Development of fluorine-18-labeled 5-HT(1A) antagonists

Article abstract of DOI:10.1021/jm980456f

We have synthesized five fluorinated derivatives of WAY 100635, N-{2- [4-(2-methoxyphenyl)piperazino]ethyl)-N-(2-pyridyl)cyclohexanecarboxamide (4a), using various acids in place of the cyclohexanecarboxylic acid (CHCA, 2a) in the reaction scheme. The five acids are 4-fluorobenzoic acid (FB, 2b), 4-fluoro-3-methylbenzoic acid (MeFB, 2c), trans-4- fluorocyclohexanecarboxylic acid (FC, 2d), 4-(fluoromethyl)benzoic acid (FMeB, 2e), and 3-nitro-4(fluoromethyl)benzoic acid (NFMeB, 2f) (see Scheme 1). These compounds were radiolabeled with fluorine-18, and their biological properties were evaluated in rats and compared with those of [¹¹C]carbonyl WAY 100635 ([carbonyl-¹¹C]4a). [Carbonyl-¹¹C]4a cleared the brain with a biological half-life averaging 41 min. The metabolite-corrected blood radioactivity had a half-life of 29 min. [¹⁸F]FCWAY ([¹⁸F]4d) gave half- lives and intercepts comparable to [carbonyl¹¹C]4a in the brain, but the blood clearance was faster. [¹⁸F]FBWAY ([¹⁸F]4b) showed an early rapid net efflux from the whole brain, clearing with a biological half-life of 35 min. The metabolite-corrected blood half-life was 41 min. The comparable whole brain and blood half-lives for Me[¹⁸F]FBWAY ([¹⁸F]4c) were 16 and 18 min, respectively. For each compound, the corresponding carboxylic acid was identified as a major metabolite in blood. Fluoride was also found after injection of [¹⁸F]4d. However, for all compounds there was a good correlation (R > 0.97) between the differential uptake ratio (DUR, (%ID/g) x body weight (g)/100) in individual rat brain regions at 30 min after injection and the concentration of receptors as determined by in vitro quantitative autoradiography in rat. Specific binding ratios [region of interest (ROI)/cerebellum-1] in control studies for cortex (Ctx) and hippocampus (H) were higher for [carbonyl¹¹C]4a and [¹⁸F]4d compared to [¹⁸F]4b and [¹⁸F]4c. [¹⁸F]4d has similar pharmacokinetic properties and comparable specific binding ratios to [carbonyl-¹¹C]4a. Fifty nanomoles of 4a blocked only 30% of the specific binding of [¹⁸F]4d, while complete blockade was obtained from co-injection of 200 nmol of 4a (H/Cb-1 from 17.2 to 0.6). [¹⁸F]4b and [¹⁸F]4c showed lower specific binding ratios than [carbonyl-¹¹C]4a and [¹⁸F]4d. [¹⁸F]4c was superior to [¹⁸F]4b since its specific binding was more readily blocked by 4a. These studies suggest that [¹⁸F]4c should be a useful compound to assess dynamic changes in serotonin levels while [¹⁸F]4d, with its high contrast and F-18 label, should provide better statistics and quantification for static measurement of 5-HT(1A) receptor distribution.

Full text of DOI:10.1021/jm980456f

1576
J . Med. Chem. 1999, 42, 1576-1586
Develop m en t of F lu or in e-18-La beled 5-HT1A An ta gon ists
Lixin Lang,* Elaine J agoda, Bernard Schmall, Bik-Kee Vuong, H. Richard Adams, David L. Nelson,†
Richard E. Carson, and William C. Eckelman
Positron Emission Tomography Department, Clinical Center, National Institutes of Health, 10 Center Drive,
Bethesda, Maryland 20892, and Eli Lilly & Company, Indianapolis, Indiana 46285
Received August 5, 1998
We have synthesized five fluorinated derivatives of WAY 100635, N-{2-[4-(2-methoxyphenyl)-
piperazino]ethyl}-N-(2-pyridyl)cyclohexanecarboxamide (4a ), using various acids in place of
the cyclohexanecarboxylic acid (CHCA, 2a ) in the reaction scheme. The five acids are
4-fluorobenzoic acid (FB, 2b), 4-fluoro-3-methylbenzoic acid (MeFB, 2c), trans-4-fluorocyclo-
hexanecarboxylic acid (FC, 2d ), 4-(fluoromethyl)benzoic acid (FMeB, 2e), and 3-nitro-4-
(fluoromethyl)benzoic acid (NFMeB, 2f) (see Scheme 1). These compounds were radiolabeled
with fluorine-18, and their biological properties were evaluated in rats and compared with
those of [11C]carbonyl WAY 100635 ([carbonyl-11C]4a ). [Carbonyl-11C]4a cleared the brain with
a biological half-life averaging 41 min. The metabolite-corrected blood radioactivity had a half-
life of 29 min. [18F]FCWAY ([18F]4d ) gave half-lives and intercepts comparable to [carbonyl-
11C]4a in the brain, but the blood clearance was faster. [18F]FBWAY ([18F]4b) showed an early
rapid net efflux from the whole brain, clearing with a biological half-life of 35 min. The
metabolite-corrected blood half-life was 41 min. The comparable whole brain and blood half-
lives for Me[18F]FBWAY ([18F]4c) were 16 and 18 min, respectively. For each compound, the
corresponding carboxylic acid was identified as a major metabolite in blood. Fluoride was also
found after injection of [18F]4d . However, for all compounds there was a good correlation (R >
0.97) between the differential uptake ratio (DUR, (%ID/g) × body weight (g)/100) in individual
rat brain regions at 30 min after injection and the concentration of receptors as determined by
in vitro quantitative autoradiography in rat. Specific binding ratios [region of interest (ROI)/
cerebellum-1] in control studies for cortex (Ctx) and hippocampus (H) were higher for [carbonyl-
11C]4a and [18F]4d compared to [18F]4b and [18F]4c. [18F]4d has similar pharmacokinetic
properties and comparable specific binding ratios to [carbonyl-11C]4a . Fifty nanomoles of 4a
blocked only 30% of the specific binding of [18F]4d , while complete blockade was obtained from
co-injection of 200 nmol of 4a (H/Cb-1 from 17.2 to 0.6). [18F]4b and [18F]4c showed lower specific
binding ratios than [carbonyl-11C]4a and [18F]4d . [18F]4c was superior to [18F]4b since its specific
binding was more readily blocked by 4a . These studies suggest that [18F]4c should be a useful
compound to assess dynamic changes in serotonin levels while [18F]4d , with its high contrast
and F-18 label, should provide better statistics and quantification for static measurement of
5-HT1A receptor distribution.
In tr od u ction
erties and often high affinity to other biogenic amine
receptor systems. Clearly, a highly selective antagonist
is needed to characterize this receptor subtype. The
term “silent antagonist” has been introduced to distin-
guish true 5-HT1A receptor antagonists from those that
may also be partial agonists. In addition, most of these
ligands showed relatively weak binding to the 5-HT1A
receptor and cannot be used for studies using radioli-
gands in vivo with positron emission tomography (PET).
Serotonin receptors, specifically the 5-HT1A subtype,
have been implicated in anxiety, dementia, schizophre-
nia, and depression.1 It has been difficult to fully isolate
the effects of the 5-HT1A system because of the large
number of 5-HT receptor subtypes2 and the difficulty
in defining the agonist and antagonist properties of
potential subtype-selective ligands using physiological
models.1 Studies with the 5-HT1A receptor agonist 8-OH-
DPAT [8-hydroxy-2-(di-n-propylamino)tetralin] showed
it to be selective for this subtype and defined many
functional responses that are believed to be a result of
5-HT1A receptor activation. However, many of the
antagonists used to date (spiperone, propanolol, and
pindolol) are nonselective, and some of the ligands that
were initially characterized as antagonists (BMY7378,
spiroxatrine, NAN190, UH301, SDZ216525, S14063)
were subsequently shown to have partial agonist prop-
In 1992, Crouzel et al. reviewed ligands and tracers
for PET studies of the 5-HT system.3 The first selective
5-HT1A receptor silent antagonist, WAY 100135 (N-tert-
butyl-3-[4-(2-methoxyphenyl)-1-piperazinyl]-2-phenyl-
propanamide), has an IC50 of 15 nM.4 The first high-
affinity, subtype-selective, silent antagonist appears to
be WAY 100635 (4a ).5 The carboxamide derivative was
radiolabeled with 123I by Kung et al. and shown to have
appropriate properties for in vivo imaging.6 Laporte et
al. showed that the tritiated 4a had the appropriate
binding characteristics in the mouse brain,7 whereas
Hume et al. showed that the same compound had the
* Corresponding author: Bldg 10, Rm 1C497, NIH, 10 Center Dr,
MSC1180, Bethesda, MD 20892.
