Evaluation of [11C]hemicholinium-15 and [18F]hemicholinium-15 as new potential PET tracers for the high-affinity choline uptake system in the heart

Article abstract of DOI:10.1016/j.bmc.2006.11.014

[¹¹C]Hemicholinium-15 ([¹¹C]HC-15) and [¹⁸F]hemicholinium-15 ([¹⁸F]HC-15) have been synthesized as new potential PET tracers for the heart high-affinity choline uptake (HACU) system. [¹¹C]HC-15 was prepared by N-[¹¹C]methylation of the appropriate precursor, 4-methyl-2-phenyl-morpholin-2-ol, using [¹¹C]CH₃OTf in 55-70% radiochemical yield decay corrected to end of bombardment (EOB) and 2-3 Ci/μmol specific activity at end of synthesis (EOS). [¹⁸F]HC-15 was prepared by N-[¹⁸F]fluoromethylation of the precursor using [¹⁸F]FCH₂OTf in 20-30% radiochemical yield decay corrected to EOB and >1.0 Ci/μmol specific activity at EOS. The biodistribution of both compounds was determined in rats at 20 min post-intravenous injection, and the results show the heart region uptakes 1.32 ± 0.75%ID/g in R-ventricle for [¹¹C]HC-15 and 1.28 ± 0.81%ID/g in L-ventricle for [¹⁸F]HC-15, respectively. The dynamic PET imaging studies of [¹¹C]HC-15 in rats were acquired 60 min post-intravenous injection of the tracer using the IndyPET-II scanner. For the blocking experiments, the rats were intravenously pretreated with 3.0 mg/kg of unlabeled HC-15 prior to [¹¹C]HC-15 injection. [¹¹C]HC-15 rat heart PET studies show rapid heart uptake to give clear heart images. The rat heart PET blocking studies found no significant blocking effect. The dynamic PET studies in normal and ablated dogs were performed using Siemens PET scanner with [¹³N]NH₃, [¹¹C]HC-15, and [¹⁸F]HC-15. PET studies in dogs of both [¹¹C]HC-15 and [¹⁸F]HC-15 also show significant heart uptake and give images of the heart. However, there is no significant change in [¹¹C]HC-15 L-ventricle uptake following radiofrequency ablation in the dog. These results suggest that the localization of HC-15 tracers in the heart is mediated by non-specific processes, and the visualization of HC-15 tracers on the heart is related to non-specific binding of HACU.

Full text of DOI:10.1016/j.bmc.2006.11.014

Bioorganic & Medicinal Chemistry 15 (2007) 1289–1297
Evaluation of [11C]hemicholinium-15 and [18F]hemicholinium-15
as new potential PET tracers for the high-affinity choline
uptake system in the heart
Mingzhang Gao,a Michael A. Miller,a Timothy R. DeGrado,a Bruce H. Mock,a
John C. Lopshire,b Joshua G. Rosenberger,b Cristian Dusa,b Mithilesh Kumar Das,b
William J. Groh,b Douglas P. Zipes,b Gary D. Hutchinsa and Qi-Huang Zhenga,*
aDepartment of Radiology, Indiana University, Indianapolis, IN 46202, USA
bKrannert Institute of Cardiology, Indiana University, Indianapolis, IN 46202, USA
Received 9 October 2006; revised 31 October 2006; accepted 8 November 2006
Available online 10 November 2006
Abstract—[11C]Hemicholinium-15 ([11C]HC-15) and [18F]hemicholinium-15 ([18F]HC-15) have been synthesized as new potential
PET tracers for the heart high-affinity choline uptake (HACU) system. [11C]HC-15 was prepared by N-[11C]methylation of the appro-
priate precursor, 4-methyl-2-phenyl-morpholin-2-ol, using [11C]CH3OTf in 55–70% radiochemical yield decay corrected to end of
bombardment (EOB) and 2–3 Ci/lmol specific activity at end of synthesis (EOS). [18F]HC-15 was prepared by N-[18F]fluoromethy-
lation of the precursor using [18F]FCH2OTf in 20–30% radiochemical yield decay corrected to EOB and >1.0 Ci/lmol specific activity
at EOS. The biodistribution of both compounds was determined in rats at 20 min post-intravenous injection, and the results show the
heart region uptakes 1.32 0.75%ID/g in R-ventricle for [11C]HC-15 and 1.28 0.81%ID/g in L-ventricle for [18F]HC-15, respective-
ly. The dynamic PET imaging studies of [11C]HC-15 in rats were acquired 60 min post-intravenous injection of the tracer using the
IndyPET-II scanner. For the blocking experiments, the rats were intravenously pretreated with 3.0 mg/kg of unlabeled HC-15 prior
to [11C]HC-15 injection. [11C]HC-15 rat heart PET studies show rapid heart uptake to give clear heart images. The rat heart PET
blocking studies found no significant blocking effect. The dynamic PET studies in normal and ablated dogs were performed using
Siemens PET scanner with [13N]NH3, [11C]HC-15, and [18F]HC-15. PET studies in dogs of both [11C]HC-15 and [18F]HC-15 also show
significant heart uptake and give images of the heart. However, there is no significant change in [11C]HC-15 L-ventricle uptake follow-
ing radiofrequency ablation in the dog. These results suggest that the localization of HC-15 tracers in the heart is mediated by
non-specific processes, and the visualization of HC-15 tracers on the heart is related to non-specific binding of HACU.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
vious work, we have synthesized and studied a series of
PET heart imaging agents that target either sympathetic
or parasympathetic nerve such as [11C]meta-hydroxy-
ephedrine ([11C]HED), 5-[11C]methoxybenzovesamicol
([11C]MOBV), [11C]neostigmine, [11C]edrophonium,
[11C]pyridostigmine, and N-[11C]methyl-3-[[(dimethyla-
mino)carbonyl]oxy]-2-(20,20-diphenylpropionoxymethyl)
pyridinium triflate ([11C]MDDP).4–9 In this ongoing
study, we targeted high-affinity choline uptake (HACU)
system in the heart autonomic nervous system and
developed radiolabeled hemicholinium-15 (HC-15)
derivatives as new in vivo tracers for HACU system.
