Radiolabeling of a high potency cannabinoid subtype-1 receptor ligand, N-(4-fluoro-benzyl)-4-(3-(piperidin-1-yl)-indole-1-sulfonyl) benzamide (PipISB), with carbon-11 or fluorine-18

Article abstract of DOI:10.1002/jlcr.1491

PipISB [N-(4-fluoro-benzyl)-4-(3-(piperidin-1-yl)-indole-1-sulfonyl) benzamide, 9] was identified as a selective high potency CB₁ receptor ligand. Here we describe the labeling of 9 with positron-emitters to provide candidate radioligands for i

Full text of DOI:10.1002/jlcr.1491

Research Article
Received 29 August 2007,
Revised 20 November 2007,
Accepted 28 November 2007
Published online 15 February 2008 in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1491
Radiolabeling of a high potency cannabinoid
subtype-1 receptor ligand, N-(4-fluoro-benzyl)
-4-(3-(piperidin-1-yl)-indole-1-sulfonyl)
benzamide (PipISB), with carbon-11 or
fluorine-18
Ã
Sean R. Donohue,a,b Christer Halldin,b Magnus Schou,b Jinsoo Hong,a
Lee Phebus,c Eyassu Chernet,c Stephen A. Hitchcock,c,d Kevin M.
Gardinier,c Kevin M. Ruley,c Joseph H. Krushinski,c John Schaus,c and
Victor W. Pikea
PipISB [N-(4-fluoro-benzyl)-4-(3-(piperidin-1-yl)-indole-1-sulfonyl)benzamide, 9] was identified as a selective high potency CB1
receptor ligand. Here we describe the labeling of 9 with positron-emitters to provide candidate radioligands for imaging brain
CB1 receptors with positron emission tomography (PET). The radiolabeling of 9 was achieved by two methods, method A with
carbon-11 and method B with fluorine-18. In method A, [11C]9 was prepared in one step from [11C]carbon monoxide, itself
prepared from cyclotron-produced [11C]carbon dioxide. In method B, [18F]9 was prepared from cyclotron-produced
[
18F]fluoride ion in a two-stage, four-step synthesis with [18F]4-fluoro-benzyl bromide as a labeling agent. The radiosynthesis
time for method A was 44min; decay-corrected radiochemical yields (RCYs) from [11C]carbon monoxide ranged from 3.1 to
11.6% and specific radioactivities ranged from 21 to 67GBq/lmol. The radiosynthesis time for method B was 115 min; RCYs
from [18F]fluoride ion ranged from 1.5 to 5.6% and specific radioactivities ranged from 200 to 348 GBq/lmol. With these
methods, [11C]9 and [18F]9 may be prepared in adequate activity and quality for future evaluation as PET radioligands.
Keywords: CB1 receptor; PET; carbon-11; fluorine-18; arylsulfonyl-substituted indole; PipISB
need for radioligands for use with PET for effective imaging of
CB1 receptors in vivo. Considerable progress has been made
Introduction
Cannabinoid type-1 (CB1) receptors are implicated in a range of
neuropsychiatric disorders and hence modulation of these
receptors is considered to have broad therapeutic potential. Most
notably, rimonabant (Acomplias or Zimultis; SR 141716A) (1,
Figure 1),1 a potent and selective CB1 receptor inverse agonist, has
demonstrated efficacy for the treatment of obesity2 and smoking
abatement.3 When CB1 receptor inverse agonists are given alone or
in combination with other receptor ligands, the treatment
potential may extend to memory deficits,4 neuroinflammatory
disorders,4 depression,5 anxiety,6 stress5 and schizophrenia.7
Based on the use of in vitro techniques, substantial progress
has been made in understanding CB1 receptor activation,
inactivation and inverse activation. Nevertheless, how changes
in CB1 receptor populations and occupancy activate a patho-
physiological or therapeutic response in living subjects is not
well understood. Also, greater clarity is needed to delineate
which neurobiological disorders are likely related to CB1
receptor abnormalities. The development of brain CB1 receptor
imaging with positron emission tomography (PET), as a ‘higher-
order’ in vivo assay, is therefore of importance; there is a strong
recently in the development of promising PET radioligands for
CB1 receptors. These include inverse agonists from the 1,5-
diarylpyrazole platform of rimonabant (e.g. [123/124I]AM281 (2),8,9
[11C]JHU75528 (3)10,11 and [11C]JHU75575 (4),11 Figure 1) and
from two other structurally distinct platforms (i.e. [11C]MePPEP
(5)12 and [18F]MK-9470 (6), Figure 1).13,14 Earlier attempts at
radioligand development had focused on the agonist, D8-THC
aMolecular Imaging Branch, National Institute of Mental Health, National
Institutes of Health, Bethesda, MD 20892, USA
b
Karolinska Institutet, Department of Clinical Neuroscience, Psychiatry Section,
Karolinska Hospital, S-17176 Stockholm, Sweden
cLilly Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285, USA
dPresent address: Amgen, One Amgen Way, Thousand Oaks, CA 91320, USA
*Correspondence to: Sean R. Donohue, Molecular Imaging Branch, National
Institute of Mental Health, National Institutes of Health, Building 10, Rm B3
C346A, 10 Center Drive, Bethesda, MD 20892-1003, USA.