† Eli Lilly & Co.
10.1021/jm980456f CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/17/1999

Fluorine-18-Labeled 5-HT1A Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9 1577
appropriate distribution in the rat brain.8 In the latter
paper, binding potentials were derived and correlated
with the receptor concentration as determined using in
vitro studies. In addition, Mathis et al. studied [O-meth-
yl-11C]4a in rat and rhesus monkeys.9 In the monkey,
C-11 radioactivity increased in the frontal cortex up to
20 min and then slowly cleared. The cortex-to-cerebel-
lum ratio reached a maximum of 5.5. However, Osman
et al. showed that a radioactive metabolite of [O-methyl-
11C]4a crossed the blood-brain barrier but did not bind
to the 5-HT1A receptor.10,11 This radioactive metabolite
was identified as [O-methyl-11C]WAY 100634, N-{2-[4-
(2-methoxyphenyl)piperazino]ethyl}-2-pyridinamine (1).
As a result, further labeling attempts concentrated on
incorporating the C-11 or F-18 in the carboxamide
moiety. The first approach was to radiolabel 4a at the
carbonyl carbon of the cyclohexanecarboxamide.12 This
compound showed reduced nonspecific binding and
higher target-to-nontarget ratios and gave only radio-
labeled carboxylic acid metabolites. Preliminary PET
experiments in our laboratory showed that transient
equilibrium was not reached in tissues with the highest
specific binding using [carbonyl-11C]4a in rhesus mon-
keys. Furthermore, receptor parameter quantification
was sensitive to accurate plasma metabolite correction,
leading to the use of a reference tissue method.13
Because of its longer half-life, we therefore investigated
the possibility of using a F-18-radiolabeled silent an-
tagonist with comparable or slightly weaker affinity for
the 5-HT1A receptor. We concentrated on labeling the
analogues of 4a at the carboxamide moiety in order to
minimize metabolites that cross the blood-brain bar-
rier.
evoke any component of the 5-HT syndrome or induced
hypothermia elicited by agonists. Forster et al. con-
cluded from these experiments that 4a was a silent
antagonist, a term which distinguishes true antagonist
from partial agonist.
As a result of these preliminary data, we have
synthesized four additional derivatives of 4a using
various acids to replace the cyclohexanecarboxylic acid
moiety in the reaction scheme. The four acids are
4-fluoro-3-methylbenzoic acid (MeFB, 2c), 4-(fluorom-
ethyl)benzoic acid (FMeB, 2e), 3-nitro-4-(fluoromethyl)-
benzoic acid (NFMeB, 2f), and trans-4-fluorocyclohex-
anecarboxylic acid (FC, 2d) (Scheme 1).18 The comparable
N-{2-[4-(2-methoxyphenyl)piperazino]ethyl}-N-(2-pyridyl)-
cyclohexanecarboxamide is abbreviated by adding WAY
to the abbreviation: e.g., the amide resulting from the
reaction of 4-fluoro-3-methylbenzoic acid is abbreviated
as MeFBWAY.) For comparison, [carbonyl-11C]4a was
also synthesized by radiolabeling in the carbonyl posi-
tion with C-11 using the procedure of Pike et al.19 We
first determined the metabolite profile in rat brain and
blood and then studied saturable binding for those
compounds shown to have minimal concentration of
metabolites in the rat brain.
Resu lts a n d Discu ssion
Ch em istr y. Reference compounds were prepared for
the five fluorinated analogues according Scheme 1.
Compounds 4a -4e were prepared by reacting 1 with
the corresponding acid chloride 3a -3e in methylene
chloride. Each compound was purified and analyzed by
elemental analysis, proton NMR, and MS and found to
be a single compound of the proposed formula. Among
these five compounds, compound 4d was particularly
difficult to make due to the low yield of 4-fluorocyclo-
hexanecarboxylic acid (2d ) with a six-step reaction
sequence. The cyclohexene derivative is the major
product during the fluorination process with <10% of
the products as the fluoro compound.
Compounds 6a -6c were prepared by coupling glycine
sodium salt with the corresponding acid chloride as the
reference compounds of possible metabolites. Five sub-
strates (5b-5f) for radiolabeling were prepared accord-
ing to Scheme 2. The acids were protected with pen-
tamethylbenzyl ester from either pentamethylbenzyl
chloride or pentamethylbenzyl alcohol. This protecting
group can be easily removed with a small amount of
trifluoroacetic acid during the radiolabeling process.
Compounds 5e and 5f prepared from pentamethylbenzyl
alcohol and 1,3-dicyclohexylcarbodiimide had relatively
low yields compared to compounds 7 and 8 due to the
formation of anhydrides. Compounds 5b and 5c are two
very stable triflate salts prepared from 7 and 8 with
trifluoromethanesulfonate. Extra care is required for
preparing these two substrates since a trace amount of
acid can cause the hydrolysis of the protecting group.
Compound 5d was prepared from 9 with 4-nitrobenze-
nesulfonyl chloride. Mesylate and tosylate derivatives
were also prepared from 9, but the radiolabeling ef-
ficiency for mesylate and tosylate was much lower than
that of 5d (<5%). Compound 9 was prepared starting
from ethyl 4-hydroxycyclohexanecarboxylate which is a
mixture of cis and trans isomers. The ethyl ester (cis
and trans mixture) was first hydrolyzed to the acid with
Zhuang et al. had shown that various benzoyl sub-
stituents affected the Ki of the compound as measured
in rat hippocampal homogenates.14 WAY 100635 (4a )
had a Ki of 0.84 nM, whereas the p-fluorobenzoyl
analogue had a Ki of 3.3 nM and the p-iodobenzoyl
analogue had a Ki of 2.6 nM. We have previously
reported that 4a and the 4-fluorobenzoyl analogue
(FBWAY, 4b) in which the cyclohexane carbonyl was
replaced by the 4-fluorobenzoyl moiety have high affin-
ity for the 5-HT1A receptor.15 We found using an in vitro
assay (NovaScreen) that 4a had a Ki of 0.98 nM for the
5-HT1A receptor, 578 nM for the 5HT2 receptor, and 86.8
nM for the dopamine D2 receptor, which was the highest
binding non-5-HT biogenic amine receptor. The 4b had
a Ki of 1.2 nM for the 5-HT1A receptor, 130 nM for the
5-HT2 receptor, and 151 nM for the R1-adrenoceptor,
which was the highest binding non-5-HT biogenic amine
receptor for this compound. In addition, Thielen and
Frazer concluded that 4b was a silent antagonist in at
least two functional measures of postsynaptic 5-HT1A
receptor activation.16 Recently, Forster et al. have
carried out a complete pharmacological profile of 4a .17
In a radioligand binding assay, 4a had a high affinity
(pIC50 ) 8.87) with good specificity, with R1-adrenocep-
tor binding being the strongest at a pIC50 of only 6.64.
They also carried out a number of physiological tests
that showed the antagonism of 4a . In the guinea pig
ileum, 4a antagonized the 5-HT1A receptor-mediated
inhibition of electrically evoked twitch. WAY 100635
(4a ) also inhibited dorsal raphe nucleus 5-HT neuronal
firing induced by 8-OH-DPAT. In addition, 4a did not

1578 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9
Lang et al.
Sch em e 1
base and then converted to pentamethylbenzyl ester (cis
and trans mixture) using pentamethylbenzyl chloride.
There are two distinct proton NMR signals of this
mixture for proton at the 4 position of the cyclohexane
ring with chemical shifts at 3.6 and 3.9 ppm, respec-
tively.
The mixture was recrystallized using methylene
chloride/hexane to give pure 9 with a chemical shift at
3.9 ppm for the proton at the 4 position. Compound 9
can be converted to lactone 8 indicating that this
compound is a cis isomer. Compound 10 was prepared
from 9 using diethylaminosulfur trifluoride with inver-
sion of configuration to give the trans isomer.
Ra d ioch em istr y. Radiolabeling was carried out us-
ing F-18 produced by the 18O(p,n) reaction. The two
fluoromethylbenzoic acid derivatives were radiolabeled
using the bromide precursor (5e and 5f),20 and the two
fluorobenzoic acid derivatives were radiolabeled using
the trimethylammonium precursor (5b and 5c) following
standard procedures (Scheme 3).21 [18F]2d was prepared
using the 4-cis-nosylate derivative of the pentamethyl-
benzyl ester of cyclohexanecarboxylic acid (5d ). After
radiolabeling the benzoic acid or the carboxylic acid with
F-18, the acid chloride was prepared and reacted with
the amine intermediate, WAY 100634 (1). Shiue et al.22
previously radiolabeled the nitro precursor of 4b, but
we obtained low radiochemical yields using their one-
step method to make [18F]4c. All radiolabeled products
were purified by HPLC with radiochemical yields rang-
ing 10-20% and total synthesis time of 80-90 min. The
specific activity, determined by using on-line measure-
ments of radioactivity and UV absorption, was >1000
Ci/mmol (EOB) for the F-18 compounds and >250 Ci/
mmol (EOB) for the C-11 compounds. The retention
times in minutes for the compounds on the same
reversed-phase HPLC system with 65% methanol and
0.1 M ammonium formate (4a ) 30, 4c ) 29, 4b ) 19,
4d ) 18, and 4e ) 16) give an indication of their relative
lipophilicity. In this chromatography system 4d appears
to be more polar than 4a and 4c.