The HACU system is located specifically in presynaptic
cholinergic nerve terminals and was found to be the
regulating and rate-limiting step in the synthesis of
acetylcholine, thus, it is a biomarker of the functional
status of the cholinergic presynaptic terminals.10 Like
Cardiac autonomic nervous system dysfunction plays a
significant role in the genesis of ventricular tachyar-
rhythmias and sudden cardiac death in patients who
have coronary heart disease.1 Non-invasive imaging of
heart autonomic nervous system is an active area of cur-
rent molecular imaging research.2 Radiotracer methods
using biomedical imaging modalities such as positron
emission tomography (PET) for heart imaging have
been extensively investigated and validated.3 In our pre-
Keywords: [11C]Hemicholinium-15; [18F]Hemicholinium-15; Positron
emission tomography; Heart imaging; High-affinity choline uptake.
*
Corresponding author. Tel.: +1 317 278 4671; fax: +1 317 278
9711; e-mail: qzheng@iupui.edu
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.11.014

1290
M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
hemicholinium-3 (HC-3), a golden standard for studying
the HACU system in vitro, HC-15 is an inhibitor of
HACU and is one-half of the molecule of HC-3 with sim-
ilar in vitro affinity of potent HC-3.11 To study HACU
system in vivo, we proposed to develop radiolabeled
HC-3 and HC-15 derivatives as PET probes. Because
HC-3 is a bis-quaternary ammonium salt, it is relatively
difficult to radiolabel. Thus, we synthesized radiolabeled
HC-15 derivatives first. HC-15 labeled with a positron
emitting radionuclide either carbon-11 or fluorine-18
may enable non-invasive monitoring of HACU system
in the heart to study cardiac parasympathetic innerva-
tion using PET imaging technique. We present here our
initial investigation on synthesis, biodistribution, and
PET imaging of [11C]hemicholinium-15 ([11C]HC-15)
and [18F]hemicholinium-15 ([18F]HC-15) as new poten-
tial in vivo heart HACU system biomarkers.
2.2. Tissue biodistribution studies
All animal experiments were performed under a proto-
col approved by the Indiana University (IU) Institution-
al Animal Care and Use Committee (IACUC). The
in vivo biodistribution of [11C]HC-15 and [18F]HC-15
was determined in Fisher rats at 20 min post-intrave-
nous (iv) injection, and the data are listed in Table 1.
The heart region uptakes included L-atrium, R-atrium,
septum, L-ventricle, R-ventricle, and A-V node. The sig-
nificant heart ventricle uptakes were 1.32 0.75%ID/g
in R-ventricle for [11C]HC-15 and 1.28 0.81%ID/g in
L-ventricle for [18F]HC-15, respectively. The data pre-
sented here represented the average value in four rats.
Highest uptakes were found in liver, kidney, and urine.
Uptake in myocardial regions was moderate for both
11C and 18F tracers. In comparison with the uptakes in
liver, kidney, and urine, the uptake in brain, blood,
and muscle was very low. Both [11C]HC-15 and
[18F]HC-15 are radiolabeled quaternary ammonium
salts, therefore, it is not possible for the tracers to cross
the blood–brain barrier (BBB). The biodistribution data
of both tracers in brain confirmed the lipophilicity of the
tracers,18 which are consistent with their HPLC reten-
tion times.