E-mail: donohues@intra.nimh.nih.gov
J. Label Compd. Radiopharm 2008, 51 146–152
Copyright r 2008 John Wiley & Sons, Ltd.

S. R. Donohue et al.
R5
CN
N
O
NH
O
R2
HN
O
O
N
N
N
H
N
R3
N
F
R1
O
F3C
O
R4
1
2
3
4
5
5
6
1
2
3
, R = Cl, R = Me, R = Cl, R = Cl, R = CH2
1
2
, R = I, R = Me, R3 = Cl, R4 = Cl, R5 = O
1
2
, R = OMe, R = CN, R3 = Cl, R4 = Cl, R5 = CH2
4, R1 = OMe, R2 = CN, R3 = Br, R4 = H, R5 = CH2
N
N
O
OH
F
O
S
O
N
O
O
O
O
N
HN
7
8
9
Figure 1. Structures of rimonabant (1), AM281 (2), JHU75528 (3), JHU75575 (4), MePPEP (5), MK-9470 (6), D8-THC (7), WIN-55212-8 (8) and PipISB (9).
(7, Figure 1)15 and the aminoalkylindole core of WIN55212-2 (8, 1H-indole (10), as previously described,18 and showed both high
Figure 1),16,17 but were unsuccessful.
potency at CB1 receptors (Kb = 1.5 nM) and high selectivity
Recently, a series of arylsulfonyl-substituted indoles was versus CB2 receptors (44667) (Table 1). Overall, these properties
disclosed as a new class of CB1 receptor inverse agonists with of 9 compared well with other tested imaging radioligands
promising therapeutic potential.18 We sought to explore this (Table 1) and encouraged us to pursue the labeling of 9 with
class of CB1 receptor ligand to develop a radioligand for positron-emitters to provide candidate radioligands for PET
neuroreceptor imaging with PET.
imaging of brain CB1 receptors. The precursors (11, 13) that we
We considered that one of these ligands, PipISB [N-(4-fluoro- required for use in radiolabeling were synthesized from 10 in
benzyl)-4-(3-(piperidin-1-yl)-indole-1-sulfonyl)benzamide; 9] might few steps and in moderate yields (Figure 2).
compare well with previously tested imaging radioligands with
[11C]9 was prepared in one step from [11C]carbon monoxide
regard to critical parameters,19–21 such as high potency and through palladium-catalyzed coupling between the aryl halide
selectivity. Furthermore, we considered 9 to be amenable to 11 and 4-fluoro-benzylamine (14) (Figure 3) based on a
labeling with carbon-11 (b1, t1/2 = 20.4 min) or fluorine-18 (b1, previously reported general methodology.24 The radiosynthesis
t
1/2 = 109.7min). The opportunity to label with either positron was performed in a lead-shielded automated Synthia radio-
emitter may be advantageous for avoiding troublesome metabo- chemistry apparatus22 that had been equipped with an
lites in vivo and for matching the half-life of the radiolabel to the autoclave for performing carbonylations with [11C]carbon
pharmacokinetics of the radioligand. Here, we report the potency monoxide.25 The required [11C]carbon monoxide was obtained
and selectivity of PipISB (9) for binding to CB1 receptors and also by passing cyclotron-produced [11C]carbon dioxide over heated
its radiolabeling by two different pathways (methods A and B). In molybdenum.26 [11C]9 was purified with reverse-phase high-
method A, [11C]PipISB ([11C]9) was prepared within a commercially performance liquid chromatography (HPLC) (Figure 4). The
available Synthia apparatus22 from [11C]carbon monoxide. In fraction containing [11C]9 (tR = 10 min) was evaporated to
method B, [18F]PipISB ([18F]9) was prepared within a custom-made dryness and formulated for intravenous injection first by
remotely controlled apparatus23 using [18F]4-fluoro-benzyl bro- dissolution in sterile saline containing hydroxypropyl-b-cyclo-
mide as the labeling agent.
dextrin (to avoid sticking of the radioligand to the glass vessel)
and then final sterile filtration. The overall radiosynthesis time
was 44 min. Non-optimized decay-corrected radiochemical
yields (RCYs) of formulated [11C]9 from [11C]carbon monoxide
ranged from 3.1 to 11.6%. [11C]9 was obtained in acceptable
radiochemical purity (498%) and chemical purity (490%) and
was free of labeling precursors. Specific radioactivities at the end
of synthesis (EOS) ranged between 21 and 67 GBq/mmol.