In Vitr o Assa ys. Three compounds (4a , 4c, and 4d )
were analyzed by various methods to determine the
inhibition constant (Ki). The Ki was determined by the
inhibition of G-protein activation for all three com-
pounds (Table 1, column 1). In addition, compounds 4a
and 4c were studied in duplicate in separate studies.
For all studies, the affinity of 4a was similar to that
obtained by 4d , whereas 4c had a weaker affinity.
Compound 4c also showed R1 binding with a Ki of 13.2
( 3.7 nM.
Similar rank order was obtained when competitive
binding assays were used (Table 1, columns 2, 3). The
MDS Panlabs assays agreed with the previous data
obtained by other methods. The data from the competi-
tive assay (column 2) was carried out using cloned cells,
as was the competitive assay (column 3). In general,
compound 4d has a stronger binding affinity than
compound 4c by a factor of 3-8. FCWAY (4d ) had an
affinity similar to that obtained for WAY100635.
Meta bolites. The chemical form of the radioactive
compound in brain and blood was determined by solvent
extraction and TLC. In these metabolite studies of brain
and blood in rat, we proceeded with testing of an
analogue if: (1) the ex vivo extraction efficiency did not
decrease with time and on average g85% of the radio-
activity present in the brain was present in the aceto-
nitrile:water solution and (2) >95% of the extracted
radioactivity in the brain was present as the injected
compound. The percentage of radioactivity for both [18F]-
FMeBWAY ([18F]4e) and N[18F]FMeBWAY ([18F]4f) in
acetonitrile from brain tissue was less than 77% at 5

Fluorine-18-Labeled 5-HT1A Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9 1579
Sch em e 2
min and decreased with time (<44% at 30 min). The
mechanism for this retention has not been elucidated
although low extraction of fluoride is most likely.
Therefore, further studies were not pursued. For all five
radiolabeled compounds, the chemical form of the
extracted material in whole brain was >95% parent
compound from 5 to 30 min. Because of the small
percentage difference, no correction was made on the
reported DUR values for brain (Figure 1).
The radioactivity in the blood was extracted at high
efficiencies for Me[18F]FBWAY ([18F]4c) (83-90%), [18F]-
FBWAY ([18F]4b) (84-92%), and [11C]4a (90-97%),
whereas for [18F]FCWAY ([18F]4d ) the extraction de-
creased from 65% to 33% from 5 to 30 min, most likely
due to the percent fluoride present.
corrected for extraction efficiency assuming that the
radioactivity remaining in the pellet was in the form of
fluoride (Figure 2b). Fluoride is extracted with aceto-
nitrile at an efficiency of 18 ( 5% in blood between 5
and 30 min after injection of pure F-18 fluoride, whereas
the extraction from brain tissue is 5 ( 0.7% under the
same conditions (n ) 8). Since fluoride does not cross
the blood-brain barrier to an appreciable extent, the
fluoride observed in brain tissue in this experiment is
most likely due to intravascular radioactivity.
Biod istr ibu tion . The radioactivity in brain and
blood samples is presented in terms of the DUR [(%ID/
g) × body weight (g)/100] in order to normalize for the
differences in rat weight (Figures 1 and 2). The corre-
sponding carboxylic acid and other more polar metabo-
lites were found in blood in all studies. For 4a (Figure
1a) the radioactivity in whole brain was maximal at the
first time point (5 min) and decreased over time with a
biological half-life averaging 41 min. The parent com-
pound had a half-life of 29 min in blood (Figure 1a). The
ratio of the brain to metabolite-corrected blood (∼4.2)
was lower than previously observed7 most likely because
of the relatively high mass injected (see Experimental
For Me[18F]FB ([18F]2c), a metabolite of Me[18F]-
FBWAY ([18F]4c), the extraction from brain tissue was
between 72% and 80%, whereas the blood extraction
efficiency was between 80% and 85%. For the analogous
metabolite of [18F]FCWAY ([18F]4d ), [18F]FC ([18F]2d ),
the extraction from brain and blood was below 85%,
presumably because of the rapid defluorination of this
metabolite. The values for both brain and blood were

1580 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9
Lang et al.
Sch em e 3
Ta ble 1. Inhibition Constants (Ki, nM) Obtained by in Vitro
Assay
being substantially lower than the blood concentration.
For FCWAY (4d ), the major metabolite, trans-4-fluoro-
cyclohexanecarboxylic acid (2d ) and its metabolites,
most likely the glycine conjugates and fluoride, also
showed minimal uptake in the brain (compared to
FCWAY, 4d ) and cleared with a similar half-life from
blood and brain (Figure 2b). Note that, unlike Me[18F]-
FB ([18F]2c), the brain-to-blood ratio for [18F]FC ([18F]-
2d ) is approximately 1.0, demonstrating uptake into
brain tissue. The pKa of the carboxylic acids is in the
range of 4.2-4.9; thus at physiological pH, the concen-
tration ratio of charged to uncharged molecules is
approximately 1000:1. However, the uncharged mol-
ecules presumably cross the blood-brain barrier, even-
tually resulting in equilibrium between brain and
blood.23 We have previously reported brain localization
in monkey studies using [11C]CHCA ([11C]2a ).24
5-HT
GTP-γ-S
Ki vs
MDS Panlabs
assay
compound
[3H]8-OH-DPAT
WAY 100635 (4a ) 1.09 (n ) 2) 0.59 (n ) 2)
0.34 ( 0.06
MeFBWAY (4c)
4.02 (n ) 2) 3.96 (n ) 2)
2.9 ( 0.35 1.14 ( 0.36
1.04 ( 0.08 0.515 ( 0.047
0.791 ( 0.113
0.247 ( 0.085
FCWAY (4d )
Section). [18F]FBWAY ([18F]4b) had an early rapid net
efflux (Figure 1b). The second component of the whole
brain clearance has a biological half-life of 35 min with
high uptake and a blood clearance of 41 min. The
comparable whole brain and blood half-lives for Me[18F]-
FBWAY ([18F]4c) were shorter, i.e., 16 and 18 min,
respectively (Figure 1c). The [18F]FCWAY ([18F]4d ) gave
a brain half-life comparable to [11C]4a (Figure 1d). The
blood half-time was faster (16 min).
The biodistribution of the radiofluorinated metabo-
lites was also studied in the same paradigm using TLC
conditions that separated the benzoic acid from metabo-
lites (most likely the glycine conjugates) to determine
the pharmacokinetics and the blood-brain barrier
penetration of these metabolites. The fluorobenzoic acid
metabolites of Me[18F]FBWAY ([18F]4c) and [18F]FB-
WAY ([18F]4d ), i.e., Me[18F]FB (Figure 2a) and [18F]FB
(data not shown), respectively, both cleared rapidly from
the brain and the blood with the brain concentration
Recently, there have been other analogues of 4a
prepared for use as radiolabeled silent antagonists.
Zhuang et al.25 prepared cyclic amides in an attempt to
alter the metabolism, but this led to non-5-HT1A receptor
binding in vivo. Wilson et al.,26 on the basis of metabo-
lism studies using a liver cytosol and microsomal
preparation of human liver, chose the [2.2.2]bicyclooc-
tane-1-carboxylic acid as the compound with the least
metabolism to the carboxylic acid. However, there was
extensive metabolism to an unidentified compound that
was identical whether the compound was labeled in the

Fluorine-18-Labeled 5-HT1A Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9 1581
F igu r e 1. (a) Distribution of [11C]4a in rat brain (9) and blood (O) expressed as DUR [(%ID/g) × body weight (g)/100]. The
equation for the metabolite-corrected blood is 0.40 e(-0.024*time), and the equation for the brain is 1.41 e(-0.017*time); analyzed with
TLC system A (n ) 2). (b) Distribution of [18F]4b in rat brain (9) and blood (O) expressed as DUR. The equation for the metabolite-
corrected blood is 0.15 e(-0.017*time), and the equation for the brain is 0.69 e(-0.020*time). This was calculated using the points at 15,
30, and 60 min; analyzed with TLC system A (n ) 4). (c) Distribution of [18F]4c in rat brain (9) and blood (O) expressed as DUR
and corrected for metabolites. The equation for the metabolite-corrected blood is 0.31 e(-0.039*time), and the equation for the brain
is 1.26 e(-0.044*time); analyzed by TLC using 70% propanol, 30% ammonia as solvent (n ) 4). (d) Distribution of [18F]4d in rat brain
(9) and blood (O) expressed as DUR. The equation for the metabolite-corrected blood is 0.25 e(-0.041*time), and the equation for the
brain is 1.81 e(-0.019*time). The blood data were also corrected for extraction efficiency; analyzed with TLC system A (n ) 4).
carboxylic acid moiety or at the O-methyl position. The
compound appeared to be more lipophilic than the
parent. Pike et al. have labeled the desmethyl form of
4a and found it to have identical binding properties to
that of 4a . However, they did not find the carboxylic
acid metabolite.27 It is not clear if any of these com-
pounds have a superior metabolite profile, i.e., metabo-
lites that are more polar than the parent compound and
do not cross the blood-brain barrier.
will eventually match that in blood, and a constant
tissue-to-blood concentration ratio will be achieved, a
condition termed transient equilibrium. The time to
achieve this state depends on the transport and binding
kinetics of the tracer. In general, higher-affinity com-
pounds have a slower dissociation rate and will require
a longer period of time to achieve transient equilibrium.
