 6.75

F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
789
(d, J = 8.6 Hz, 1H), 5.25 (br s, 1H), 3.92 (s, 3H), 3.84–
3.90 (m, 4H), 3.54 (td, J = 2.5, 11.4 Hz, 1H), 2.87(br s,
1H), 2.72 (br s, 1H), 1.41 (s, 9H); 13C NMR (CDCl3,
100 MHz), d 157.3, 154.9, 136.6, 133.3, 131.9, 128.7,
128.5, 127.4, 121.3, 120.7, 115.2, 81.7 (br), 80.2, 77.9,
67.0, 52.2, 44.2 (br), 43.2 (br), 28.5. HRMS [MH]+
Calcd for C24H30NO6: 428.2068, found: 428.2067.
(45 mg), which was used without any further
purification.
4.8. (2S,3S)-2-[a-(2-Methoxyphenylthio)phenylmethyl]mor-
pholine (3)
Trifluoroacetic acid (0.13 mL, 1.68 mmol) was added
dropwise to a solution of 12 (45 mg, 0.11 mmol) ob-
tained as above in CH2Cl2 (2 mL) at 0 ꢁC. The reaction
mixture was allowed to reach room temperature and
stirred for another 2 h. Three milliliters of 1 M NaOH
solution was then slowly added at 0 ꢁC and the mixture
was extracted with CH2Cl2. The extracts were com-
bined, dried over Na2SO4, and the solvent was evapo-
rated under reduced pressure. The crude product was
purified by flash chromatography on silica eluted with
MeOH/CH2Cl2 (10:90) to afford 3 as colorless oil
(25 mg, 42% in two-step): 1H NMR (CDCl3,
400 MHz), d 7.06–7.19 (m, 7 H), 6.65–6.69 (m, 2H),
4.39 (d, J = 8.2 Hz, 1H), 3.97–4.01 (m, 1H), 3.79–3.84
(m, 1H), 3.78 (s, 3H), 3.63 (td, J = 2.8, 11.1 Hz, 1H),
2.86 (td, J = 3.2, 11.1 Hz, 1H), 2.77 (d, J = 12.0 Hz,
1H), 2.61–2.64 (m, 2H), 1.92 (s, br, 1H); 13C NMR
(CDCl3, 100 MHz), d 159.0, 139.1, 134.3, 128.8, 128.3,
127.4, 122.4, 120.7, 110.5, 79.8, 68.6, 55.8, 54.7, 49.7,
45.8. HRMS [MH]+ Calcd for C18H22NO2S: 316.1366,
found: 316.1363.
4.5. (2S,3S)-2-[a-(2-(Carbomethoxy)phenoxy)benzyl]mor-
pholine (2)
Trifluoroacetic acid (0.11 mL, 1.44 mmol) was added
dropwise to a solution of 10 (41 mg, 0.096 mmol) in
CH2Cl2 (2 mL) at 0 ꢁC. The reaction mixture was al-
lowed to reach room temperature and stirred for an-
other 2 h. 3 mL of 1 M NaOH solution was then
slowly added at 0 ꢁC and the mixture was extracted with
CH2Cl2. The extracts were combined, dried over
Na2SO4, and the solvent was evaporated under reduced
pressure. The crude product was purified by flash chro-
matography on silica eluted with MeOH/CH2Cl2 (10:90)
to afford 2 as colorless oil (19 mg, 60%): 1H NMR
(CDCl3, 400 MHz), d 7.75 (dd, J = 1.6, 7.6 Hz, 1H),
7.21–7.37 (m, 6H), 6.88 (td, J = 1.0, 7.6 Hz, 1H), 6.75
(d, J = 8.2 Hz, 1H), 5.24 (d, J = 5.4 Hz, 1H), 3.97–4.02
(m, 2H), 3.91 (s, 3H), 3.66–3.72 (m, 1H), 2.84–2.88 (m,
3H), 2.68–2.73 (m, 1H); 13C NMR (CDCl3, 100 MHz),
d 167.1, 157.3, 136.7, 133.4, 131.9, 128.7, 128.5, 127.3,
121.1, 120.8, 115.2, 81.8, 78.2, 67.4, 52.2, 46.3, 45.1.
HRMS [MH]+ calcd for C19H22NO4: 328.1543, found:
328.1541.
4.9. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-hydroxyphe-
nylthio)benzyl]morpholine (13)
A reaction mixture of 11 (71 mg, 0.2 mmol), 2-hydroxy-
thiophenol (50 mg, 0.4 mmol), and Cs2CO3 (130 mg,
0.4 mmol) in DMF (2 mL) was stirred at room temper-
ature for 18 h. The crude product was directly purified
by flash chromatography on silica eluted with hexane/
EtOAc (80:20) to afford 13 as colorless oil (73 mg,
91%): 1H NMR (CDCl3, 400 MHz), d 7.52 (s, 1H),
7.21–7.25 (m, 3H), 6.98–7.02 (m, 3H), 6.91–6.94 (m,
1H), 6.68 (td, J = 1.2, 7.6 Hz, 1H), 4.06–4.11 (m, 1H),
3.83–3.86 (m, 1H), 3.66–3.74 (m, 4H), 2.95–3.02 (m,
1H), 2.52 (m, 1H), 1.35 (s, 9H); 13C NMR (CDCl3,
100 MHz), d 158.4, 154.5, 138.6, 138.2, 132.1, 129.0,
128.1, 128.0, 120.3, 115.4, 80.4, 76.5, 67.1, 57.8 (br),
47.0 (br), 43.8 (br), 28.5. HRMS [MH]+ calcd for
C22H28NO4S: 402.1734, found: 402.1739.