Results and discussion
The computed lipophilicity of PipISB (9) (c Log D7.4 = 5.10) is high
but is lower than those of all other radioligands reported as
being promising for imaging CB1 receptors, with only one
exception (Table 1). Ligand 9 was prepared from 3-pyridine-1-yl-
J. Label Compd. Radiopharm 2008, 51 146–152
Copyright r 2008 John Wiley & Sons, Ltd.
www.jlcr.org

S. R. Donohue et al.
Table 1. Properties of CB1 receptors ligands 1–6 and PipISB (9)
a
Potency at CB1 receptors (KB or IC50) Potency at CB2 receptors (KB or IC50) CB1 vs CB2 c Log D7.4
a
Ligand
(nM)
(nM)
selectivity
b
b
b
1( Rimonabant)
2 (AM281)
6.1 KB
8.9 KB
46776 KB
1069 KB
41111b
120b
54
84.2
610
6.95b
5.66b
5.34
4.80
5.60
5.76
5.10
b
3 (JHU75528)
4 (JHU75575)
5 (MePPEP)
6 (MK-9470)
9 (PipISB)
31.3 KB
21.3 KB
0.6 KB
1710 KB
1640 KB
c
c
363 KB
44 IC50
d
d
0.7 IC50
1.5 KB
63
44667
47000 KB
aIn vitro potencies determined through GTPg35S assay.
bData from reference [33].
cData from reference [12].
dData from reference [13].
The radioactive residue was dissolved in
a
mixture of
ethanol–propylene glycol with sterile sodium phosphate buffer
solution and then sterile filtered.30 The overall radiosynthesis
time was 115 min. Non-optimized RCYs of formulated [18F]9
ranged from 1.5 to 5.6% from [18F]fluoride ion. [18F]9 was obtained
in high radiochemical purity (498%) and chemical purity (490%)
and was free of labeling precursors. Specific radioactivity (EOS)
ranged from 200 to 348 GBq/mmol. Product identity was confirmed
by its co-elution with reference 9 in analytical HPLC.
N
N
a
N
O
S
N
H
O
I
10
11
b
Experimental
Materials
All solvents and reagents were purchased from commercial
sources and used without further purification. Compounds 3
and 4 were prepared at NIMH using previously published
procedures.11 PipISB (9) and 3-piperidine-1-yl-1H-indole (10)
were prepared according to Smith et al.18 4-(Trimethylamino)-
benzaldehyde trifluoromethanesulfonate was prepared as de-
scribed previously.31
N
N
c
N
O
N
O
S
O
S
O
O
CN
NH2
12
13
General methods
Figure 2. Synthesis of precursors 11 and 13. Conditions and yields: a) 4-iodo-
benzenesulfonyl chloride, t-BuOK, dioxane, RT, 2h, yield 41%; b) 4-cyano-
benzenesulfonyl chloride, t-BuOK, dioxane, RT, 4h, yield 27%; c) NaOH, EtOH-
H2O, 30% H2O2, RT, 18h, yield 52%.
Column chromatography was performed on silica gel columns
(35–65 mm; Isco Inc., Lincoln, NE, USA). 1H NMR spectra
(400 MHz) were measured on an INOVA spectrometer (Varian,
Palo Alto, CA, USA). Abbreviations s, d, dd, m and br denote
singlet, doublet, double doublet, multiplet and broad, respec-
tively. Mass detection was performed with an 1100 Series
LC/MSD single quadrupole spectrometer (Agilent, Santa Clara,
CA, USA) with ESI interface. LC-MS was performed with a
reversed-phase column (Luna C18: 150 ꢀ 2 mm, 5 mm; Phenom-
enex, Torrance, CA, USA) eluted with a mixture of aq. 10 mM
ammonium formate and acetonitrile (25:75 v/v) at 150 mL/min.
After electrospray ionization of the eluted test sample, ions
between m/z 150 and 750 were acquired. LC analyses, unless
otherwise stated, were performed with a heated (507101C)
reversed phase column (Xterra C18:2.1 ꢀ 50 mm, 3.5 mm; Waters
Corp., Milford, MA, USA) eluted at 1 mL/min with a gradient of
MeCN–MeOH (50:50 v/v) (A) and 0.2% aq. HCO2NH4 (B), with A
increased linearly from 5 to 100% v/v over 7 min and then held for
1 min. The chemical purities of 11 and 13 were assessed with an
identical chromatographic system eluted with a mixture of A and
B, with A held at 1% for 2 min, then increased linearly to 30% over
Product identity was confirmed with liquid chromatography-
mass spectrometry (LC-MS) of the carrier and observation of co-
elution of radioligand with reference compound 9 in analytical
HPLC.
[18F]9 was prepared by treating the carboxamide 13 with
[18F]4-fluoro-benzyl bromide ([18F]17) in an overall two-stage,
four-step synthesis. In the first stage (Figure 5), [18F]17 was
synthesized in three steps from cyclotron-produced [18F]
fluoride ion, as described previously,27,28,29 but within
custom-made remotely controlled apparatus23.
a
In the second stage, brief treatment of the carboxamide 13
with [18F]17 in the presence of sodium hydride gave [18F]9 in
about 36% RCY (n = 2) (Figure 5). The dichloromethane was
evaporated off, leaving a dimethylformamide (DMF) solution of
the crude reaction product. [18F]9 was purified from the DMF
solution with reversed phase HPLC (Figure 6). The fraction
containing [18F]9 (tR = 15.8 min) was evaporated to dryness.