This relationship can be seen in these data. For the
lower affinity compounds, [18F]FBWAY ([18F]4b) and
Me[18F]FBWAY ([18F]4c), the tissue and blood clearance
For receptor-binding ligands that bind reversibly,
following a bolus injection, the clearance rate in tissue

1582 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9
Lang et al.
F igu r e 3. Comparison of the %ID/g in various tissues from
rat brain at 30 min after injection with 5-HT1A receptor
concentration obtained using quantitative autoradiography in
vitro. The rank order of receptor concentration from in vitro
studies was hippocampus > frontal cortex > hypothalamus >
cerebellum ) striatum, and a similar rank order was found
in the studies in vivo. The concentration of receptors was taken
from Khawaja27 who determined the receptor concentration
from in vitro quantitative autoradiography in rats.
min after injection in rat and the concentration of
receptors as determined by in vitro quantitative auto-
radiography in rats (Figure 3).28 The rank order of
receptor concentration from in vitro studies was hip-
pocampus > frontal cortex > hypothalamus > cerebel-
lum ) striatum, and a similar rank order was found in
the studies in vivo. The concentration of receptors was
taken from Khawaja27 who determined the receptor
concentration from in vitro quantitative autoradiogra-
phy in rats.
In general, the two cyclohexanecarboxamide ana-
logues ([11C]4a and [18F]4d ) had higher brain-to-blood
ratios than the two phenylcarboxamide analogues ([18F]-
4b and [18F]4c) reflecting the in vitro binding affinities
and specific binding ratios. Blocking experiments with
co-injected 50 nmol of 4a , reported as %ID/g in the
target organ and as a ratio of %ID/g in the ROI minus
%ID/g in the cerebellum (Cb) divided by the %ID/g in
the Cb (ROI/Cb-1), showed >80% reduction of specific
binding in the hippocampus for Me[18F]FBWAY ([18F]-
4c), whereas [18F]FBWAY ([18F]4b) showed only ∼60%
reduction of specific binding in the hippocampus. The
reason for this is not clear given the similar pharma-
cokinetics, but because of this, Me[18F]FBWAY ([18F]-
4c) was pursued for evaluation in nonhuman primates.
Other receptors seem not to be involved in brain
retention of Me[18F]FBWAY ([18F]4c). [18F]4c had an
affinity of 13 nM for the R1-adrenoceptor, but co-
injection of 200 nmol of prazosin did not decrease the
uptake of MeFBWAY (4c) in brain. The uptake of
prazosin in brain is relatively low (0.06% ID/g) but
would still represent a competing dose of ∼120 nM in
rat brain.29 MeFBWAY (4c) bound to other biogenic
amine receptors with affinities > 100 nM.
F igu r e 2. (a) Distribution of [18F]2c in rat brain (9) and blood
(O) expressed as DUR. The equation for the metabolite-
corrected blood is 0.93 e(-0.175*time), and the equation for the
metabolite-corrected brain is 0.24 e(-0.208*time); analyzed with
TLC system B (n ) 4). (b) Distribution of [18F]2d in rat brain
(9) and blood (O) expressed as DUR and corrected for metabo-
lites. The equation for the metabolite-corrected blood is 0.87
e(-0.156*time), and the equation for the metabolite-corrected brain
is 0.38 e(-0.108*time); analyzed with TLC system A (n ) 4). Both
brain and blood values were corrected for extraction efficiency
assuming that the radioactivity remaining in the pellet was
in the form of fluoride.
rates are well-matched (Figure 1b,c) and transient
equilibrium was achieved.
However, for the compounds with higher affinity, [11C]-
4a and [18F]FCWAY ([18F]4d ), the brain clearance rate
was slower than that in blood suggesting that transient
equilibrium had not been achieved due to slower equi-
librium between bound and free ligand in tissue. For
both types of compounds there was a good correlation
between the DUR in individual parts of the brain at 30
When blocked with 50 nmol of 4a , [18F]FCWAY ([18F]-
4d ) showed a decrease in the hippocampus tissue to 54%
of the value obtained for the no-carrier-added study
(Table 2). The high Ctx/Cb-1 and H/Cb-1 ratios for [18F]-
FCWAY ([18F]4d ) in the presence of 50 nmol of 4a are
inconsistent with their similar pharmacokinetics (Fig-

Fluorine-18-Labeled 5-HT1A Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9 1583
Ta ble 2. Biodistribution (%ID/g) of Radiolabeled WAY Derivatives at 30 min in Rats with and without Co-injected WAY 100635
(n g 4)
derivative
18F]4b
+50 nmol of 4a
18F]4c
+50 nmol of 4a
11C]4a
+50 nmol of 4a
18F]4d
cortex
hippocampus
hypothalamus
cerebellum
Ctx/Cb-1
H/C6-1
[
0.085 ( 0.010
0.048 ( 0.001
0.047 ( 0.007
0.018 ( 0.007
0.426 ( 0.022
0.044 ( 0.001
0.823 ( 0.028
0.267 ( 0.018
0.069 ( 0.064
0.195 ( 0.049
0.074 ( 0.007
0.118 ( 0.016
0.017 ( 0.004
0.760 ( 0.118
0.132 ( 0.036
1.298 ( 0.068
0.701 ( 0.053
0.085 ( 0.006
0.066 ( 0.011
0.016 ( 0.041
ND
0.034 ( 0.003
0.024 ( 0.002
0.020 ( 0.004
0.016 ( 0.003
0.045 ( 0.011
0.045 ( 0.004
0.064 ( 0.006
0.048 ( 0.004
0.054 ( 0.006
1.5
1.0
1.3
0.1
8.5
0.0
11.0
4.6
0.3
4.7
2.1
4.9
0.1
15.9
1.9
19.3
13.6
0.6
[
ND
[
0.224 ( 0.037
0.036 ( 0.012
0.499 ( 0.086
0.175 ( 0.024
0.045 ( 0.009
[
+50 nmol of 4a
+200 nmol of 4a
7.0 (m, 4H), 7.2-7.3 (m, 2H),7.74 (td, 1H), 8.51 (dd, 1H); MS
ure 1a,d). However, co-injection of 200 nmol of 4a with
[18F]FCWAY ([18F]4d ), decreased the brain radioactivity
to background levels. These data are a single measure-
ment at 30 min and therefore will be dependent on the
relative pharmacokinetics of the radiolabeled compound
and 50 nmol of 4a . In addition, 4a has a higher affinity
than either FBWAY (4b) and MeFBWAY (4c) and
therefore should increase the net efflux of these two
radiolabeled compounds compared to its effect on [11C]-
4a and [18F]FCWAY ([18F]4d ).
(CI-NH3) m/z 203 (M + NH4)+, 186 (M + H)+. Anal. (C9H15
-
NO3) Calcd: C, 58.38; H, 8.11; N, 7.57. Found: C, 58.67; H,
8.20; N, 7.57.
4-F lu or o-N-{2-[4-(2-m eth oxyp h en yl)p ip er a zin o]eth yl}-
N-(2-p yr id in yl)ben za m id e (F BWAY, 4b). This compound
was prepared from 4-fluorobenzoyl chloride and 1 by the
procedure of Zhuang et al.14 in 55% yield: mp 260 °C (HCl
salt);1H NMR (CDCl3) δ 2.80 (overlapping signals, 10H), 3.84
(s, 3H), 4.30 (t, 2H), 6.90 (m, 8H), 7.39 (m, 3H), 8.44 (dd, 1H);
MS (EI) m/z 434 (M). Anal. (C25H27N14O2F‚2HCl‚5H2O) Cal-
cd: C, 58.14; H, 5.81; N, 10.85; F, 3.68; Cl, 13.76. Found: C,
58.03; H, 5.95; N, 10.24; F, 3.09; Cl, 12.85.