4.6. (2S,3R)-N-tert-Butoxycarbonyl-2-[a-bromo(phenyl)
methyl]morpholine (11)
To a solution of (2S,3S)-N-Boc-2-[a-hydroxy(phenyl)
methyl]morpholine (7) (118 mg, 0.4 mmol) in CH2Cl2
(3 mL) was added PPh3 (210 mg, 0.8 mmol) at room
temperature. The mixture was then cooled down to
0 ꢁC, and a solution of CBr4 (266 mg, 0.8 mmol) in
CH2Cl2 (1 mL) was then added dropwise at this temper-
ature. The reaction mixture was warmed up to room
temperature and stirring was continued for 30 min.
The solvent was evaporated and the crude product
was purified by flash chromatography on silica eluted
with hexane/EtOAc (80:20) to afford 11 as colorless oil
1
(107 mg, 75%): H NMR (CDCl3, 400 MHz), d 7.26–
7.41 (m, 5 H), 4.84 (d, J = 7.6 Hz, 1H), 4.36 (br s, 1H),
3.79–3.86 (m, 3H), 3.46 (td, J = 2.8, 11.4 Hz, 1H),
2.92–2.96 (m, 2H), 1.46 (s, 9H); 13C NMR (CDCl3,
100 MHz), d 154.8, 138.7, 128.8, 128.6, 80.5, 77.9,
66.9, 53.3, 47.0 (br), 43.2 (br), 28.6. HRMS [MH]+ calcd
for C16H23O3N79Br: 356.0856, found: 356.0854.
4.10. (2S, 3S)-N-tert-Butoxycarbonyl-2-[a-(2-(2-fluoro-
ethoxy)phenylthio)benzyl]morpholine (14)
A stirred mixture of 13 (30 mg, 0.075 mmol) and 2-flu-
oroethyl tosylate (22 mg, 0.097 mmol) in DMF (2 mL)
was treated with tetrabutylammonium hydroxide (1 M
in methanol, 89 lL, 0.089 mmol). After 20 min at room
temperature, the solution was stirred at 80 ꢁC for 2 h,
cooled, and quenched with aqueous Na2CO3 (0.1 M,
8 mL). The mixture was extracted with ethyl acetate,
the organic layer washed with water, dried over anhy-
drous sodium sulfate, and concentrated under reduced
pressure. The residue was loaded on a short silica col-
umn eluted with hexane/EtOAc (80:20), and 100 mL
fractions was collected. After evaporating the solvent,
colorless oil (28.5 mg) was obtained, and NMR shows
4.7. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-methoxyphe-
nylthio)phenylmethyl]morpholine (12)
A reaction mixture of 11 (68 mg, 0.19 mmol), 2-meth-
oxybenzenethiol (46 lL, 0.38 mmol), and Cs2CO3
(124 mg, 0.38 mmol) in DMF (3 mL) was stirred at
room temperature for 18 h. The crude product was di-
rectly loaded on a short silica column eluted with hex-
ane/EtOAc (80:20) to afford 12 as colorless oil

790
F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
it is the mixture of product and 2-fluoroethyl tosylate
(3.5:1), which was used without any further purification.
J = 2.8, 11.1Hz, 1H), 2.76–2.88 (m, 2H), 2.59–2.68 (m,
2H), 2.23 (t, J = 6.0 Hz, 1H), 2.17 (t, J = 6.0 Hz, 1H),
1.81 (br s, 1H); 13C NMR (CDCl3, 100 MHz), d 158.1,
139.3, 134.2, 128.7, 128.6, 128.3, 127.4, 123.1, 121.1,
111.8, 81.9, 80.0 (J = 49.6 Hz), 68.5, 64.3 (J = 5.3 Hz),
54.8, 49.7, 45.8, 30.6 (J = 19.8 Hz). HRMS [MH]+ calcd
for C20H25FNO2S: 362.1584, found: 362.1573.