2. Results and discussion
2.1. Chemistry and radiochemistry
The precursor 4-methyl-2-phenyl-morpholin-2-ol (3)
was synthesized from phenacyl bromide with 2-(meth-
ylamino)ethanol in 90% yield.12 The methylation of
compound 3 with methyl iodide provided HC-15 stan-
dard, 4,4-dimethyl-2-hydroxy-2-phenylmorpholinium
iodide (1a), in 92% yield. The one-step reaction of
phenacyl bromide with 2-(dimethylamino)ethanol pro-
vided another HC-15 standard, 4,4-dimethyl-2-hy-
droxy-2-phenylmorpholinium bromide (1b), in 95%
yield.13 The fluoromethyl-HC-15 standard, 4-(fluorom-
ethyl)-2-hydroxy-4-methyl-2-phenylmorpholinium chlo-
ride (2), was prepared from the reaction of precursor 3
with a fluoromethylated agent, chlorofluoromethane,
in 10% yield using a modification of the literature
procedure.14
2.3. Rat IndyPET-II imaging
Animal PET scanner, IndyPET-II, is a high-resolution
(<3 mm full-width at half-maximum), high sensitivity,
research PET scanner developed for small field-of-view
(FOV) imaging applications including rodent imaging
(mice, rats), intermediate size animals (dogs, pigs, pri-
mates), and dedicated human imaging applications
(brain, breast).19,20 In vivo dynamic PET imaging stud-
ies8,9 of the tracer [11C]HC-15 in young adult female
Sprague–Dawley rats were performed in the IndyPET-
II scanner for 60 min post iv injection of 0.5–1.0 mCi
of the tracer in a rat. Four rats were imaged. Each
was imaged twice, once following iv injection in the tail
vein with 3.0 mg/kg of unlabeled HC-15 in saline 30 min
prior to tracer administration and imaging. Time be-
tween blocking injection and tracer injection was
30 2 min (mean sd). Mean uptake values were calcu-
lated from region-of-interest (ROI) data for frames from
20 to 40 min following the start of the scans. Mean up-
take values were analyzed with a mixed effect model tak-
ing the blocker as the fixed effect and the animal as a
random effect.
[11C]HC-15 was prepared by N-[11C]methylation of the
precursor 3 using a reactive [11C]methylating agent,
[11C]methyl triflate ([11C]CH3OTf),15 and isolated by a
simplified solid-phase extraction (SPE) method16 in
55–70% decay corrected radiochemical yields based on
[11C]CO2, 10–15 min overall synthesis time from end
of bombardment (EOB), >99% radiochemical purity,
and 2–3 Ci/lmol specific activity at end of synthesis
(EOS). Either SiO2 or cation-exchange CM Sep-Pak car-
tridge was used in SPE purification technique.
Similarly, [18F]HC-15 was prepared by N-[18F]fluo-
romethylation of the precursor 3 using a novel and
reactive [18F]fluoromethylating agent, [18F]fluorometh-
A typical IndyPET-II coronal view image of [11C]HC-15
in a rat is shown in Figure 1. Image intensity is standard
uptake value (SUV), averaged from 20 to 60 min after
yl triflate ([18F]FCH2OTf)17, followed by a fast SPE
purification as mentioned above in 20–30% decay cor-
rected radiochemical yields based on K[18F]F, 20–
30 min overall synthesis time from EOB, >99% radio-
chemical purity, and >1.0 Ci/lmol specific activity at
EOS.
tracer injection. [11C]HC-15 rat heart PET dynamic
studies show rapid heart uptake and blood pool clear-
ance to give clear heart images.
The heart uptake normalized to injected dose/mass for
the four rats (13, 14, 15, and 16) was studied. Data from
blocked images are plotted as filled circles, while un-
blocked data are plotted as open circles. Except for rat
15, there is no effect. The p value from a mixed effect
The synthetic approach for the precursor, reference
standards, and tracers is outlined in Scheme 1.

M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
1291
OH
O
OH
I
O
O
Br
+
CH3I
THF
Ether
CH3NHCH2CH2OH
N
N
H3C
CH3
CH3
1a
3
OH
O
O
Br
Ether
(CH3)2NCH2CH2OH
+
N
Br
H3C
CH3
1b
OH
O
FCH2Cl
THF
3
N
Cl
H3C
Br2
CH2F
2
H2
AgOTf
~100%
11CO2
11CH3Br
11CH3OTf
11CH4
70-75%
>95%
AgOTf
CH2Br2
18FCH2Br
18FCH2OTf
K
18F/K222
11CH3
CH218F
H3C
H3C
N
N
O
18FCH2OTf, CH3CN
11CH3OTf, CH3CN
3
O
OH
11C]HC-15 ([11C]1)
OH
OTf
OTf
[
[
18F]HC-15 ([18F]2)
Scheme 1. Synthesis of HC-15 precursor, reference standards, and tracers [11C]HC-15 and [18F]HC-15.
animals (13 and 16) where it was possible to see the kid-
ney, there appears to be a significant blocking effect.
Since there are missing data, this sample is too small to
do a mixed effect calculation. A simple t test comparing
blocked to unblocked gives a p value of 0.35, showing
an insignificant result. The liver uptake normalized to
injected dose/mass for the four rats was studied. Filled
points were blocked with HC-15. Open points were not
blocked. These data shows a significant blocking effect
with a p value of 0.035. PET imaging studies in rats indi-
cated a significant accumulation of radiotracer in the liv-
er. This may be caused by the lipophilicity of the
quaternary ammonium tracer aforementioned. Heart
uptake (A), left kidney uptake (B), and liver uptake (C)
of [11C]HC-15 in mean ROI intensity · injected dose/
mass · 106 within 60 min scan time using the IndyPET-
II scanner in a representative study in rat 13 are shown
in Figure 2. From the graphs in Figure 2, we can see
the liver uptake is likely 5–10 times of the heart uptake,
which is consistent with their biodistribution data as
indicated in Table 1. The data are disappointing for the
purpose of demonstrating specificity of the heart uptake,
and thus we conclude that rat may not be a good model
for evaluation of these type compounds in the heart.