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 146–152

S. R. Donohue et al.
N
N
S
O
O
H2N
+
11C
HN
O
a
11C]CO
+
11
[
F
14
F
[
11C]9
Figure 3. Radiosynthesis of [11C]PipISB ([11C]9). Conditions and isolated decay-corrected radiochemical yield; a) Pd(PPh3)4, THF, 6000 psi, 150 1C, 5 min, 3.1–11.6% RCY.
Figure 4. Chromatogram from the HPLC separation of [11C]PipISB ([11C]9) (absorbance at 245 nm response, Panel A; radioactivity response, Panel B).
4min, and finally increased linearly to 100% over 1 min. Radio- between water and ethyl acetate. The aqueous layer was
thin-layer chromatography (TLC) was performed on silica gel and extracted with ethyl acetate. The ethyl acetate extracts were
analyzed with an automatic AR-2000 Imaging Scanner (Bioscan combined, washed with water and then brine, dried over
Inc., Washington, DC, USA) with Winscan 2.2 software (LabLogic sodium sulfate, filtered and concentrated under reduced
Inc., Sheffeld, UK). Radioactivity was measured in calibrated pressure to give crude product (1.5 g). Silica gel chromatography
ionization chambers (AtomLab 300, Biodex, Shirley, NY, USA) (eluent: 10% ethyl acetate–hexanes) of this crude product gave
and corrected for background and physical decay.
pure 11 as a yellow foam (600 mg, 41%). 1H NMR (dimethyl
sulfoxide (DMSO)-d6) d 7.874 (d, 1H, J = 7.9 Hz), 7.866 (d, 2H,
J = 8.8 Hz), 7.59 (d, 2H, J = 8.8 Hz), 7.50 (d, 1H, J = 7.5 Hz), 7.30 (dd,
1H, J = 7.9, 7.9 Hz), 7.20 (dd, 1H, J = 7.6, 7.6 Hz), 6.99 (s, 1H),
2.94–2.90 (m, 4H), 1.66–1.59 (m, 4H), 1.53–1.46 (m, 2H); MS (ESI)
m/z 467 (M111); HPLC, 100%.
4-(3-Piperidin-1-yl-indole-1-sulfonyl)-benzonitrile (12): Potassium
tert-butoxide (420 mg, 3.7 mmol) was added to a solution of 10
(600 mg, 3.0 mmol) in dioxane (30 mL). The mixture was stirred
at room temperature for 30 min. A solution of 4-cyanobenzene-
sulfonyl chloride (634 mg, 3.2 mmol) in dioxane (2 mL) was
added and the reaction mixture stirred at room temperature for
4 h. The reaction mixture was diluted with ethyl acetate and
then washed with saturated sodium bicarbonate and brine. The
CB1 and CB2 GTPc35S binding assays
GTPg35S binding of 9 in Sf9 cell membranes expressing human
CB1 or CB2 receptors was measured in a 96-well format with a
modified antibody capture technique described previously.32,33
Agonist EC50 and antagonist IC50 values were calculated with
Activity Base software using a four-parameter fit and KB values
(binding constants) calculated as follows:
KB ¼ IC50=½1 þ ½Agonistꢁ=EC50ꢁ
Chemistry
1-(4-Iodo-benzenesulfonyl)-3-piperidin-1-yl-1H-indole (11): Potassium organic layer was dried over sodium sulfate and then
tert-butoxide (440 mg, 3.9 mmol) was added to a cooled (ice evaporated to dryness. The residue was purified with silica gel
bath) solution of 3-piperidin-1-yl-1H-indole (10) (665 mg, chromatography (eluent: 10% ethyl acetate–hexanes) to yield 12
3.3 mmol) in dioxane (10 mL). The ice bath was removed and as a yellow solid (298 mg, 27%). 1H NMR (DMSO-d6) d 8.04 (d, 2H,
the reaction warmed to room temperature over 40 min. A J = 8.8 Hz), 7.97 (d, 2H, J = 8.8 Hz), 7.91 (d, 1H, J = 8.4 Hz), 7.51 (d,
solution of 4-iodo-benzenesulfonyl chloride (960 mg, 3.2 mmol) 1H, J = 7.9 Hz), 7.32 (dd, 1H, J = 7.7, 7.7 Hz), 7.21 (dd, 1H, J = 7.7,
in dioxane (6 mL) was added and the reaction mixture stirred at 7.7 Hz), 7.04 (s, 1H), 2.96–2.91 (m, 4H), 1.66–1.59 (m, 4H),
room temperature for 2 h. The reaction mixture was partitioned 1.54–1.46 (m, 2H).