4.11. (2S,3S)-2-[a-(2-(2-Fluoroethoxy)phenylthio)benzyl]
morpholine (4)
Trifluoroacetic acid (70 lL, 0.83 mmol) was added
dropwise to a solution of compounds (28.5 mg) obtained
as above in CH2Cl2 (2 mL) at 0 ꢁC. The reaction mixture
was allowed to reach room temperature and stirred for
another 2 h. Two milliliters of 1 M NaOH solution
was then slowly added at 0 ꢁC and the mixture was ex-
tracted with CH2Cl2. The extracts were combined, dried
over Na2SO4, and the solvent was evaporated under re-
duced pressure. The crude product was purified by flash
chromatography on silica eluted with MeOH/CH2Cl2
(10:90) to afford 4 as colorless oil (12 mg, 44% in two-
step): 1H NMR (CDCl3, 400 MHz), d 7.06–7.22 (m,
7 H), 6.68–6.73 (m, 2H), 4.90–4.95 (m, 1/2H), 4.72–
4.84 (m, 1H), 4.68–4.71 (m, 1/2H), 4.52 (d, J = 7.9 Hz,
1H), 4.09–4.27 (m, 2H), 3.98–4.02 (m, 1H), 3.79–3.84
(m, 1H), 3.64 (td, J = 2.8, 11.1 Hz, 1H), 2.84–2.91 (m,
1H), 2.76–2.80 (m, 1H), 2.67 (d, J = 6.3 Hz, 2H), 1.75
(s, 1H). HRMS [MH]+ calcd for C19H23FNO2S:
348.1428, found: 348.1418.
4.14. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-(carbome-
thoxy)phenylthio)benzyl]morpholine (16)
A reaction mixture of 11 (31 mg, 0.087 mmol), methyl
thiosalicylate (24 lL, 0.174 mmol), and Cs2CO3
(57 mg, 0.174 mmol) in DMF (2 mL) was stirred at
room temperature for 18 h. The crude product was di-
rectly purified by flash chromatography on silica eluted
with hexane/EtOAc (80:20) to afford 16 as colorless oil
(33 mg, 85%): 1H NMR (CDCl3, 400 MHz), d 7.75–
7.77 (m, 1H), 7.35–7.37 (m, 2H), 7.17–7.24 (m, 5H),
7.05–7.09 (m, 1H), 4.40 (d, J = 7.6 Hz, 1H), 3.91–3.97
(m, 1H), 3.89 (s, 3H), 3.72–3.75 (m, 3H), 3.50–3.55 (m,
1H), 2.93–2.98 (m, 1H), 2.77 (br s, 1H), 1.37 (s, 9H).
4.15. (2S,3S)-2-[a-(2-(Carbomethoxy)phenylthio)benzyl]
morpholine (6)
4.12. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-(3-fluoropro
poxy)phenylthio)benzyl]morpholine (15)
Trifluoroacetic acid (85 lL, 1.11 mmol) was added
dropwise to a solution of 16 (33 mg, 0.074 mmol) in
CH2Cl2 (2 mL) at 0 ꢁC. The reaction mixture was al-
lowed to reach room temperature and stirred for an-
other 2 h. Three milliliters of 1 M NaOH solution was
then slowly added at 0 ꢁC and the mixture was extracted
with CH2Cl2. The extracts were combined, dried over
Na2SO4, and the solvent was evaporated under reduced
pressure. The crude product was purified by flash chro-
matography on silica eluted with MeOH/CH2Cl2 (10:90)
to afford 6 as colorless oil (20 mg, 78%): 1H NMR
(CDCl3, 400 MHz), d 7.75 (dd, J = 1.6, 8.0 Hz, 1 H),
7.35 (d, J = 6.7 Hz, 1H), 7.16–7.26 (m, 5H), 7.05 (td,
J = 1.3, 7.6 Hz, 1H), 4.39 (d, J = 7.9 Hz, 1H), 3.89–
3.98 (m, 1H), 3.88 (s, 3H), 3.77–3.83 (m, 1H), 3.62
(dd, J = 2.9, 11.1 Hz, 1H), 2.84 (td, J = 3.5, 12.4 Hz,
1H), 2.74–2.77 (m, 1H), 2.61 (d, J = 6.7 Hz, 2H), 1.97
(s, 1H); 13C NMR (CDCl3, 100 MHz), d 167.3, 139.2,
138.7, 131.8, 130.8, 130.3, 129.5, 128.8, 128.6, 127.7,
124.9, 79.5, 68.6, 55.2, 52.4, 49.7, 45.7. HRMS [MH]+
calcd for C19H22NO3S: 344.1315, found: 344.1311.
A stirred mixture of 13 (33 mg, 0.082 mmol) and 3-flu-
oropropyl tosylate (26.6 mg, 0.11 mmol) in DMF
(2 mL) was treated with tetrabutylammonium hydroxide
(1 M in methanol, 98 lL, 0.098 mmol). After 20 min at
room temperature, the solution was stirred at 80 ꢁC
for 2 h, cooled, and quenched with aqueous Na2CO3
(0.1 M, 10 mL). The mixture was extracted with ethyl
acetate, the organic layer washed with water, dried over
anhydrous sodium sulfate, and concentrated under re-
duced pressure. The residue was loaded on a short silica
column eluted with hexane/EtOAc (80:20), and 100 mL
fractions were collected. After evaporating the solvent,
colorless oil (31 mg) was obtained, and NMR shows it
is the mixture of product and 3-fluoropropyl tosylate
(3.8:1), which was used without any further purification.