Table 1. Biodistribution data of [11C]HC-15 and [18F]HC-15 in Fisher
rat model (%ID/g) at 20 min post iv injection
Organ
[
11C]HC-15 (n = 4)
[
18F]HC-15 (n = 4)
Blood
Brain
Lung
0.181 0.095
0.022 0.012
1.302 1.161
6.284 2.843
3.618 1.810
0.201 0.048
0.046 0.023
3.423 1.510
4.378 2.613
Liver
Kidney
Urine
11.651 8.016
17.434 4.510
31.835 12.119
Heart regions
L-Atrium
R-Atrium
Septum
0.526 0.320
0.488 0.222
0.010 0.005
0.859 0.494
1.316 0.750
1.083 1.016
0.796 0.304
0.756 0.407
1.557 1.174
1.280 0.810
0.037 0.016
0.793 0.417
L-Ventricle
R-Ventricle
A-V Node
Solar plexus
Ske. musc.
0.938 0.371
0.091 0.066
0.744 0.212
0.145 0.063
The data presented here represented the average value in four rats.
model looking for effects from the blocking is 0.43, con-
firming what is clear from the graph. The initial PET
blocking studies of [11C]HC-15 with pretreatment drug
HC-15 in rats found only minor changes in rat heart
[11C]HC-15 retention and no significant blocking in the
data set. The left kidney uptake normalized to injected
dose/mass for the four rats was studied. In the two
2.4. Canine PET imaging
Previous work from our laboratory has shown that
parasympathetic fibers run along the left ventricular

1292
M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
Figure 1. A coronal view image of [11C]HC-15 distribution in a rat
acquired with the IndyPET-II scanner for 60 min. A green line
surrounds the heart region. Image intensity is SUV averaged from 20
to 60 min after tracer injection.
endocardial surface.21 Electroanatomic mapping and
radiofrequency ablation (RFA) catheter techniques ap-
plied to the dog generated canine parasympathetic
denervation model.22 To further determine the exact
binding site of the HC-15 tracers, dog PET imaging
studies were performed. PET baseline images of
[13N]NH3, [11C]HC-15, and [18F]HC-15 in a normal
dog are shown in Figure 3. The images on the left-hand
side are short-axis (SA) slices, while those on the right
are horizontal long-axis (HLA) slices. The color scales
are approximately similar for the [13N]NH3 and
[11C]HC-15, and the [18F]HC-15 color scale has been
turned up to show the uptake. Both [13N]NH3 and
[11C]HC-15 show significant heart uptake and give clear
heart images, while relatively low heart uptake was seen
with [18F]HC-15.
Due to the low uptake of [18F]HC-15 in the normal dog,
only [13N]NH3 and [11C]HC-15 were administered in the
ablated dog. [13N]NH3-PET in the dog served as a refer-
Figure 2. Heart uptake (A), left kidney uptake (B), and liver uptake
(C) of [11C]HC-15 in mean ROI intensity · injected dose/mass · 106
within 60 min scan time using the IndyPET-II scanner in a represen-
tative study in rat 13. Blocking of HACU was effected by pretreatment
of rat with HC-15.
ence imaging study for the blood flow. PET images of
[13N]NH3 in a dog before RFA and after RFA are
shown in Figure 4. The first row images are at HLA
view, the second row images are at vertical long-axis
(VLA) view, and the third row images are at SA view.
There is no significant change in the images before
RFA and after RFA. Kinetics of [13N]NH3 within
30 min scan time in a dog before RFA and after RFA
for two ROIs on the anterior L-ventricle (LV) wall are
shown in Figure 5. The open black circles are from the
normal, pre-ablation scan, and the filled green circles
are from the scan following the first ablation. The figure
shows there is a slight increase in tracer uptake follow-
ing RFA, but there is no significant change of the kinet-
ics of [13N]NH3 in a dog before RFA and after RFA.
Figure 7. The open black (pre-ablation) and filled green
(post-ablation) triangles are the [11C]HC-15 data, and
the open black (pre-ablation) squares are the [18F]HC-
15 data. Figure 7 shows that [11C]HC-15 has higher
heart LV uptake than [18F]HC-15, which is consistent
with their PET images as shown in Figure 3, and there
is no significant change of the kinetics in [11C]HC-15
LV uptake in the dog pre- and post-ablation.
3. Conclusion
PET images of [11C]HC-15 in a dog before RFA and
after RFA at HLA, VLA, and SA views are shown in
Figure 6. Likewise, there is no significant change in
the images before RFA and after RFA. Kinetics of
[11C]HC-15 within 90 min scan time in a dog pre- and
post-ablation, and kinetics of [18F]HC-15 within
90 min scan time in the same normal dog are shown in
In summary, the synthetic procedures for new C-11 and
F-18 labeled HC-15 tracers have been well developed.
Biodistribution studies show both [18F]HC-15 and
[11C]HC-15 have certain uptakes in heart regions and
would distribute to brain very little. Initial PET studies
in blocked and unblocked rats demonstrate that myo-

M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
1293
Figure 3. PET baseline images of [13N]NH3, [11C]HC-15, and [18F]HC-15 in a normal dog. The [13N]NH3 images show homogeneous perfusion of the
myocardium. Moderate uptake of [11C]HC-15 was observed throughout the myocardium, while relatively low myocardial uptake was seen with
[
18F]HC-15.