J. Label Compd. Radiopharm 2008, 51 146–152
Copyright r 2008 John Wiley & Sons, Ltd.
www.jlcr.org

S. R. Donohue et al.
O
O
b
a
c
H
H
OH
Br
[
18F]F-
+
+
18F
18F
18F
Me3N
-
CF3SO3
18
18
18
15
F]
16
17
[ F]
[
[
F]
d
N
N
S
O
O
O
HN
18F
[
18F]9
Figure 5. Radiosynthesis of [18F]PipISB ([18F]9). Conditions and yields: a) Kryptofix 2.2.2, K2CO3, DMSO, 110 1C, 7 min; b) aq. NaBH4, RT, 5 min; c) PPh3Br2, RT, 5 min; d) NaH,
DMF-DCM, RT, 3 min, 1.5–5.6% RCY (from [18F]fluoride ion).
evaporated to dryness. The residue was purified with silica gel
chromatography (eluent: 50% ethyl acetate–hexanes) to yield 13
1
as a yellow solid (78 mg, 52%). H NMR (DMSO-d6) d 8.04 (br s,
1H), 7.93 (d, 2H, J = 8.8 Hz), 7.92 (d, 1H, J = 8.2 Hz), 7.87 (d, 2H,
J = 8.7 Hz), 7.54 (br s, 1H), 7.49 (d, 1H, J = 7.9 Hz), 7.30 (dd, 1H,
J = 7.7, 7.7 Hz), 7.19 (dd, 1H, J = 7.6, 7.6 Hz), 7.03 (s, 1H), 2.95–2.90
(m, 4H), 1.66–1.59 (m, 4H), 1.53–1.46 (m, 2H); MS (ESI) m/z 384
(M111); HPLC, 100%.
Calculation of c log D7.4
The c Log D7.4 values of ligands (Table 1) were calculated with
Pallas 3.0 software (Compudrug International Inc., San Francisco,
CA, USA).
Radiochemistry instrumentation
Labeling with [11C]carbon monoxide was performed within a
lead-shielded automated Synthia apparatus (Uppsala Imanet,
Uppsala, Sweden), equipped with an autoclave module.22 The
Synthia consists of a Gilson Aspec AL, XYZ-robot, a Compact PCI
running a Visual C11 program, an evaporation unit, specially
designed modules (for labeling chemistry, solid-phase extrac-
tion, HPLC purification and sterile filtration) and photo-diodes
for monitoring radioactivity levels.
Figure 6. Chromatogram from the HPLC separation of [18F]PipISB ([18F]9) (upper
trace shows absorbance response and lower trace radioactivity response).
Labeling with [18F]fluoro-benzyl bromide was performed within
a lead-shielded custom-made remotely controlled apparatus
(Laboratech, Stockholm, Sweden).23 This apparatus is essentially
based on a stainless-steel box housing control electronics and
supporting on its outside face nine 3-way valves (Bu¨rkert-
Contromatic AB, Malmo, Sweden), a heating block with a cooling
gun, two reaction vessels with moving needles and a gas manifold.
4-(3-Piperidin-1-yl-indole-1-sulfonyl)-benzamide (13): Sodium
hydroxide pellets (520 mg) were dissolved in water (3.9 mL)
and the solution was chilled to ꢂ101C. A solution of 12 (145 mg,
0.4 mmol) in ethanol (7.8 mL) was added to the sodium
hydroxide solution. Then hydrogen peroxide solution (30%,
390 mL) was added over 20 min. The mixture was slowly warmed
to room temperature, stirred for 18 h, quenched with a sodium
bisulfite solution (20% w/v, 3 mL), and then stirred for 30 min at
room temperature. The ethanol was removed under vacuum
and the resulting aqueous solution partitioned between ethyl
acetate and water. The water was removed and the organic layer
washed with brine, dried over sodium sulfate and then
Radiochemistry
NCA [11C]carbon monoxide: No-carrier-added (NCA) [11C]carbon
dioxide was produced at the National Institutes of Health (NIH)
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 146–152

S. R. Donohue et al.
with a PETtrace cyclotron (GE Medical Systems, Milwaukee, WI, heated to 1101C for 7 min with a solution of (4-trimethylamino)
USA) to implement the 14N(p,a)11C reaction on nitrogen (300 psi) benzaldehyde trifluoromethanesulfonate (7 mg, 22.4 mmol) in
containing a low concentration of oxygen (1%). Typically, proton DMSO (500 mL). The reaction mixture was then cooled, diluted
irradiations (16.5 MeV, 45 mA) for 20 min gave about 52 GBq of with water (2 ꢀ 7 mL) and eluted through a short cartridge (Sep-
[11C]carbon dioxide. [11C]Carbon monoxide was generated by Paks Plus C-18, AccellTM; Waters Corp.) that had been precondi-
passing the [11C]carbon dioxide in helium carrier gas (20 mL/ tioned by elution first with ethanol (10 mL) and then deionized
min) over molybdenum wire (3.5–5.8 g; Strem Chemicals, water (10 mL). [18F]4-Fluorobenzaldehyde ([18F]15) (silica gel-
Newport, MA, USA) contained in a quartz column (7 mm TLC: Rf = 0.59, chloroform) was retained on the cartridge.