Con clu sion
4-F lu or o-N-{2-[4-(2-m eth oxyp h en yl)p ip er a zin o]eth yl}-
3-m eth yl-N-(2-p yr id in yl)ben za m id e (MeF BWAY, 4c). To
a solution of 1 (2.0 g, 6.4 mmol) in 20 mL of methylene chloride
were added 1 mL of triethylamine (7.2 mmol) and 1.1 g (6.4
mmol) of 4-fluoro-3-methylbenzoyl chloride (Buu-Hoi and
J acquignon)30 in 10 mL of methylene chloride, and the mixture
was refluxed for 24 h. After the mixture cooled, 100 mL of
methylene chloride was added and the solution was placed in
a separatory funnel. The solution was treated two times with
100 mL of 1 N NaHCO3 solution and two times with 100 mL
of 10% aqueous NaCl. The organic phase was dried with Na2-
SO4. Evaporation of the solvent yielded 4c as a viscous oil in
92% yield: 1H NMR (CDCl3) δ 2.18 (s, 3H), 2.79 (overlapping
signals, 10H), 3.85 (s, 3H), 4.29 (t, 2H), 6.91 (m, 8H), 7.37 (m,
2H), 8.23 (dd, 1H); MS (EI) m/z 448 (M). Anal. (HCl salt)
(C26H29N4O2F‚3HCl‚2H2O) Calcd: C, 52.57; H, 6.07; N, 9.44;
F, 3.20; Cl, 17.94. Found: C, 52.81; H, 6.31; N, 10.03; F, 2.66;
Cl, 17.86.
4-(F lu or om eth yl)-N-{2-[4-(2-m eth oxyp h en yl)p ip er a zi-
n o]et h yl}-3-m et h yl-N-(2-p yr id in yl)b en za m id e (F MeB-
WAY, 4e). This compound was prepared from 4-(fluoromethyl)-
benzoyl chloride and 1 in 60% yield as an oil. The benzoyl
chloride was synthesized from thionyl chloride and 4-(fluo-
romethyl)benzoic acid:19 1H NMR (CDCl3) δ 2.80 (overlapping
signals, 10H), 3.85 (s, 3H), 4.43 (t, 2H), 5.32 (d, 2H), 7.34 (m.
5H), 8.44 (dd, 1H); MS (EI) m/z 448 (M). Anal. (C26H29N4O2F‚
2H2O) Calcd: C, 64.46; H, 6.82; N, 11.57; F, 3.93. Found: C,
64.10; H, 6.17; N, 11.01; F, 3.75.
We have prepared five F-18-labeled potential 5-HT1A
receptor binding ligands. Despite the success in using
[18F]fluoromethylbenzoyl derivatives in other applica-
tions in vivo, two of these derivatives ([18F]4e and [18F]-
4f) were not readily extracted from either plasma or
brain and were abandoned. Of the remaining three, Me-
[18F]FBWAY ([18F]4c) appears to have an advantage
over [18F]FBWAY ([18F]4b) because of its higher specific
binding in rat as shown by competition studies with 4a .
In vitro test results were consistent with a high-affinity
antagonist for the 5-HT1A receptor. [18F]FCWAY ([18F]-
4d ) appears to have the same pharmacokinetic proper-
ties as [11C]4a . With the longer half-life of F-18, there
is the possibility that [18F]4d should provide improved
imaging qualities and quantification accuracy for 5-HT1A
receptor over [11C]4a .
Exp er im en ta l Section
Gen er a l. Compound 1 was obtained from Med-Life Sys-
tems, Inc. Upper Darby, PA. All other chemicals were pur-
chased from Aldrich Chemical Co., Milwaukee, WI. NMR
spectra were obtained on a Varian Gemini-2000 (200-MHz)
instrument with tetramethylsilane as the internal standard.
Mass spectra were recorded on a Hewlett-Packard 5989B mass
spectrometer linked to a Hewlett-Packard 5890 series II plus
gas chromatograph. Elemental analyses were performed by
Galbraith Laboratories (Knoxville, TN). Thin-layer chroma-
tography of the radioactive products was performed on What-
man LK6DF silica gel glass-backed plates (5 × 20 cm, 250 µm).
TLC radiochromatograms were obtained using a Bioscan
System 200 imaging scanner and a Fuji Bio-imaging Analysis
System 1500. Semipreparative HPLC was carried out using a
Perkin-Elmer series 200 LC pump. UV absorbance was
monitored with a Waters 486 UV detector, and radioactivity
was monitored using a Beckman model 170 radioisotope
detector. The purification of radiolabeled compounds was
performed using a reversed-phase semipreparative Rainin
Microsorb 5-µm (10 × 250 mm) C18 column.
Compound NFMeBWAY (4f) was prepared with a similar
procedure as an oil.
Cycloh exa n eca r bon ylglycin e (6a ). This solid compound
was prepared by the procedure used to synthesize 4-fluoroben-
zoylglycine utilizing cyclohexanecarbonyl chloride: 1H NMR
(200 MHz, CD3CN) δ 1.31 (m, 5H), 1.69 (m, 5H), 2.16 (m, 1H).
3.83 (d, 2H), 6.00 (s, 1H); MS (CI-NH3) m/z 203 (M + NH4)+,
186 (M + H)+. Anal. (C9H15NO3) Calcd: C, 58.38; H, 8.11; N,
7.57. Found: C, 58.67; H, 8.20; N, 7.57.
4-F lu or oben zoylglycin e (6b). This compound was pre-
pared by reacting sodium glycinate in aqueous NaOH with
4-fluorobenzoyl chloride. The 4-fluorobenzoylglycinate was
neutralized with HCl. After filtration, a white solid was
obtained that was dissolved in acetonitrile. The acetonitrile
was evaporated, and chloroform was added to the residue. The
mixture was placed onto a silica gel column and eluted with
chloroform, which removed 4-fluorobenzoic acid. Then elution
with acetonitrile gave 4-fluorobenzoylglycine as a white solid
(37%): 1H NMR (CD3CN) δ 4.05 (d, 2H), 7.22 (m, 2H), 7.38 (s,
N-{2-[4-(2-Meth oxyp h en yl)p ip er a zin o]eth yl}-N-(2-p y-
r id yl)cycloh exa n eca r boxa m id e] (WAY 100635, 4a ). This
compound was prepared from cyclohexanecarbonyl chloride
and WAY 100634 (1) by the procedure of Pike et al.:18 1H NMR
(CDCl3) δ 0.9-1.3 (m, 4H), 1.4-1.8 (m, 6H), 2.2 (t, 1H), 2.5-
2.7 (m, 6H), 3.0 (s, 4H, broad), 3.82 (s, 3H), 3.98 (t, 2H), 6.8-

1584 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9
Lang et al.
1H), 7.87 (m, 2H); MS (CI-NH3) m/z 215 (M + NH4)+, 198 (M
+ H)+. Anal. (C9H8NO3F) Calcd: C, 54.82; H, 4.06; N, 7.11; F,
9.65. Found: C, 54.80; H, 4.12; N, 6.89; F, 9.33.
4-F lu or o-3-m eth ylben zoylglycin e (6c). This solid com-
pound was prepared by the procedure used to synthesize
4-fluorobenzoylglycine utilizing 4-fluoro-3-methylbenzoyl chlo-
ride in 14% yield: 1H NMR (CD3CN) δ 2.31 (s, 3H). 4.05 (d,
2H), 7.14 (t, 1H). 7.35 (s, 1H), 7.70 (m, 2H); MS (CI-NH3) m/z
229 (M + NH4)+, 212 (M + H)+. Anal. (C10H10NO3F) Calcd: C,
56.87; H, 4.74; N, 6.64; F, 9.01. Found: C, 56.69; H, 4.80; N,
6.63; F, 8.66.
droxycyclohexanecarboxylate) in 50 mL of N,N-dimethylfor-
mamide was added pentamethylbenzyl chloride (6.5 g, 33.1
mmol) and triethylamine (3.3 g, 32.6 mmol). The mixture was
stirred at room temperature overnight. The reaction mixture
was poured into a 400-mL 10% sodium bicarbonate solution
to give a white precipitate. The solid product was collected by
filtration and dried under vacuum. The cis and trans mixture
was recrystallized from methylene chloride/hexane to 1.5 g
(14.8%) of pure cis isomer as white crystals: mp 148-149 °C;
1H NMR (CDCl3) δ 1.6-1.8 (m, 6H), 1.8-2.1 (m, 2H), 2.1-2.3
(m, 15H), 2.3-2.5 (m, 1H), 3.85 (m, 1H), 5.12 (s, 2H); MS (EI)
304 (M+), 160, 145.