4.13. (2S,3S)-2-[a-(2-(3-Fluoropropoxy)phenylthio)benzyl]
morpholine (5)
Trifluoroacetic acid (75 lL, 0.97 mmol) was added
dropwise to a solution of compounds (31 mg) obtained
as above in CH2Cl2 (2 mL) at 0 ꢁC. The reaction mixture
was allowed to reach room temperature and stirred for
another 2 h. Two milliliters of 1 M NaOH solution
was then slowly added at 0 ꢁC and the mixture was ex-
tracted with CH2Cl2. The extracts were combined, dried
over Na2SO4, and the solvent was evaporated under re-
duced pressure. The crude product was purified by flash
chromatography on silica eluted with MeOH/CH2Cl2
(10:90) to afford 5 as colorless oil (17 mg, 57% in two-
step): 1H NMR (CDCl3, 400 MHz), d 7.05–7.23 (m,
7 H), 6.67–6.73 (m, 2H), 4.73–4.81 (m, 1H), 4.62–4.69
(m, 1H), 4.35 (d, J = 8.0 Hz, 1H), 4.02–4.11 (m, 2H),
3.98 (d, J = 10.8 Hz, 1H), 3.78–3.83 (m, 2H), 3.64 (td,
4.16. (2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-(3-methoxy-
3-oxopropylthio)phenoxy)benzyl]morpholine (18)
To a mixture of 8 (40 mg, 0.136 mmol), methyl 3-(2-
hydroxyphenylthio)propanoate (58 mg, 0.27 mmol),
and triphenylphosphine (71 mg, 0.27 mmol) in dry
THF (3 mL) was added diisopropyl azodicarboxylate
(53 lL, 0.27 mmol) at 0 ꢁC. The reaction mixture was al-
lowed to reach room temperature and stirred for 24 h.
The crude product was concentrated under reduced
pressure and was purified by flash chromatography on
silica eluted with hexane/EtOAc (80:20) to afford 18 as
colorless oil (40 mg, 60%): 1H NMR (CDCl3,
400 MHz), d 7.24–7.38 (m, 6H), 6.96–7.00 (td, J = 1.6,
8.3 Hz, 1H), 6.80–6.84 (m, 1H), 6.63–6.65 (m, 1H),

F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
791
5.25 (br s, 1H), 3.74–3.93 (m, 4H), 3.67 (s, 3H), 3.52–
3.65 (m, 1H), 3.15–3.27 (m, 2H), 2.92 (m, 1H), 2.72–
2.78 (m, 1H), 2.65 (t, J = 7.6 Hz, 2H), 1.41 (s, 9H).
HRMS [MH]+ calcd for C26H34NO6S: 488.2101, found:
488.2090.
one (pore size 0.2 lm), to a 30 mL sterile vial containing
10 mL of saline and is ready for PET study.
4.20. Radiosynthesis of [11C]2 and [11C]6
(S,S)-2((N-(tert-Butoxycarbonyl)morpholin-2-yl) (phenyl)-
methoxyl)benzoic acid (19) or (S,S)-2((N-(tert-Butoxy-
carbonyl)morpholin-2-yl)(phenyl)methylthio)benzoic acid
(20) (1 mg) was dissolved in DMF (0.26 mL) followed
by addition of 0.1 M Bu4NOH (aq) (23 lL). The solu-
tion was placed in a sealed conical vial and cooled to
0 ꢁC, and [11C]CH3I was bubbled through the solution
at 0 ꢁC. The solution was heated at 100 ꢁC for 10 min,
trifluoroacetic acid (0.21 mL) was added, and the solu-
tion was heated at 100 ꢁC for 7 min, cooled to 0 ꢁC,
and neutralized by addition of 6 M NH4OH (0.6 mL).
The solution was purified by semipreparative HPLC
(Water XTerra Prep RP18 5 lm, 19 · 100 mm). The
retention time of [11C]2 was 11 min with buffered mobile
phase consisting 52:48:0.1 v/v/v MeOH/H2O/Et3N at a
8 mL/min flow rate. The retention time of [11C]6 was
12.5 min with buffered mobile phase consisting
53:47:0.1 v/v/v MeOH/H2O/Et3N at a 9-mL/min flow
rate. The dose formulation was conducted by the same
procedure described above.