ID ꢀ 12 mm OD) while heated in a furnace at 8451C.31 The Aqueous sodium borohydride (30 mg/mL) solution was slowly
resulting [11C]carbon monoxide was trapped on GC-grade silica passed through the cartridge and then nitrogen was passed
(19 mg, 100/120, GC grade, Alltech, Raleigh, NC, USA) contained through for 5 min. The generated and retained [18F]4-fluoro-
in a stainless-steel tube (1 mm ID ꢀ 26 mm) that was cooled in benzyl alcohol ([18F]16) (silica gel-TLC: Rf = 0.15, chloroform) was
liquid nitrogen.
then eluted from the column with dichloromethane (2 mL),
NCA [18F]fluoride ion: NCA [18F]fluoride ion was produced at passed through an extraction tube (3 mL; Supelco, Stockholm,
the Karolinska Hospital with a PETtrace cyclotron (GE Medical SE) that had been filled with potassium carbonate (1 g in upper
Systems, Uppsala, Sweden) to implement the 18O(p,n)18F part) and magnesium sulfate (800 mg in lower part), and finally
reaction on 18O-enriched water (95 atom %, 2 mL). Typically, emptied into a vial containing triphenyldibromophosphorane
proton bombardments (16.4 MeV, 35 mA) for 20 min gave 27 GBq (90 mg, 213 mmol). The reaction mixture was vortexed with
of [18F]fluoride ion.
remote control at room temperature for 5 min. The crude [18F]4-
[11C]PipISB ([11C]9): Trapped [11C]carbon monoxide was fluoro-benzyl bromide ([18F]17) (silica gel-TLC: Rf = 0.83, chloro-
released from the silica gel by heating the gel to 601C. The form) was then purified by passage through a Sep-Paks Plus
[11C]carbon monoxide was transferred into a stream of helium silica cartridge.
(20 mL/min) into the high-pressure stainless-steal micro-auto-
[18F]PipISB ([18F]9): The purified [18F]4-fluoro-benzyl bromide
clave [which had been pre-washed with tetrahydrofuran (THF) was immediately reacted with precursor (13) (1mg, 2.6 mmol)
(10 mL) and then dried by heating to 1501C for 5 min)]. A THF dissolved in DMF (300mL) containing sodium hydride (1mg, 60%
solution (200 mL) of tetrakistriphenylphosphine palladium(0) in dispersion) at room temperature for 3 min. The remaining
(1.5 mg, 1.3 mmol), aryl iodide (11) (2.1 mg, 4.5 mmol) and 4- dichloromethane was removed by heat under a nitrogen flow.
fluoro-benzylamine (14) (20 mL, 180 mmol) was injected into the HPLC mobile phase (700mL), consisting of MeCN-–aq. HCO2NH4
loading vessel of the apparatus and then flushed into the micro- (0.1M) (55:45 v/v), was added to the reaction vial containing the
autoclave with a steam of helium (90 mL/min). The autoclave remaining DMF solution and the resultant solution was
was sealed and heated under pressure (41 MPa, 6000 psi) for transferred to a HPLC loop. The [18F]9 was purified on a m-
5 min in a 1501C bath. The autoclave contents were then Bondapak C-18 column (300ꢀ 7.8 mm, 10mm; Waters Corp.)
emptied into an evacuated V-vial and the THF (0.2 mL plus eluted at 7 mL/min. Mobile phase was evaporated off from the
0.2 mL rinse) was evaporated off under nitrogen. The crude fraction containing radioligand [18F]9 (tR = 15.8min). The residue
residue was dissolved in MeCN–aq. HCO2NH4 (0.1 M) (65:35 v/v) was re-dissolved in a mixture of ethanol–propylene glycol (30:70
and injected onto a Luna C-18 column (10 ꢀ 250 mm; Phenom- v/v, 3 mL) with sterile sodium phosphate buffer solution (0.2M,
enex, Torrance, CA, USA) eluted at 6 mL/min with a gradient of pH= 7.4, 5 mL), and filtered through a sterile filter (MillexsGV,
65% acetonitrile (C) and 35% HCO2NH4 (0.1 M) (D) with C 0.22mm pore size; Millipore Corp., Corrigtwohill, Co. Cork, Ireland).
increasing linearly to 80% over 10 min. The fraction containing
The radiochemical and chemical purities of [18F]9 were
m-Bondapak C-18 column
[11C]9 (tR = 9.8 min) was collected and the mobile phase was determined by HPLC on
a
removed by rotary evaporation at 60 1C. The residue of [11C]9 (300 ꢀ 3.9 mm, 10 mm; Waters corp.) eluted with MeCN–aq.
was reconstituted in a solution of hydroxypropyl-b-cyclodextrin HCO2NH4 (0.1 M) (55:45 v/v) at 3 mL/min. The identity of [18F]9
(TCI America, Portland, OR, USA) (10% w/v) in sterile saline for was confirmed by its co-elution with reference 9 (tR = 9.95 min)
injection (5–10 mL), in readiness for intravenous injection into under the same HPLC conditions.
monkey.