Figure 4. PET images of [13N]NH3 in a dog before RFA and after RFA. There is no significant change in the images before RFA and after RFA.
cardial uptake of [11C]HC-15 is not blocked by HC-15
administration. Initial PET studies in normal and ablat-
ed dogs indicate there is no significant change in
[13N]NH3 or [11C]HC-15 LV uptake before RFA and
after RFA-certainly nothing at all like that seen with
[11C]HED4 following denervation. [11C]HC-15 change
is similar to [13N]NH3 change, although it does slightly
trend in the direction we expect if the tracer was indicat-
ing denervation. There is no evidence of an uptake de-
fect coincident with the location of the ablation.

1294
M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
Figure 7. Kinetics of [11C]HC-15 and [18F]HC-15 within 90 min scan
time in a dog. The triangles are the [11C]HC-15 data, and the squares
are the [18F]HC-15 data. Black and green are as: pre- and post-RFA.
The heart uptake of [11C]HC-15 is higher than that of [18F]HC-15, and
there is no significant change in [11C]HC-15 heart uptake following
RFA.
Figure 5. Kinetics of [13N]NH3 within 30 min scan time in a dog pre-
ablation scan and post-ablation scan. The plot shows the kinetics for
[
13N]NH3 for two ROIs on the anterior LV wall. They are for the same
dog. The open black circles are from the normal, pre-ablation scan,
and the filled green circles are from the scan following the ablation.
There is a slight increase in tracer uptake following RFA.
macokinetics and more potent affinity for the HACU,
since it represents at present the most potent known syn-
thetic inhibitor of the HACU system. These same prop-
erties are often beneficial in a diagnostic radiotracer. To
translate therapeutic agents for diagnostic use, our fu-
ture works to develop HACU-specific, hemicholinium-
based PET probes will be the in vivo biological evalua-
tion of radiolabeled HC-3 derivatives, [11C]hemicholini-
um-3 ([11C]HC-3) and [18F]hemicholinium-3 ([18F]HC-
3), to determine the exact binding site of the potent
HACU inhibitors.
Taken collectively, these preliminary results suggest that
the heart uptakes of the HC-15 tracers are most likely
the result of non-specific binding. Therefore, we con-
clude the HC-15 compounds were unable to show spe-
cific binding of HACU, and the heart uptakes are due
to non-specific binding. The reason that radiolabeled
HC-15 compounds did not specifically bind to the
HACU system might be related to their relatively low
in vitro affinity in comparison with the really potent
HC-3 compounds.11 Among these HACU inhibitors,
HC-3 should possess the combination of favorable phar-
Figure 6. PET images of [11C]HC-15 in a dog before RFA and after RFA. There is no significant change in the images before RFA and after RFA.

M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
1295
4. Experimental
10 mmol) in anhydrous tetrahydrofuran (THF, 40 mL)
was added iodomethane (2.13 g, 15 mmol). The mixture
was stirred at room temperature for 8 h, during the time
a white solid precipitated. Then the solid was isolated by
filtration, washed with THF (2· 5 mL), and dried under
vacuum to provide compound 1a as a white solid
(3.08 g, 92%), mp 183–185 ꢁC. 1H NMR (300 MHz,
DMSO-d6) d: 3.15 (s, 3H, CH3), 3.26 (d, J = 13.2 Hz,
1H, CHH), 3.43 (s, 3H, CH3), 3.56 (d, J = 5.1, 2H,
CH2), 3.62 (d, J = 13.2, 1H, CHH), 4.08 (d,
J = 12.5 Hz, 1H, CHH), 4.38–4.47 (m, 1H, CHH), 7.33
(d, J = 1.5, 1H, OH), 7.41–7.47 (m, 3H, Ph-H), 7.55–
7.58 (m, 2H, Ph-H).
4.1. General
All commercial reagents and solvents were used without
further purification. [11C]CH3OTf was made according
to a literature procedure.15 Melting points were deter-
mined on a MEL-TEMP II capillary tube apparatus
and are uncorrected. 1H NMR spectra were recorded on
a Bruker QE 300FT-NMR spectrometerusing tetrameth-
ylsilane (TMS) as an internal standard. Chemical shift
data for the proton resonances were reported in parts
per million (ppm, d scale) relative to internal standard
TMS (d 0.0), and coupling constants (J) are reported in
hertz (Hz). The low-resolution mass spectra (LRMS)
were obtained using a Bruker Biflex III MALDI-TOF
mass spectrometer, and the high-resolution mass spectra
(HRMS) were obtained using a Kratos MS80 mass
spectrometer. Chromatographic solvent proportions are
expressed on a volume: volume basis. Thin-layer chroma-
tography was run using Analtech silica gel GF uniplates
(5 · 10 cm2). Plates were visualized by UV light. Normal
phase flash chromatography was carried out on EM Sci-
ence silica gel 60 (230–400 mesh) with a forced flow of
the indicated solvent system in the proportions described
below. All moisture- and/or air-sensitive reactions were
performed under a positive pressure of nitrogen main-
tained by a direct line from a nitrogen source.
4.2.3.