4.17. (S, S)-2((N-(tert-Butoxycarbonyl)morpholin-2-yl)
(phenyl)methoxyl)benzoic acid (19)
A mixture of 10 (17 mg, 0.04 mmol) and NaOH (16 mg,
0.4 mmol) in DMF (1.5 mL) and H2O (0.5 mL) was
heated at 110 ꢁC overnight. After cooling to room tem-
perature, the crude product was directly purified by flash
chromatography on silica eluted with CH2Cl2/MeOH
(95:5) to afford 19 as colorless oil (11 mg, 67%), which
was solidified by cooling in a freezer: 1H NMR (CDCl3,
400 MHz), d 8.11 (dd, J = 1.7, 7.8 Hz, 1 H), 7.31–7.42
(m, 4 H), 7.28 (td, J = 0.9, 2.0 Hz, 2 H), 7.05 (td,
J = 0.9, 7.9 Hz, 1H), 6.71 (d, J = 8.3 Hz, 1H), 5.06 (d,
J = 7.3 Hz, 1H), 3.98–4.02 (m, 1H), 3.82 (m, 2H),
3.53–3.64 (m, 2H), 2.98–3.04 (m, 1H), 2.80 (m, 1H),
1.39 (m, 9H). HRMS [MH]+ calcd for C23H28NO6:
414.1911, found: 414.1914.
4.18. (S, S)-2((N-(tert-Butoxycarbonyl)morpholin-2-yl)
(phenyl)methylthio)benzoic acid (20)
4.21. Radiosynthesis of [11C]3
A solution of 16 (15 mg, 0.03 mmol) and 1 M NaOH
(0.3 mL, 0.3 mmol) in ethanol (1.5 mL) was refluxed
overnight. After cooling to room temperature, the crude
product was directly purified by flash chromatography
on silica eluted with CH2Cl2/MeOH (95:5) to afford 20
as colorless oil (8 mg, 57%): 1H NMR (CDCl3,
400 MHz), d 7.92 (d, J = 7.3 Hz, 1 H), 7.12–7.33 (m,
8H), 4.35(d, J = 7.6 Hz, 1H), 3.95 (d, J = 11.4 Hz, 1H),
3.78 (m, 3H), 3.51–3.58 (m, 1H), 2.91–2.98 (m, 2H),
1.36 (s, 9H). HRMS [MH]+ calcd for C23H28NO5S:
430.1683, found: 430.1688.
(2S,3S)-N-tert-Butoxycarbonyl-2-[a-(2-hydroxyphenyl-
thio)benzyl]morpholine (13) (1 mg) was dissolved in
DMF (0.25 mL) followed by addition of 0.1 M Bu4-
NOH (aq) (22 lL). The solution was placed in a sealed
conical vial and cooled to 0 ꢁC, and [11C]CH3I was bub-
bled through the solution at 0 ꢁC. The solution was
heated at 100 ꢁC for 10 min, 6 M HCl(aq) (0.21 mL)
was added, and the solution was heated at 100 ꢁC for
7 min, cooled to 0 ꢁC, and neutralized by addition of
6 M NH4OH (0.22 mL). The solution was purified by
semipreparative HPLC (Water XTerra Prep RP18
5 lm, 19 · 100 mm; 55:45:0.1 v/v/v MeOH/H2O/Et3N;
9 mL/min; tR = 10 min). The dose formulation was con-
ducted by the same procedure described above as for
[11C]1 .
4.19. Radiosynthesis of [11C]1
To a suspension of 18 (1 mg, 2.0 lmol) in THF (0.3 mL)
was added potassium tert-butoxide (4 lL, 1.0 M in
THF) at room temperature. The reaction vial was
sealed, mixed, and kept at room temperature for
30 min until [11C]CH3I was delivered as a gas and bub-
bled through the solution with ice bath cooling. After
delivery of [11C]CH3I, the reaction vial was warmed to
40 ꢁC for 2 min, 6 M HCl (0.21 mL) was added, and
the solution was heated at 100 ꢁC for 7 min, cooled to
0 ꢁC, and neutralized by addition of 6 M NH4OH
(0.22 mL). The solution was purified by semipreparative
HPLC (Water XTerra Prep RP18 5 lm, 19 · 100 mm;
56:44:0.1 v/v/v MeOH/H2O/Et3N; 9 mL/min; tR
(range) = 10.5–12.5 min). The desired fractions were
combined, diluted with double volume of water and
loaded onto a Waters C18 SepPak cartridge that was
washed with saline (0.9% NaCl, 40 mL) and ethanol
(0.5 mL). The radioactive product was washed out of
the cartridge by absolute ethanol (1.5 mL) into a sealed
sterile vial containing 3.5 mL of saline. The resulting
solution was transferred under argon pressure through
a Millipore filter (pore size 1.0 lm) followed by a smaller
4.22. Radiosynthesis of 2-[18F]fluoroethyl brosylate and 3-
[18F]fluoropropyl brosylate
[18F]HF was produced with a Siemens 11 MeV RDS 112
cyclotron by employing the 18O(p,n)18F reaction in
H218O. The [18F]HF was transferred to a chemical pro-
cessing control unit (CPCU), collected on a trap/release
cartridge (DW-TRC, D&W Inc., Oakdale, TN), re-
leased with K2CO3(aq) (0.9 mg in 0.6 mL H2O), and
added to a CH3CN solution of Kryptofix 222 (5 mg,
in 1 mL). The solution was placed in a 110 ꢁC oil bath,
the solvent was evaporated under a N2(g) flow, and
CH3CN (3 mL) was added and evaporated in order to
azeotropically dry the Kryptofix 222/K18F. 1,2-Dibrosy-
lethane or 1,3-dibrosylpropane (4 mg in 1 mL CH3CN)
was added, the reaction mixture was heated at 90 ꢁC
for 10 min, and the 2-[18F]fluoroethyl brosylate or 3-
[18F]fluoropropyl brosylate trapped on a silica Sep-Pak
(previously prepped with 10 mL Et2O). The 2-[18F]flu-

792
F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
oroethyl brosylate or 3-[18F]fluoropropyl brosylate was
eluted with Et2O, and Et2O solution was transferred to
a hot cell under N2(g) pressure and collected in a V-tube
to give 2-[18F]fluoroethyl brosylate or 3-[18F]fluoropro-
pyl brosylate in 74% or 67% radiochemical yield (de-
cay-corrected from transfer of H18F(aq) to the CPCU).