P en ta m eth ylben zyl 4-(N,N-Dim eth yla m in o)ben zoa te
(7b). To a solution of 4-(N,N-dimethylamino)benzoic acid (0.57
g, 3.4 mmol) in 15 mL of N,N-dimethylformamide were added
pentamethylbenzyl chloride (PMBC) (0.65 g, 3.3 mmol) and
triethylamine (TEA) (0.46 mL, 3.3 mmol). The mixture was
stirred at room temperature overnight. The reaction mixture
was poured into 80 mL of 10% sodium bicarbonate solution to
give white precipitation. The solid product was collected by
filtration and dried under vacuum. The crude product was
recrystallized from methylene chloride/hexane to give white
cis-P en t a m et h ylben zyl 4-[(4-Nit r oben zen esu lfon yl)-
oxy]cycloh exh a n eca r boxyla te (5d ). To a solution of 9 (0.31
g, 0.1 mmol) in 10 mL of methylene chloride were added
4-nitrobenzenesulfonyl chloride (0.25 g, 0.11 mmol) and tri-
ethylamine (0.16 mL, 0.11 mmol). The mixture was stirred at
room temperature overnight. At the end of the reaction, the
solid was filtered and solvent evaporated, and the residue was
purified by flash chromatography on silica gel to give 0.25 g
(50.1%) of white crystals after recrystallization from methylene
1
1
chloride/hexane: mp 142 °C dec; H NMR (CDCl3) δ 1.4-1.6
crystals (0.7 g, 63.6%): mp 154-155 °C; H NMR (CDCl3) δ
(m, 4H), 1.9-2.1 (m, 4H), 2.2-2.3 (m, 15H), 4.55 (m, 1H), 5.22
2.2-2.4 (3s, 15H), 3.01 (s, 6H), 5.42 (s, 2H), 6.60 (d, 2H), 7.90
(d, 2H); MS (EI) 325 (M+), 160, 148.
(s, 2H), 8.1 (d, 2H), 8.4 (d, 2H).
P en t a m et h ylb en zyl 4-(N,N-Dim et h yla m in o)-3-m et h -
ylben zoa te (7c). To a solution of 4-(N,N-dimethylamino)-3-
methylbenzoic acid (0.70 g, 3.9 mmol) in 15 mL of N,N-
dimethylformamide were added pentamethylbenzyl chloride
(0.73 g, 3.7 mmol) and triethylamine (0.52 mL, 3.7 mmol). The
mixture was stirred at room temperature overnight. The
reaction mixture was poured into a 100-mL 10% sodium
bicarbonate solution to give white precipitation. The solid
product was collected by filtration and dried under vacuum.
The crude product was recrystallized from methylene chloride/
hexane to give white crystals (1.0 g, 75.8%): mp 99-101 °C;
1H NMR (CDCl3) δ 2.2-2.4 (3s, 15H), 2.95 (s, 6H), 5.48 (s, 2H),
6.95 (d, 1H), 8.0 (dd, 1H), 8.4 (d, 1H); MS (EI) 339 (M+), 179,
160, 145, 131.
tr a n s-P en t a m et h ylb en zyl 4-F lu or ocycloh exa n eca r -
boxyla te (10). To a solution of 9 (1.8 g, 5.9 mmol) in 100 mL
of methylene chloride was added diethylaminosulfur trifluoride
(DAST) (0.79 mL, 6.0 mmol). The mixture was stirred at room
temperature for 2 h, and the reaction was quenched with
water. The methylene chloride was washed with 50 mL of 1
N HCl twice and dried over anhydrous sodium sulfate. The
solvent was evaporated, and the residue was chromatographed
on silica gel eluting with 1% ethyl acetate in hexane to give
120 mg of product as a white solid (6.7%): mp 110-111 °C;
1H NMR (CDCl3) δ 1.4-1.7 (m, 4H), 1.8-2.2 (m, 4H), 2.2-2.4
(m, 15H), 4.3-4.7 (md, 1H, fluorine coupling), 5.25 (s, 2H);
MS (EI) 306 (M+), 160, 145.
tr a n s-4-F lu or ocycloh exa n eca r coxylic Acid (F C, 2d ).
Compound 10 (120 mg, 0.39 mmol) was dissolved in 0.5 mL of
TFA and was then diluted with 20 mL of water after 2 min.
The white solid formed during dilution was filtered and
aqueous layer extracted with 5 × 20 mL of methylene chloride.
The organic solvent was combined and dried over anhydrous
sodium sulfate. The solvent was evaporated to give 44 mg
(77%) of product as a semisolid: 1H NMR (CDCl3) δ 1.4-1.7
(m, 4H), 1.9-2.3 (m, 4H), 2.3-2.5 (m, 1H), 4.4-4.7 (md,1H,
fluorine coupling), 10.9 (s, 1H); MS (EI) 146 (M+) 126, 108,
97.
tr a n s-4-F lu or o-N-{2-[4-(2-m eth oxyp h en yl)p ip er a zin o]-
eth yl}-N-(2-p yr id yl)cycloh exa n eca r boxa m id e (F CWAY,
4d ). To a solution of 2d (44 mg) in 5 mL of 1,2-dichloroethane
was added 0.2 mL R,R-dichloromethyl methyl ether. The
mixture was refluxed for 2 h until all acid was converted to
acid chloride. The solvent and unreacted R,R-dichloromethyl
methyl ether were evaporated under reduced pressure, and
the residue was redissolved in 5 mL of methylene chloride.
The above solution was added to 120 mg of 1 in 5 mL of
methylene chloride containing 50 µL of triethylamine. The
mixture was refluxed for another 2 h. The solvent was
evaporated and residue redissolved in 1 mL of ethyl acetate
and chromatographed on silica gel eluting with ethyl acetate
containing 0.5% triethylamine. The fractions containing the
product were pooled together, and the solvent was evaporated
to give 100 mg of product as an oil (75%):1H NMR (CDCl3) δ
1.1-1.4 (m, 2H), 1.6-2.0 (m, 4H), 2.0-2.3 (m,2H), 2.5-2.7 (m,
6H) 3.0 (s, 4H, broad), 3.85 (s, 3H), 3.98 (t, 2H),4.3-4.7 (md,
1H, fluorine coupling), 6.8-7.0 (m, 4H), 7.2-7.4 (m, 4H), 7.8
(td, 1H), 8.55 (dd, 1H); MS (EI) 440 (M+), 425, 311, 278, 249,
218, 205, 190, 162. Anal. (C25H33FN4O2F) Calcd: C, 68.16; H,
7.55; N, 12.72. Found: C, 68.02; H, 7.82, N, 12.52.
P en ta m eth yl 3-Meth yl-4-(tr im eth yla m m on iu m tr iflu o-
r om eth a n esu lfon a te)ben zoa te (5c). To a solution of 7 (255
mg, 0.75 mmol) in 5 mL of anhydrous methylene chloride was
added methyl trifluoromethanesulfonate (90 µL, 0.8 mmol),
and the solution was stirred at room temperature overnight.
At the end of the reaction, the solvent was evaporated under
reduced pressure and the residue was recrystallized from
methylene chloride/hexane to give 142 mg (37.5%) of white
crystals: mp 216 °C dec; 1H NMR (CDCl3) δ 2.1-2.3 (m, 15H),
2.75 (s, 3H), 3.8 (s, 9H), 5.5 (s, 2H), 7.8-8.1 (m, 3H).
P en ta m eth yl 4-(Tr im eth yla m m on iu m tr iflu or om eth -
a n esu lfon a te)ben zoa te (5b). Compound 5b was prepared
by a similar procedure from 8 with a 51.2% yield: mp 219 °C
1
dec; H NMR (CDCl3) δ 2.2-2.4 (3s, 15H), 3.75 (s, 9H), 5.52
(s, 2H), 7.85 (d, 2H), 8.2 (d, 2H).
P en ta m eth ylben zyl 4-(Br om om eth yl)ben zoa te (5e). To
a solution of 4-(bromomethyl)benzoic acid (300 mg, 1.0 mmol)
and 1,3-dicyclohexylcarbodiimide (DCC) (210 mg, 1.0 mmol)
in 4 mL of methylene chloride was added pentamethylbenzyl
alcohol (200 mg, 1.1 mmol). The mixture was stirred at room
temperature overnight. The solution was filtered through a
small silica gel filter (1 mL) and diluted with hexane. The
product (160 mg, 30.4%) was obtained from the above solution
1
as a white crystal: mp 166-168 °C; H NMR (CDCl3) δ 2.2-
2.4 (3s, 15H), 4.48 (s, 1H), 5.49 (s, 2H), 7.42 (d, 2H), 8.10 (d,
2H); MS (EI) 376 (M+), 295, 197, 160, 145.