The radiochemical and chemical purities of [11C]9 were Specific radioactivity determination
assessed with HPLC on a Luna C-18 column (4.6 ꢀ 250 mm,
Phenomemex) eluted with MeCN–aq. HCO2NH4 (0.1 M) (80:20
Specific radioactivities (GBq/mmol) at EOS were determined with
analytical radio-HPLC, calibrated for absorbance (l = 254 nm)
response per mass of ligand, using the same HPLC conditions
v/v) at 2 mL/min (tR = 4.7 min). The identity of the product was
confirmed by its co-elution with reference 9 in the HPLC system
that were used to determine the radiochemical and chemical
and also LC-MS of associated carrier in NCA material.
purities for [11C]9 or [18F]9, respectively. The radioactivity of the
[18F]4-Fluoro-benzyl bromide ([18F]9): NCA [18F]fluoride ion was
radioligand peak (decay corrected) (GBq) was divided by the
separated from irradiated 18O-enriched water on an anion
exchange resin (Sep-Paks Light QMA cartridge, AccellTM; Waters
Corp.) that had been preconditioned by elution first with
mass of the associated carrier peak (mmol).
potassium carbonate solution (0.5 M, 10 mL) and then deionized
with water (10 mL). The [18F]fluoride ion was transferred to a
Conclusion
10-mL vial by elution of the QMA resin with acetonitrile–water PipISB (9) was identified as a highly potent and selective CB1
(94:4 v/v, 2 mL) containing Kryptofixs 2.2.2 (9.5 mg, 25 mmol; receptor ligand with a calculated lipophilicity similar to other
Aldrich Chem. Co.) and potassium carbonate (1.7 mg, 12 mmol). PET radioligands targeting brain CB1 receptors and was
Solvents were evaporated off at 1301C under a nitrogen flow. successfully labeled with No-carrier-added (NCA) carbon-11 or
The resulting [18F]fluoride–K1–Kryptofix 2.2.2 complex was fluorine-18, hence permitting future evaluation of the two
J. Label Compd. Radiopharm 2008, 51 146–152
Copyright r 2008 John Wiley & Sons, Ltd.
www.jlcr.org

S. R. Donohue et al.
generated radioligands for in vivo imaging of brain CB1
receptors with PET.
S. Krause, A. Vanko, A. Van Hecken, P. Dupont, I. De Lepeleire,
P. Rothenberg, S. A. Stoch, J. Cote, W. K. Hagmann, J. P. Jewell,
L. S. Lin, P. Liu, M. T. Goulet, K. Gottesdiener, J. A. Wagner, J. de
Hoon, L. Mortelmans, T. M. Fong, R. J. Hargreaves, Proc. Natl. Acad.
Sci. USA 2007, 104, 9800–9805.
Acknowledgements
[14] P. Liu, L. S. Lin, T. G. Hamill, J. P. Jewell, T. J. Lanza Jr, R. E. Gibson,
S. M. Krause, C. Ryan, W. Eng, S. Sanabria, X. Tong, J. Wang,
D. A. Levorse, K. A. Owens, T. M. Fong, C. P. Shen, J. Lao, S. Kumar,
W. Yin, J. F. Payack, S. A. Springfield, R. Hargreaves, H. D. Burns,
M. T. Goulet, W. K. Hagmann, J. Med. Chem. 2007, 50,
3427–3430.
The authors thank other members of the PET group at the Karolinska
Institutet and of the Molecular Imaging Branch (NIMH) for assistance.
This study was supported in part by the Intramural Research
Program of the National Institutes of Health (NIH), specifically the
National Institute of Mental Health (NIMH) and was also performed [15] A. Charalambous, G. Marciniak, C. Y. Shiue, S. L. Dewey, D. J.
Schlyer, A. P. Wolf, A. Makriyannis, Pharmacol. Biochem. Behav.
1991, 40, 503–507.
[16] P. G. Willis, O. A. Pavlova, S. I. Chefer, D. B. Vaupel, A. G. Mukhin,
A. G. Horti, J. Med. Chem. 2005, 48, 5813–5822.
[17] J. Dhawan, H. Deng, S. J. Gatley, A. Makriyannis, T. Akinfeleye,
M. Bruneus, A. A. Dimaio, A. N. Gifford, Synapse 2006, 60, 93–101.
[18] J. B. Allen, A. K. Amegadzie, K. M. Gardinier, G. S. Gregory, S. A.
Hitchcock, P. J. Hoogstraat, W. J. Jones Jr, D. L. Smith, PCT Int. App.
WO2005066126, 2005.
[19] C. Halldin, B. Gulyas, O. Langer, L. Farde, Q. J. Nucl. Med. 2001, 45,
130–152.
[20] V. W. Pike, J. Psychopharm. 1993, 7, 139–158.
[21] M. Laruelle, M. Slifstein, Y. Huang, Mol. Imaging Biol. 2003, 5,
363–375.
under a CRADA agreement between NIMH, the Karolinska Institutet
and Eli Lilly and Co. Mr Sean R. Donohue received support through
the National Institutes of Health–Karolinska Institutet Graduate
Training Partnership in Neuroscience. We thank the Clinical PET
Center (NIH) for performing cyclotron irradiations.
References
[1] M. Rinaldi-Carmona, F. Barth, M. Heaulme, D. Shire, B. Calandra,
C. Congy, S. Martinez, J. Maruani, G. Neliat, D. Caput, P. Ferrara,
P. Soubrie, J. C. Breliere, G. Le Fur, FEBS Lett. 1994, 350, 240–244.
[2] S. P. Vickers, G. A. Kennett, Curr. Drug Targets 2005, 6, 215–223.