The V-tube was placed in an 80 ꢁC oil bath and the
Et2O evaporated with an Ar flow.
ard Cobra II automated gamma-counter and decay cor-
rected. The measurement was repeated three times. The
logP7.4 values were calculated for each replicate with the
following equation: log P7.4 = log[(counts in octanol
phase)/(counts in buffer phase)]. The logP7.4 measure-
ments from each replication were averaged to give the
logP7.4 values for the radiotracer.
4.26. In vitro competition assays
4.23. Radiosynthesis of [18F]4 and [18F]5
Competition assays were performed based on methods
reported previously.31,32 Cell membranes prepared from
a human embryonic kidney cell line (HEK-293) stably
expressing hNET or hSERT (a gift from Randy Blakely,
Ph.D., Vanderbilt University) and Madin Darby canine
kidney cells stably expressing hDAT (a gift from Dr.
Gary Rudnick, Yale University) were used in these as-
says. Cells were grown to confluency in DMEM con-
taining 10% fetal bovine serum and Geneticin sulfate
and then harvested using pH 7.4 phosphate-buffered sal-
ine (PBS) containing 0.53 mM ethylenediaminetetraace-
tic acid at 37 ꢁC. Cell pellets were prepared through
centrifugation at 400g for 10 min, the supernatant was
decanted, and the pellets were homogenized with a Poly-
tron PT3000 (Brinkman, Littau, Switzerland) at
11,000 rpm for 12 s in 30 vol of PBS. The resulting cell
membrane suspensions were centrifuged at 43,000g for
10 min, the supernatants were decanted, and the result-
ing pellets were stored at À70 ꢁC until used in the assays.
A solution of (2S,3S)-N-tert-butoxycarbonyl-2-[a-(2-
hydroxyphenylthio)benzyl]morpholine (13) (1 mg) and
0.1 M Bu4NOH (aq) (23 lL) in DMF (0.26 mL) was
added to 2-[18F]fluoroethyl brosylate or 3-[18F]fluoro-
propyl brosylate. The reaction mixture was heated at
100 ꢁC for 10 min, 6 M HCl (0.21 mL) was added, and
the solution was heated at 100 ꢁC for another 7 min,
cooled to 0 ꢁC, and neutralized by addition of 6 M
NH4OH (0.22 mL). The solution was purified by semi-
preparative HPLC (Water XTerra Prep RP18 5 lm,
19 · 100 mm). The retention time of [18F]4 was tR
(range) = 23–27 min with buffered mobile phase consist-
ing 50:50:0.1 v/v/v MeOH/H2O/Et3N at a 5.5-mL/min
flow rate. The retention time of [18F]5 was tR
(range) = 19–22 min with buffered mobile phase consist-
ing 55:45:0.1 v/v/v MeOH/H2O/Et3N at a 7-mL/min
flow rate. The dose formulation was conducted by the
same procedure described above as for [11C]1. The total
synthesis time was 120 min from EOB with an 11%
radiochemical yield for [18F]4 and a 9% radiochemical
yield for [18F]5 (decay-corrected from transfer of
[18F]fluoroethyl brosylate and [18F]fluoropropyl brosy-
late to the hot cell).
Competition assays were performed in 13 · 100 mm
polystyrene tubes in a 2.0 mL final volume consisting
of 1.7 mL of assay buffer, 100 lL of competing ligand
in assay buffer, 100 lL of radioligand in assay buffer,
and 100 lL of cell membrane suspension (corresponding
to 30–70 lg of protein) in assay buffer. Cell membrane
pellets were characterized prior to competition assays
to determine membrane concentrations that gave opti-
mal signal while not significantly affecting the concentra-
tion of the free radioligand. The cell membrane pellets
were resuspended in the appropriate volume of assay
buffer through brief homogenization using a Polytron
PT3000. Competing ligands were assayed in triplicate
at 12 concentrations ranging from 10À13 to 10À5 M.
To ensure solubility, the competing ligands were dis-
solved in 1:1 ethanol/5 mM HCl and then serially di-
luted in 5 mM HCl.