4,4-Dimethyl-2-hydroxy-2-phenylmorpholinium
bromide (1b). Compound 1b was synthesized using a
modification of the literature method.13 Phenacyl bro-
mide (6.22 g, 31.2 mmol) and 2-(dimethylamino)ethanol
(2.78 g, 31.2 mmol) were added into dry ether (200 mL),
stirred at room temperature for 24 h. A white solid was
precipitated, filtered, and washed with dry ether (2·
20 mL) to give compound 1b (8.54 g, 95%), mp 187–
188 ꢁC. 1H NMR (300 MHz, DMSO-d6) d: 3.16 (s,
3H, CH3), 3.26 (d, J = 12.5 Hz, 1H, CHH), 3.43 (s,
3H, CH3), 3.57 (d, J = 5.9, 2H, CH2), 3.63 (d,
J = 13.2, 1H, CHH), 4.07 (d, J = 12.5 Hz, 1H, CHH),
4.36–4.46 (m, 1H, CHH), 7.32 (d, J = 1.5, 1H, OH),
7.39–7.45 (m, 3H, Ph-H), 7.56 (dd, J = 1.8, 7.7 Hz,
2H, Ph-H).
Analytical HPLC was performed using a Prodigy (Phe-
nomenex) 5 lm C-18 column, 4.6 · 250 mm2; 3:1:3
CH3CN:MeOH:20 mM, pH 6.7 KHPO4 (buffer solu-
4.2.4.
4-(Fluoromethyl)-2-hydroxy-4-methyl-2-phen-
ꢀ
ylmorpholinium chloride (2). To a 50 mL pressure tube
containing anhydrous THF (15 mL) at ꢀ45 ꢁC was add-
ed compound 3 (0.50 g, 2.6 mmol), chlorofluorometh-
ane (Synquest Labs, Alachua, FL) was bubbled
through the solution for 20 min. Then the tube was
sealed. The mixture was allowed to slowly warm to
room temperature, stirred for 60 h, during the time a
white solid precipitated. The solid was filtered, washed
with dry THF (2 mL), and dried under vacuum to pro-
vide compound 2 as a white solid (68 mg, 10%), mp
tion) mobile phase; flow rate 1.5 mL/min; and UV
(254 nm) and c-ray (NaI) flow detectors. Semi-prep
SiO2 Sep-Pak cartridge and cation-exchange CM Sep-
Pak cartridge were obtained from Waters Corporate
Headquarters, Milford, MA. Sterile Millex-GS
0.22 lm vented filter unit was obtained from Millipore
Corporation, Bedford, MA.
4.2. Synthesis of precursors and reference standards
1
153–155 ꢁC. H NMR (300 MHz, DMSO-d6) d: 3.16–
4.2.1. 4-Methyl-2-phenyl-morpholin-2-ol (3). Compound
3 was synthesized using a modification of the literature
method.12 Phenacyl bromide (19.9 g, 0.10 mol) was dis-
solved in ether (200 mL) and added dropwise with stir-
ring to 2-(methylamino)ethanol (15.2 g, 0.20 mol),
which was dissolved in ether (180 mL). The mixture
was stirred at room temperature for 24 h. The superna-
tant liquid was decanted or filtered from the precipitated
N-substituted ethanolamine hydrobromide, washed with
brine (3· 100 mL), and dried over MgSO4. The solution
was concentrated to give low melting point solid com-
pound 3 (17.3 g, 90%), Rf = 0.76 (1:9, MeOH/CH2Cl2),
mp 51–52 ꢁC. 1H NMR (300 MHz, CDCl3) d: 2.16–
2.56 (m, 2H, CH2), 2.29 (s, 3H, CH3), 2.77 (dd,
J = 11.1, 20.2 Hz, 2H, CH2), 3.82 (dd, J = 2.9, 11.8 Hz,
1H, CHH), 4.23 (dt, J = 2.9, 11.8 Hz, 1H, CHH), 4.74
(s, 1H, OH), 7.31–7.38 (m, 3H, Ph-H), 7.62 (dd,
J = 1.7, 8.1 Hz, 2H, Ph-H).
3.78 (m, 7H, 2 · CH2 and CH3), 4.05–4.18 (m, 1H,
CHH), 4.42–4.52 (m, 1H, CHH), 5.46–5.62 (m, 1H,
FCHH), 5.85–6.14 (m, 1H, FCHH), 7.40–7.63 (m, 6H,
Ph-H and OH). LRMS (ESI): 226.1 (M+, 100%).
HRMS (ESI): Calcd for C12H17O2NF 226.1237. Found
226.1229.
4.3. Synthesis of target tracers
4.3.1. [11C]Hemicholinium-15 ([11C]HC-15, [11C]1). Pre-
cursor 3 (0.1–0.2 mg) was dissolved in acetonitrile
(250 lL). The mixture was placed in a sealed reaction
vessel. No carrier-added (high specific activity)
[11C]CH3OTf was passed through the reaction solution,
which was cooled at ꢁ0 ꢁC, until radioactivity reached a
maximum (ꢁ3 min), and then the reaction mixture was
heated at 80 ꢁC for 5 min. The reaction tube was con-
nected to the SiO2 Sep-Pak or CM Sep-Pak. The labeled
product mixture solution was passed onto the Sep-Pak
cartridge for SPE purification by gas pressure. The reac-
tion vessel and Sep-Pak cartridge were washed with eth-
4.2.2. 4,4-Dimethyl-2-hydroxy-2-phenylmorpholinium
iodide (1a). To a solution of compound 3 (1.93 g,

1296
M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
anol (5 mL) and water (2 mL), and the washing solution
was discarded to a waste bottle. The final product [11C]1
was eluted from the SiO2 Sep-Pak with 90:8:2 H2O/
EtOH/HOAc (2–4 mL) or the CM Sep-Pak with saline
(2–4 mL) and sterile-filtered through a 0.22 lm cellulose
acetate membrane and collected into a sterile vial. The
pH of the production solution from the SiO2 Sep-Pak
was adjusted to 5.5–7.0 with 8.4% NaHCO3 solution
(0.2–0.4 mL). Total radioactivity was assayed and the
total volume (2.5–5.0 mL) was noted. The overall syn-
thesis time was 10–15 min. The radiochemical yields de-
cay corrected to EOB, from 11CO2, were 55–70%, the
radiochemical purity was >99%, and the chemical purity
of the target tracer was >95% measured by analytical
HPLC. Retention times in the analytical HPLC
system were: tR 3 = 2.38 min, tR 1 = 1.78 min, tR
[11C]1 = 1.78 min.