P en ta m eth ylben zyl 4-(Br om om eth yl)-3-n itr oben zoa te
(5f). Compound 5f was prepared by a similar procedure as
above from 4-(bromomethyl)-3-nitrobenzoic acid and penta-
1
methylbenzyl alcohol with a 29% yield: mp 136-137 °C; H
NMR (CDCl3) δ 2.2-2.4 (m, 15H), 4.8 (s, 1H), 5.55 (s, 2H),
7.62 (d, 1H), 8.22 (dd, 1H), 8.62 (d, 1H).
cis-P en ta m eth ylben zyl 4-Hyd r oxycycloh exa n eca r box-
yla te (9). To a solution of a mixture of cis- and trans-4-
hydroxycyclohexanecarboxylic acid (4.8 g, 32.4 mmol) (obtained
from hydrolysis of the mixture of cis- and trans-ethyl 4-hy-
Syn t h esis of F -18-La b eled WAY 100635 An a logu es.
Anhydrous [18F]fluoride ion was prepared from aqueous fluo-
ride produced by irradiating [18O]H2O. A typical resolubiliza-
tion procedure is described below.

Fluorine-18-Labeled 5-HT1A Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9 1585
To a 1-mL V-vial containing 3 µmol of potassium carbonate
in 15 µL of water and 6 µmol of Kryptofix-2.2.2 in 30 µL of
acetonitrile was added 300-500 µL of activity (30-50 mCi).
The water was remove with argon flow three times using
anhydrous acetonitrile on a 105 °C heating block. To above
test tube containing the anhydrous [18F]fluoride ion was
added3 mg of substrate in 0.1 mL of acetonitrile. The vial was
sealed and heated on the heating block for 10 min. The reaction
mixture was cooled to room temperature and diluted with 1
mL of ether. The ether solution was passed through a small
silica gel column (0.5 mL) to a 5-mL V-vial and rinsed with
another 1 mL of ether. The solvent was evaporated with argon
flow, and the pentamethylbenzyl ester was hydrolyzed with
0.1 mL of trifluoroacetic acid in 2 min. The trifluoroacetic acid
was removed with argon flow, and 40 mL of R,R-dichloromethyl
methyl ether was added to the reaction vial. The vial was
sealed using a cap with Teflon seal, heated at 90 °C for 5 min,
and cooled at room temperature for 2 min. The vial cap was
opened, and the vial was put back to the heating block for
another 3-5 min until the vial became dry. The vial was cooled
again, and 5 mg of substrate in 0.2 mL of acetonitrile
containing 5 µL of triethylamine was added. The vial was
sealed and heated at 90 °C for 5 min. The vial was cooled in
water and the reaction mixture diluted with 0.3 mL of HPLC
solvent (40-50% acetonitrile in 5 mM phosphate buffer
containing triethylamine at a flow rate of 5 mL/min) and
injected onto the reversed-phase HPLC column. The fraction
containing the product was collected, diluted with 2 volume
of water, and passed through a 1-mL C-18 Bond Elut column.
The column was washed with 10 mL of water, and the product
trapped on the column was eluted off with 0.5 mL of ethanol.
The [35S]GTP-γS saturation binding to cell membranes was
also determined. Receptor-linked G protein activation at 5-HT
receptors was determined by measuring the stimulation of [35S]-
GTP-γS binding. The Ki for the test compound was determined
by inhibition studies of this activation.32
Ra t Biod istr ibu tion Stu d ies. Sprague-Dawley male rats,
200-300 g, were injected intravenously while awake. Carbon
dioxide was used for euthanasia according to the NIH guide-
lines for animal use. In the biodistribution studies the rats
were injected via the tail vein with 50 µCi of F-18 or 200 µCi
of C-11 radioligand or co-injected with 50 µCi of F-18 or 200
µCi of C-11 radioligand and 50 nmol of cold WAY 100635 (4a ).
The rats were sacrificed at 30 min, and the brain was
immediately placed in 0.3 M sucrose on ice. The blood and
tissues were excised from each animal and weighed. The brain
was dissected on ice, and the various brain regions were
weighed. The radioactive content of the blood and various
tissues was assessed by gamma counting.
Ra t Meta bolite Stu d ies. The rats were injected intrave-
nously with 100 µCi of the F-18 compounds or 400 µCi of [11C]-
4a and sacrificed at 5, 10, 15, 30, or 60 min. Blood and brain
were removed, weighed, and counted to determine the total
activity. After the blood was centrifuged, the serum was
removed, mixed with an equal volume of acetonitrile, and
centrifuged. The brain was placed in an equal volume of
acetonitrile and homogenized for 15-30-s bursts. Following
centrifugation, the radioactive content of the supernatant and
pellets was determined. The supernatants were applied to TLC
plates, developed, and placed on a phosphorimaging plate
overnight. The plates were scanned the next day using a Fuji
Bio-imaging Analysis System 1500.
Meta bolite An a lysis by TLC. Brewster et al.31 used two
silica gel TLC systems to identify cyclohexanecarboxylic acid
(CHCA, 2a ) and its metabolites. The first used acetone/light
petroleum/acetic acid (20:40:1 by vol), and the second used
1-butanol/acetic acid/water (4:1:1 by vol). They identified
hexahydrohippurate, 3,4,5,6-tetrahydrohippurate, hippurate,
cyclohexanecarbonyl glucuronide, benzoyl glucuronide, and the
parent cyclohexanecarboxylate. In the first TLC system, they
reported Rf values of 0.44 for hexahydrohippurate, 0.31 for the
hippurate, and 0.1-0.2 for the glucuronide.
Ack n ow led gm en t. We gratefully acknowledge the
NIMH/NovaScreen Psychotherapeutic Drug Discovery
and Development Program (Contract No. NO1MH2003)
for the in vitro assays of 5-HT1A ligands. We would also
like to thank Dr. Dale Kiesewetter and Dr. Mark
Sassaman for valuable suggestions and the critical
reading of this manuscript.
In all systems we observed compounds mostly likely to be
conjugates analogous to Brewster’s work. We prepared 6a -
6c and determined their Rf values in the various chromato-
graphic systems. Radioactivity was observed at the same Rf
value and is reported as conjugate in the following analysis.
Likewise standards of the potential metabolites were also run
to determine the Rf values. We used two silica gel-based TLC
systems: (A) 1-propanol/ammonia/acetonitrile (50:30:20) where
the Rf values were 4a 0.90, CHCA and conjugate 0.7, 4c 0.97,
MeFB 0.9, conjugate 0.85. 4d and its metabolites have
comparable Rf values to those found for 4a with fluoride at
0.4. We also used Brewster’s first system (system B) [acetone/
light petroleum ether/acetic acid (20:40:1)] where the Rf values
were 4a 0.05, CHCA and conjugate at 0.7, 4c 0.1, MeFB 0.6,
MeFB conjugate 0.05. For 4d in the same system, fluoride,
the parent compound, and the conjugates stayed at the origin
and FC traveled to 0.65. We also carried out HPLC separations
using a semipreparative C18 reversed-phase column eluted
with 65-70% methanol in 0.1 M ammonium formate buffer
at pH 6.5.
Refer en ces
(1) Fletcher, A.; Clife, I. A.; Dourish, C. T. Silent 5-HT1A receptor
antagonists: utility as research tools and therapeutic agents.
TIPS 1993, 141, 441-448.
(2) Alexander, S. P. H.; Peters, J . A. 1998 Receptor and Ion Channel
Nomenclature Supplement. TIPS 1998.
(3) Crouzel, C.; Guillaume, M.; Barre, L.; Lemaire, C.; Pike, V. W.
Ligands and tracers for PET studies of the 5-HT system-current
status. Nucl. Med. Biol. 1992, 19, 857-879.
(4) Fletcher, A.; Bill, D. J .; Cliffe, I. A.; Dover, G. M.; Forster, E. A.;
Haskins, J . T.; J ones, D.; Mansell, H. L.; Reilly, Y. WAY-
100135: a novel, selective antagonist at presynaptic and postsyn-
aptic 5-HT1A receptors. Eur. J . Pharmacol. 1993, 237, 283-291.
(5) Khawaja, X.; Evans, N.; Reilly, Y.; Ennis, C.; Minchin, M. C. W.
Characterization of the binding of [3H] WAY100635, a novel
5-hydroxytryptamine 1a receptor antagonist, to rat brain. J .
Neurochem. 1995, 64, 2716-2726.
(6) Kung, M.-P.; Zhuang, Z.-P.; Frederick, D.; Kung, H. F. In vivo
binding of [123I] 4-(2′-methoxyphenyl)-1-[2′-(N-2”-pyridinyl)-p-
iodobenzamido]ethylpiperazine, p-MPPI, to 5-HT1A receptors in
rat brain. Synapse 1994, 18, 359-366.