4.24. Quality control of purified radioligands
The radiochemical purities of radioligands were deter-
mined by an analytical radio-HPLC (waters Nova-Pak
C18 3.9 · 150 mm) using a UV detector and a radioactiv-
ity detector with or without coinjection with unlabeled
reference compounds. The specific activities were deter-
mined by the UV absorbance of the radioactive peaks as
compared with standard curves of unlabeled reference
compounds. The analytical HPLC column was eluted
with 70:30:0.1 v/v/v MeOH/H2O/Et3N for [11C]1,
[18F]4, and [18F]5, and 65:35:0.1 v/v/v MeOH/H2O/
Et3N for [11C]2, [11C]3, and [11C]6 at a 1 mL/min flow
rate. The retention times of radioligands were 4.08 min
for [11C]1, 3.53 min for [11C]2, 4.41 min for [11C]3,
3.55 min for [18F]4, 3.70 min for [18F]5, and 3.70 min
for [11C]6, respectively.
For NET assays, the assay buffer consisted of 53 mM
Tris buffer, 320 mM NaCl, and 5.3 mM KCl (pH 7.4
at 4 ꢁC) and the equilibrium incubation time was 4 h
at 4 ꢁC. The radioligand used for NET assays
was [3H]nisoxetine obtained from Dupont NEN
(3000 GBq/mmol). For SERT assays, the assay buffer
consisted of 53 mM Tris buffer, 126 mM NaCl, and
5.3 mM KCl (pH 7.9 at room temperature) and the equi-
librium incubation time was 2 h at room temperature.
The radioligand used for SERT assays was [3H]citalo-
pram obtained from Dupont NEN (Boston, MA,
3100 GBq/mmol). For the DAT assay, the assay buffer
consisted of 42 mM sodium phosphate buffer and
320 mM sucrose (pH 7.4 at room temperature) and the
equilibrium incubation time was 1 h at room tempera-
4.25. Lipophilicity measurements
Measurement of distribution coefficients of radiolabeled
compounds was performed according to a previously re-
ported procedure.30 Briefly, the test tubes containing
2 mL of 1-octanol, 2 mL of 0.02 M sodium phosphate
buffer (pH 7.4), and ꢀ5–10 lCi portion of the radio-
tracer were vortexed for 10 min at room temperature
and then centrifuged for 5 min. Samples (0.5 mL) from
the 1-octanol and buffer layers were counted in a Pack-

F. Zeng et al. / Bioorg. Med. Chem. 16 (2008) 783–793
793
15. Wilson, A. A.; Johnson, D. P.; Mozley, D.; Hussey, D.;
Ginovart, N.; Nobrega, J.; Garcia, A.; Meyer, J.; Houls, S.
Nucl. Med. Biol. 2003, 30, 85.
16. Schou, M.; Halldin, C.; Sovago, J.; Pike, V. W.; Gulyas,
B.; Mozley, P. D.; Johnson, D. P.; Hall, H.; Innis, R. B.;
Farde, L. Nucl. Med. Biol. 2003, 30, 707.
17. Schou, M.; Halldin, C.; Sovago, J.; Pike, V. W.; Hall, H.;
Gulyas, B.; Mozley, P. D.; Dobson, D.; Shchukin, E.;
Innis, R. B.; Farde, L. Synapse 2004, 53, 57.
18. Seneca, N.; Gulyas, B.; Varrone, A.; Schou, M.; Airaksi-
nen, A.; Tauscher, J.; Vandenhende, F.; Kielbasa, W.;
Farde, L.; Innis, R. B.; Halldin, C. Psychopharmacology
2006, 188, 119.
19. Kanegawa, N.; Kiyono, Y.; Kimura, H.; Sugita, T.;
Kajiyama, S.; Kawashima, H.; Ueda, M.; Kuge, Y.; Saji,
H. Eur. J. Nucl. Med. Mol. Imaging 2006, 33, 639.
20. Tamagnan, G. D.; Brenner, E.; Alagille, D.; Staley, J.
K.; Haile, C.; Koren, A.; Early, M.; Baldwin, R. M.;
Tarazi, F. I.; Baldessarini, R. J.; Jarkas, N.; Goodman,
M. M.; Seibyl, J. P. Bioorg. Med. Chem. Lett. 2007, 17,
533.
ture. The radioligand used for DAT assays was
125I]RTI-55 obtained from Dupont NEN (Boston,
MA, 2200 Ci/mmol, 2.0 nM final concentration).
[
All of the assays were initiated by the addition of the cell
membrane suspension. At the end of the incubation, the
assays were terminated by the addition of ꢀ5 mL of as-
say buffer at 4 ꢁC followed by rapid vacuum filtration
with 3· 5 mL washes with assay buffer at 4 ꢁC through
GF/B filters (Whatman Inc., Clifton, NJ) presoaked in
assay buffer containing 0.3% polyethyleneimine. Data
from the competition curves were analyzed, and Ki val-
ues were calculated using GraphPad Prism software
(GraphPad software, San Diego, CA). The Ki values
were reported as the geometric mean of at least three
separate assays, each in duplicates, for each compound.