frame durations were defined as 300 s for entire
3600 s scan. For the blocking experiments, the rats
were pretreated by iv injection in the tail vein with
3.0 mg/kg of unlabeled HC-15 cold drug in saline
30 min prior to iv injection of 0.5 mCi [11C]HC-15 in
the tail vein. After tracer administration, the animals
were handled as described above.
4.6. Dog denervation procedure
Mongrel dogs (20–30 kg) were premedicated with mor-
phine sulfate (2 mg/kg, im) and cephazolium sodium
(3 mg/kg, im), and anesthetized with sodium thiopental
(25 mg/kg, iv). A cuffed endotracheal tube was inserted
and anesthesia was maintained with 1–3% isoflurane.
The right groin was shaved and scrubbed with betadine
and alcohol. Sterile surgical technique was followed
throughout the procedure. An 8 French arterial sheath
was placed in right femoral artery using the modified
Seldinger technique. An electroanatomic mapping and
RFA catheter (CARTO BioSenseꢂ) was placed in left
ventricle, with creation of nearly circumferential line
of ablation approximately 1 cm below the atrioventric-
ular groove,22 avoiding the area of cardiac bundle
branches. Eighty-seven applications of RFA were used
to create line of ablation, with disruption of endocardi-
al electrical conduction confirmed after ablation com-
pleted. After a 5–7 day recovery period, repeat PET
imaging was performed to re-assess the degree of para-
sympathetic innervation. The tracer kinetic models
were employed in these studies to produce indices of
regional choline transporter levels in the heart. The
denervation procedure was performed at IU Krannert
Institute of Cardiology clean procedure room and cath-
eterization laboratory and was approved by the institu-
tional IACUC.
4.3.2. [18F]Hemicholinium-15 ([18F]HC-15, [18F]2). Pre-
cursor
3 (0.6–1.0 mg) was dissolved in acetone
(0.6 mL), and the mixture was placed in a sealed reac-
tion vessel. No carrier-added (high specific activity)
[18F]fluoromethyl bromide prepared by a literature
method23 was passed through a silver triflate column
at 200 ꢁC to form [18F]FCH2OTf.17 [18F]FCH2OTf was
trapped in the reaction solution, and then the reaction
mixture was heated at 40 ꢁC for 10 min. The subsequent
purification will be the same as above for the tracer
[11C]1 to provide the final product [18F]2. The overall
synthesis time was 20–30 min. The radiochemical yields
decay corrected to EOB were 20–30%. Retention times
in the analytical HPLC system were: tR 3 = 2.38 min,
tR 2 = 1.80 min, tR [18F]2 = 1.80 min.
4.4. Biodistribution
The Fisher rats (200–300 g) were injected intravenously
with sub-pharmacologic doses (1–3 mCi) of [11C]HC-
15 and [18F]HC-15 via the tail vein while under con-
4.7. Canine PET imaging
The Siemens CTI EXACT HR+ PET scanner at IU
PET facility was used for the canine PET imaging
studies. A normal Mongrel dog underwent 30-min
dynamic PET imaging with [13N]NH3 for myocardial
blood flow, and 90-min dynamic PET imaging with
[11C]HC-15 and [18F]HC-15 for baseline imaging stud-
ies. The same dog before RFA and after RFA re-
ceived 30-min dynamic PET scan with [13N]NH3 to
image myocardial blood flow, and 90-min dynamic
PET scan with [11C]HC-15 to image parasympathetic
nerve endings.
scious restraint. At 20 min post-injection, rats were sac-
rificed by decapitation under halothane anesthesia, their
tissues quickly excised, weighed, and the decay-correct-
ed radioactive content measured using a Packard Cobra
Quantum gamma counter. The results are expressed as
percentage of injected dose (%ID) and as percentage
of injected dose per mass of tissue (%ID/g). For calcula-
tion of total blood activity, blood mass was assumed to
be 7% of the body mass.
4.5. Rat PET imaging
The IndyPET-II scanner designed and developed with-
in IU PET facility19,20 was used for rat PET imaging
studies. The young adult female Sprague–Dawley rat
(250–300 g) was anesthetized with acepromazine
(0.2 mg/kg, im) and torbugesic (0.2 mg/kg, im). 0.5–
1.0 mCi of [11C]HC-15 was administered intravenously
to the rat via the tail vein. The PET images of the
tracer were acquired in IndyPET-II scanner by the or-
dered subsets expectation maximization (OSEM) using
6 subsets/4 iterations for 60 min dynamic scans from a
rat post iv injection of 0.5 mCi of the tracer, and
Acknowledgments
This work was partially supported by the Donald W.