[3] C. Cohen, E. Kodas, G. Griebel, Pharmacol. Biochem. Behav. 2005,
81, 387–395.
[4] R. N. Takahashi, F. A. Pamplona, M. S. Fernandes, Neurosci. Lett.
2005, 380, 270–275.
[5] J. M. Witkin, E. T. Tzavara, R. J. Davis, X. Li, G. G. Nomikos, Trends
Pharmacol. Sci. 2005, 26, 609–617.
[6] B. Costa, A. E. Trovato, M. Colleoni, G. Giagnoni, E. Zarini, T. Croci,
Pain 2005, 116, 52–61.
[7] G. Griebel, J. Stemmelin, B. Scatton, Biol. Psychiatry 2005, 57,
261–267.
[8] S. J. Gatley, R. Lan, N. D. Volkow, N. Pappas, P. King, C. T. Wong, A.
N. Gifford, B. Pyatt, S. L. Dewey, A. Makriyannis, J. Neurochem.
1998, 70, 417–423.
[22] P. Bjurling, R. Reineck, G. Westerberg, A. D. Gee, J. Sutcliffe,
˚
B. Langstrom, Proceedings of the VIth Workshop on Targetry and
Target Chemistry, TRIUMF 1995, 282–284.
¨
[23] R. N. Krasikova, V. V. Zaitsev, P. Truong, M. Schou, C. Halldin, Eur. J.
Nucl. Med. Mol. Imaging 2002, 29, S376.
˚
¨
[24] T. Kihlberg, B. Langstrom, J. Org. Chem. 1999, 64, 9201–9205.
˚
¨
[25] T. Kihlberg, B. Langstrom, Patent, US 2004/0197257 A1, 2004.
[26] S. K. Zeisler, M. Nader, A. Theobald, F. Oberdorfer, Appl. Radiat. Isot.
1997, 48, 1091–1095.
[27] C. Lemaire, P. Damhaut, A. Plenevaux, R. Cantineau, L. Christiaens,
M. Guillaume, Appl. Radiat. Isot. 1992, 43, 485–494.
[28] K. Hatano, T. Ido, R. Iwata, J. Label. Compd. Radiopharm. 1991, 29,
373–380.
[29] R. N. Krasikova, V. V. Zaitsev, S. M. Ametamey, O. F. Kuznetsova,
O. S. Fedorova, I. K. Mosevich, Y. N. Belokon, S. Vyskocil,
S. V. Shatik, M. Nader, P. A. Schubiger, Nucl. Med. Biol. 2004, 31,
597–603.
[30] C. Foged, C. Halldin, J. Hiltunen, C. Braestrup, C. Thomsen,
H. C. Hansen, T. Suhara, S. Pauli, C. G. Swahn, P. Karlsson,
S. Larsson, L. Farde, Nucl. Med. Biol. 1996, 23, 201–209.
[9] G. Berding, U. Schneider, P. Gielow, R. Buchert, F. Donnerstag, W.
Brandau, W. H. Knapp, H. M. Emrich, K. Muller-Vahl, Psychiatry Res.
2006, 147, 249–256.
[10] A. G. Horti, H. Fan, H. Kuwabara, J. Hilton, H. T. Ravert, D. P. Holt, M.
Alexander, A. Kumar, A. Rahmim, U. Scheffel, D. F. Wong, R. F.
Dannals, J. Nucl. Med. 2006, 47, 1689–1696.
[11] H. Fan, H. T. Ravert, D. P. Holt, R. F. Dannals, A. G. Horti, J. Label.
Compds. Radiopharm. 2006, 49, 1021–1036.
[12] F. Yasuno, A. K. Brown, S. S. Zoghbi, J. H. Krushinski, E.
Chernet, J. Tauscher, J. M. Schaus, L. A. Phebus, A. K. Chesterfield,
C. C. Felder, R. L. Gladding, J. Hong, C. Halldin, V. W. Pike, R. B.
´
[31] O. Langer, F. Dolle, H. Valette, C. Halldin, F. Vaufrey, C. Fuseau,
˚
`
C. Coulon, M. Ottaviani, K. Nagren, M. Bottlaender, B. Maziere,
C. Crouzel, Bioorg. Med. Chem. 2001, 9, 677–694.
[32] N. W. DeLapp, J. H. McKinzie, B. D. Sawyer, A. Vandergriff, J.
Falcone, D. McClure, C. C. Felder, J. Pharmacol. Exp. Ther. 1999, 289,
946–955.
Innis, Neuropsychopharmacology
sj.npp.1301402).
2007
(DOI:
10.1038/
[33] S. R. Donohue, C. Halldin, V. W. Pike, Bioorg. Med. Chem. 2006, 14,
3712–3720.
[13] H. D. Burns, K. Van Laere, S. Sanabria-Bohorquez, T. G. Hamill,
G. Bormans, W. S. Eng, R. Gibson, C. Ryan, B. Connolly, S. Patel,
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 146–152