Acknowledgments
21. Lin, K.-S.; Ding, Y.-S.; Kim, S.-W.; Kil, K.-E. Nucl. Med.
Biol. 2005, 32, 415.
This project was supported by an investigator initiated
research grant from Wyeth Laboratories. The authors
are supported by NIH 1-R21-MH-66622-01, MH-
42088, RR-000039, MH-39415, and MH-52899.
22. In a paper, see Ref. 25, J. Boot et al., reported that the
racemic mixture of compound 3 containing (S,S) and
(R,R) enantiomers was prepared and resolved using chiral
HPLC. One enantiomer was found to be potent and
selective toward NET, whereas the opposite enantiomer
was potent for both the NET and SERT. However, the
assignment of absolute stereochemistry of resolved enan-
tiomers was not reported in this paper.
References and notes
23. Brenner, E.; Baldwin, R. M.; Tamagnan, G. Org. Lett
2005, 7, 937.
24. Mitsunobu, O. Synthesis 1981, 1.
25. Boot, J.; Cases, M.; Clark, B. P.; Findlay, J.; Gallagher, P.
T.; Hayhurst, L.; Man, T.; Montalbetti, C.; Rathmell, R.
E.; Rudyk, H.; Walter, M. W.; Whatton, M.; Wood, V.
Bioorg. Med. Chem. Lett. 2005, 15, 699.
1. Blakely, R. D.; DeFelice, L. J.; Hartzell, H. C. J. Exp.
Biol. 1994, 196, 263.
2. Pacholcyzk, T.; Blakely, R. D.; Amara, S. G. Nature 1991,
350, 350.
3. Galli, A.; DeFelice, L. J.; Duke, B. J.; Moore, K. R.;
Blakely, R. D. J. Exp. Biol. 1995, 198, 2197.
4. Ordway, G. A.; Klimek, V. CNS Spectrums 2001, 6, 697.
5. Ressler, K. J.; Nemeroff, C. B. Biol. Psychiatry 1999, 46,
1219.
26. Wilson, A. A.; DaSilva, J. N.; Houle, S. Appl. Radiat. Isot.
1995, 46, 765.
27. Millington, J. E.; Pattison, F. L. M. Can. J. Chem. 1956,
34, 1532.
6. Lee, C. M.; Snyder, S. H. Proc. Natl. Acad. Sci. U.S.A.
1981, 78, 5250.
28. Zessin, J.; Gucker, P.; Ametamey, S. M.; Steinbach, J.;
Brust, P.; Vollenweider, F. X.; Johannsen, B.; Schubiger,
P. A. J. Label. Compd. Radiopharm. 1999, 42, 1301.
7. Biegon, A.; Rainbow, T. J. Neurosci. 1983, 3, 1069.
8. Van Dort, M. E.; Kim, J. H.; Tluczek, L.; Wieland, D. M.
Nucl. Med. Biol. 1997, 24, 707.
´
29. Schou, M.; Pike, V. W.; Varrone, A.; Gulyas, B.; Farde,
9. Tejani-Butt, S. M. J. Phar. Exp. Ther. 1992, 260, 427.
10. Haka, M. S.; Kilbourn, M. R. Nucl. Med. Biol. 1989, 16,
771.
L.; Halldin, C. J. Label. Compd. Radiopharm. 2006, 49,
1007.
30. Wilson, A. A.; Houle, S. J. Label. Compd. Radiopharm.
1999, 42, 1277.
11. Ding, Y.-S.; Lin, K.-S.; Logan, J.; Benveniste, H.; Carter,
P. J. Neurochem. 2005, 94, 337.
31. Dischino, D. D.; Welch, M. J.; Kilbourn, M. R.; Raichle,
M. E. J. Nucl. Med. 1983, 24, 1030.
12. Kung, M.-P.; Choi, S.-R.; Hou, C.; Zhuang, Z.-P.;
Foulon, C.; Kung, H. F. Nucl. Med. Biol. 2004, 31, 533.
13. McConathy, J.; Owen, M. J.; Kilts, C. D.; Malveaux, E.
J.; Camp, V. M.; Votaw, J. R.; Nemeroff, C. B.;
Goodman, M. M. Nucl. Med. Biol. 2004, 31, 705.
14. Ding, Y.-S.; Lin, K. S.; Gaza, V.; Carter, P.; Alexoff, D.;
Logan, J., et al. Synapse 2003, 50, 345.
32. Owens, M. J.; Morgan, W. N.; Plott, S. J.; Nemeroff, C. B.
J. Pharmacol. Exp. Ther. 1997, 283, 1305.
33. Plisson, C.; McConathy, J.; Martarello, L.; Malveaux, E.
J.; Camp, V. M.; Williams, L.; Votaw, J. R.; Goodman,
M. M. J. Med. Chem. 2004, 47, 1122.