Reynolds Foundation and the Indiana Genomics Initia-
tive (INGEN) of Indiana University, which is supported
in part by Lilly Endowment Inc. The authors thank Shu-
yan Wang, Kevin Perry, Susan Giger, and Tanya Mar-
tinez for their assistance in animal studies, and
Barbara Glick-Wilson and Michael Sullivan for their
assistance in the production of radiolabeled precursors

M. Gao et al. / Bioorg. Med. Chem. 15 (2007) 1289–1297
1297
11CH3OTf and K18F/Kryptofix2.2.2. The referees’ criti-
9. Wang, J.-Q.; Miller, M. A.; Mock, B. H.; Lopshire, J. C.;
Groh, W. J.; Zipes, D. P.; Hutchins, G. D.; Zheng, Q.-H.
Bioorg. Med. Chem. Lett. 2005, 15, 4510.
10. Gilissen, C.; de Groot, T. J.; Bronfman, F.; van Leuven,
F.; Verbruggen, A. M.; Bormans, G. M. J. Nucl. Med.
2003, 44, 269.
11. Smart, L. A. J. Med. Chem. 1983, 26, 104.
12. Cromwell, N. H.; Tsou, K.-C. J. Am. Chem. Soc. 1949, 71,
993.
13. Rekka, E.; Kourounakis, P. Eur. J. Med. Chem. 1989, 24,
179.
cisms and editor’s comments for the revision of the man-
uscript are greatly appreciated.
References and notes
1. Paul-Labrador, M.; Polk, D.; Dwyer, J. H.; Velasquez, I.;
Nidich, S.; Rainforth, M.; Schneider, R.; Merz, C. N.
Arch. Intern. Med. 2006, 166, 1218.
2. Knuuti, J.; Sipola, P. Q. J. Nucl. Med. Mol. Imaging 2005,
49, 97.
3. Bengel, F. M.; Schwaiger, M. J. Nucl. Cardiol. 2004, 11,
603.
4. Hutchins, G. D.; Chen, T.; Carlson, K. A.; Fain, R. L.;
Winkle, W.; Vavrek, T.; Mock, B. H.; Zipes, D. P. J. Nucl.
Med. 1999, 40, 846.
14. DeGrado, T. R.; Coleman, R. E.; Wang, S.; Baldwin, S.
W.; Orr, M. D.; Robertson, C. N.; Polascik, T. J.; Price,
D. T. Cancer Res. 2001, 61, 110.
15. Mock, B. H.; Mulholland, G. K.; Vavrek, M. T. Nucl.
Med. Biol. 1999, 26, 467.
16. Zheng, Q.-H.; Mock, B. H. Biomed. Chromatogr. 2005, 19,
671.
5. Olgin, J. E.; Sih, H. J.; Hanish, S.; Jayachandran, J.
V.; Wu, J.; Zheng, Q.-H.; Winkle, W.; Mulholland, G.
K.; Zipes, D. P.; Hutchins, G. Circulation 1998, 98,
2608.
17. Iwata, R.; Pascali, C.; Bogni, A.; Furumoto, S.; Terasaki,
K.; Yanai, K. Appl. Radiat. Isot. 2002, 57, 347.
18. Liu, X.; Wang, J.-Q.; Zheng, Q.-H. Biomed. Chromatogr.
2005, 19, 379.
6. Mulholland, G. K.; Zheng, Q.-H.; Mock, B. H.; Carlson,
K. A.; Giger, S.; Vavrek, M. T.; Fain, R. L.; Hutchins, G.
D. J. Nucl. Med. 1999, 40(5 Supple. S), 495.
7. Zheng, Q.-H.; Liu, X.; Fei, X.; Wang, J.-Q.; Mock, B. H.;
Glick-Wilson, B. E.; Sullivan, M. L.; Hutchins, G. D.
Bioorg. Med. Chem. Lett. 2003, 13, 1787.
19. Frese, T.; Rouze, N. C.; Bouman, C. A.; Sauer, K.;
Hutchins, G. D. IEEE Trans. Med. Imaging 2003, 22, 258.
20. Rouze, N. C.; Hutchins, G. D. IEEE Trans. Nucl. Sci.
2003, 50, 1491.
21. Takahashi, N.; Barber, M. J.; Zipes, D. P. Am. J. Physiol.
1985, 248, H89.
8. Wang, J.-Q.; Miller, M. A.; Fei, X.; Stone, K. L.;
Lopshire, J. C.; Groh, W. J.; Zipes, D. P.;
Hutchins, G. D.; Zheng, Q.-H. Nucl. Med. Biol.
2004, 31, 957.
22. Kaseda, S.; Zipes, D. P. Am. J. Physiol. 1988, 255, H534.
23. DeGrado, T. R.; Baldwin, S. W.; Wang, S.; Orr, M. D.;
Liao, R. P.; Friedman, H. S.; Reiman, R.; Price, D. T.;
Coleman, R. E. J. Nucl. Med. 2001, 42, 1805.