(7) Laporte, A.-M.; Lima, L.; Gozlan, H.; Hamon, M. Selective in
vivo labeling of brain 5-HT1A receptors by [3H] WAY 100635 in
the mouse. Eur. J . Pharmacol. 1994, 271, 505-514.
(8) Hume, S. P.; Ashworth, S.; Opacka-J uffry, J .; Ahier, R. G.;
Lammertsma, A. A.; Pike, V. W.; Cliffe, I. A.; Fletcher, A.; White,
A. C. Evaluation of [O-methyl-3H] WAY100635 as an in vivo
radioligand for 5-HT1A receptors in rat brain. Eur. J . Pharmacol.
1994, 271, 515-523.
(9) Mathis, C. A.; Simpson, N. R.; Mahmood, K.; Kinahan, P. E.;
Mintun, M. A. [11C]WAY 100635: A radioligand for imaging
5-HT1A receptors with positron emission tomography. Life Sci.
1994, 55, 403-407.
(10) Osman, S.; Lundkvist, C.; Pike, V. W.; Halldin, C.; McCarron,
J . A.; Swahn, C.-G.; Ginovart, N.; Luthra, S. K.; Cliffe, I. A.;
In Vitr o Stu d ies. Receptor binding studies were carried
out by NovaScreen. In general, brain tissue was used in
conjunction with the 5-HT1A subtype-specific tritiated ligand
to determine the Ki of the various compounds. In addition, Ki
values (nM) were determined against [3H]8-OH-DPAT binding
to a cloned cell line containing human 5-HT1A receptors.
Receptor binding was also determined by MDS Panlabs
(Bothell, WA). The binding assay was carried out with [3H]8-
OH-DPAT as the radioligand and cloned cells as the source of
receptor with incubation for 60 min at 25 °C. All data are
presented as averages ( SEM of at least six determinations
unless otherwise noted.

1586 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 9
Lang et al.
Fletcher, A.; Farde, L. Characterization of the radioactive
metabolites of the 5-HT1A receptor radioligand, [O-methyl-11C]
WAY100635, in monkey and human plasma by HPLC: com-
parison behavior of an identified radioactive metabolite with
parent radioligand in monkey using PET. Nucl. Med. Biol. 1996,
23, 627-634.
(21) Shiue, C.-Y.; Shiue, G. G.; Zhuang, Z. P.; Kung, M.-P.; Kung, H.
F. Synthesis of no-carrier-added 4-(2′-methoxyphenyl)-1-[2′-(N-
2′-pyridinyl)-p-[F-18]fluoro-benzamido]ethylpiperazine as a po-
tential 5-HT1a receptor ligand for PET studies. J . Nucl. Med.
1994, 35, 252P.
(22) Wagner, J . Fundamentals of Clinical Pharmacokinetics; Drug
Intelligence Publications Inc.: Hamilton, IL, 1981.
(23) Carson, R. E.; Schmall, B.; Endres, C. J .; Lang, L.; Der, M. G.;
Adams, H. R.; J agoda, E.; Herscovitch, P.; Eckelman, W. C.
Kinetic modeling of the 5-HT1A antagonist [carbonyl-C-11]WAY-
100635. J . Cereb. Flow Metab. 1997, 17, S327.
(11) Pike, V. W.; McCarron, J . A.; Lammertsma, A. A.; Hume, S. P.;
Poole, K.; Grasby, P. M.; Malizia, A.; Cliffe, I. A.; Fletcher, A.;
Bench, C. J . First delineation of 5-HT1A receptors in human
brain with PET and [11C]WAY-100635. Eur. J . Pharmacol. 1995,
283, R1-R3.
(12) Pike, V. W.; McCarron, J . A.; Lammertsma, A. A.; Osman, S.;
Hume, S. P.; Sargent, P. A.; Bench, C. J .; Cliffe, I. A.; Fletcher,
A.; Grasby, P. M. Exquisite delineation of 5-HT1A receptors in
human brain with PET and [carbonyl-11C]WAY-100635. Eur. J .
Pharmacol. 1996, 301, R5-R7.
(24) Zhuang, Z. P.; Kung, M.-P.; Mu, M.; Kung, H. F. Cyclic amide
derivatives of 4-(2′-methoxyphenyl)-1-[2′-(N-2′-pyridinyl)-p-io-
dobenzamido]ethylpiperazine (p-MPPIO) as 5-HT 1a receptor
ligands. J . Labelled Compd. Radiopharm. 1997, 88-90.
(25) Wilson, A. A.; DaSilva, J . N.; Inaba, T.; Fischer, N.; Houlew, S.
Analogues of WAY 100635 as potential radiotracers for imaging
5-HT1A receptors by positron emision tomography (PET). J .
Labelled Compd. Radiopharm. 1997, 531-533.
(26) Pike, V. W.; McCarron, J . A.; Lundkvist, C.; et al. [Carbonyl-
11C]desmethyl-WAY-100635 a new potent and selective radio-
ligand for brain 5-HT1a receptors in vivo. J . Labelled Compd.
Radiopharm. 1997, 568-570.
(27) Khawaja, X. Quantitative autoradiographic characterisation of
the binding of [3H]WAY-100635, a selective 5-HT1A receptor
antagonist. Brain Res. 1995, 673, 217-225.
(28) Eckelman, W. C.; Grissom, M.; Conklin, J .; Rzeszotarski, W. J .;
Gibson, R. E.; Francis, B. E.; J agoda, E. M.; Eng, R.; Reba, R.
C. In vivo competition studies with analogues of 3-quinuclidinyl
benzilate. J . Pharm. Sci. 1984, 73, 529-534.
(29) Buu-Hoi, N. P.; J acquignon, P. Carcinogenic nitrogen Com-
pounds: part XII. Fluorine containing 1:2- and 3:4-Benza-
cridines. J . Chem. Soc. 1952, 4173-4175.
(30) Brewster, D.; J ones, R. S.; Parke, D. V. The metabolism of
cyclohexanecarboxylate in the rat. Biochem. J . 1977, 164: 595-
600.
(13) Carson, R. E.; Schmall, B.; Endres, C. J .; Lang, L.; Der, M. G.;
Adams, H. R.; J agoda, E.; Herscovitch, P.; Eckelman, W. C.
Kinetic analysis of the 5-HT1a antagonist [carbonyl-C11]WAY-
100635. J . Nucl. Med. 1997, 38, 80P-81P.
(14) Zhuang, Z.-P.; Kung, M.-P.; Kung, H. F. Synthesis and evalu-
ation of 4-(2′-methoxyphenyl)-1-[2′-[N-(2′-pyridinyl)-p-iodoben-
zamido]ethyl]piperazine (p-MPPI):
a new iodinated 5-HT1a
ligand. J . Med. Chem. 1994, 37, 1406-1407.
(15) Schmall, B.; Lang, L.; J agoda, E.; Channing, M.; Sassaman, M.;
Eckelman, W. C. Antagonists for 5HT1a and 5HT2a receptors.
J . Nucl. Med. 1996, 37, 204P.
(16) Thielen, R. J .; Frazer, A. Effects of novel 5-HT1A receptor
antagonists on measures of postsynaptic 5-HT1A receptor
activation in vivo. Life Sci. 1995, 56, 163-168.
(17) Forster, E. A.; Cliffe, I. A.; Bill, D. J .; Dover, G. M.; J ones, D.;
Reilly, Y.; Fletcher, A. A pharmacological profile of the selective
silent 5-HT1A receptor anatagonist, WAY-100635. Eur. J . Phar-
macol. 1995, 281, 81-88.
(18) Pike, V. W.; McCarron, J . A.; Hume, S. P. Pre-clinical develop-
ment of a radioligand for studies of central 5-HT1a receptors in
vivo-[11C]WAY-100635. Med. Chem. Res. 1995, 5, 208-227.
(19) Lang, L.; Eckelman, W. C. One step synthesis of Fluorine-18
labeled [18F]-N-succinimidyl 4-(fluoromethyl)benzoate for protein
labeling. Appl. Radiat. Isot. 1994, 45, 1155-1163.
(20) Kilbourn, M. Fluorine-18 labeling of radiopharmaceuticals;
Nuclear Science Series, NAS-NS-3203; National Academy
Press: Washington, DC, 1990.
(31) Newman-Tancredi, A.; Audinot, V.; Chaput, C.; Verriele, L.;
Millan, M. J . [35S]Guanosine-5′-O-(3-thio)triphosphate binding
as a measure of efficacy at human recombinant dopamine D4.4
receptors: actions of antiparkinsonian and antipsychotic agents.
J . Pharmacol. Exp. Ther. 1997, 282, 181-191.
J M980456F