Synthesis and structure-activity relationship of a novel series of aminoalkylindoles with potential for imaging the neuronal cannabinoid receptor by positron emission tomography

Article abstract of DOI:10.1021/jm0502743

A new series of CB₁ ligands with high binding affinity (K _i = 0.7-100 nM) and moderate lipophilicity (cLogD_7.4) in the range of 2.1-4.5 has been synthesized. A structure-activity relationship study demonstrated that for the studied set of aminoalkylindoles, the molecular dipole of the ground state conformation within the series was inversely related to the affinity. The racemic ligand with highest affinity (0.7 nM), 3-(4-fluoronaphthoyl)-1-(N-methylpiperidin-2-ylmethyl)indole, was radiolabeled with ¹⁸F. This radioligand specifically labeled CB₁ receptors in mouse brain and accumulated in regions of high versus low CB ₁ receptor density in a ratio of 1.6. The displaceable radioactivity of one enantiomer in the brains of mice determined in a pretreatment study using the CB₁ antagonist N-(piperidinyl)-5-(4-chlorophenyl)-1-(2,4- dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (SR141716) was nearly double that of the race-mate for the same determination; therefore, the active enantiomer is a candidate for PET studies in animals. A pretreatement study for the other enantiomer found no displaceable radioactivity in the same group of mice; this result suggested the enantiomer was inactive.

Full text of DOI:10.1021/jm0502743

J. Med. Chem. 2005, 48, 5813-5822
5813
Synthesis and Structure-Activity Relationship of a Novel Series of
Aminoalkylindoles with Potential for Imaging the Neuronal Cannabinoid
Receptor by Positron Emission Tomography
Peter G. Willis,*,† Olga A. Pavlova,† Svetlana I. Chefer,† D. Bruce Vaupel,† Alexey G. Mukhin,† and
Andrew G. Horti‡
Neuroimaging Research Branch, Intramural Research Program, National Institute on Drug Abuse, NIH, DHHS, 5500 Nathan
Shock Drive, Baltimore, Maryland 21224, and Department of Radiology, Johns Hopkins University School of Medicine,
600 North Wolfe Street, Baltimore, Maryland 21287
Received March 25, 2005
A new series of CB1 ligands with high binding affinity (Ki ) 0.7-100 nM) and moderate
lipophilicity (cLogD7.4) in the range of 2.1-4.5 has been synthesized. A structure-activity
relationship study demonstrated that for the studied set of aminoalkylindoles, the molecular
dipole of the ground state conformation within the series was inversely related to the affinity.
The racemic ligand with highest affinity (0.7 nM), 3-(4-fluoronaphthoyl)-1-(N-methylpiperidin-
2-ylmethyl)indole, was radiolabeled with 18F. This radioligand specifically labeled CB1 receptors
in mouse brain and accumulated in regions of high versus low CB1 receptor density in a ratio
of 1.6. The displaceable radioactivity of one enantiomer in the brains of mice determined in a
pretreatment study using the CB1 antagonist N-(piperidinyl)-5-(4-chlorophenyl)-1-(2,4-dichloro-
phenyl)-4-methyl-1H-pyrazole-3-carboxamide (SR141716) was nearly double that of the race-
mate for the same determination; therefore, the active enantiomer is a candidate for PET studies
in animals. A pretreatement study for the other enantiomer found no displaceable radioactivity
in the same group of mice; this result suggested the enantiomer was inactive.
Introduction
Cannabinoid receptors belong to the superfamily of
G-protein coupled receptors. The cannabinoid subtype
1 receptor (CB1) is found in the central nervous system
as well as some nonneural tissues.1 The other well-
known cannabinoid receptor subtype (CB2) is more
localized in the immune system.1,2 ∆9-Tetrahydro-
cannabinol (∆9-THC; 1), the principle physiologically
active ingredient in marijuana, is known to activate CB1
receptors, inducing the typical euphoric feelings associ-
ated with marijuana use. Figure 1 contains examples
of the five major classes of natural and synthetic
cannabinoid ligands including ∆9-THC, the prototype
of the classical cannabinoid ligand class. (1R,3R,4R)-3-
[2-Hydroxy-4-(1,1,-dimethylheptyl)phenyl]-4-(3-hydroxy-
propyl)cyclohexan-1-ol (CP 55940; 2) is an example of a
nonclassical cannabinoid, as it only contains some of the
structural motifs required for a classical cannabinoid.
(R)-(+)-[2,3-dihydro-5-methyl-3-[(morpholinyl)methyl]-
pyrrolo[1,2,3-de]-1,4-benzoxazinyl]-(1-naphthalenyl)meth-
anone mesylate (WIN55212-2; 4) and N-(piperidinyl)-
5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-
pyrazole-3-carboxamide (SR141716; 5) are examples of
the aminoalkylindole (AAI) and 1,5-biarylpyrazole classes
of cannabinoids, respectively. These two classes of
Figure 1. Five major classes of cannabinoid ligands. In left
to right, top to bottom order: classical, nonclassical, endog-
enous, aminoalkylindoles, and 1,5-biarylpyrazoles.4
ligands are purely synthetic in nature. Figure 1 also
includes anandamide, 3, a putative endogenous ligand
for the cannabinoid receptor system.3
Cannabinoid receptor ligands produce several char-
acteristic pharmacological effects.1,5 Consequently, the
implication of therapeutic applications for cannabinoids
exists. Possibilities range from the treatment of nausea6
* Corresponding author. Tel.: +1-410-550-1440 ext. 331. Fax: +1-
410-550-1441; E-mail: pwillis@intra.nida.nih.gov.
† National Institute on Drug Abuse.
‡ Johns Hopkins University School of Medicine.
10.1021/jm0502743 CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/06/2005

5814 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Willis et al.
Figure 2. Representative CB1 radiotracers.
and appetite control7 to CNS disorders8 and smoking
cessation.7 However, the role of cannabinoid receptors
in the cause and treatment of these disorders is still
not fully understood.5
Figure 3. Structure of CB1 ligands with SAR information.
Table 1. Synthesized CB1 Ligands, Binding Affinity (Ki,
The ability to image CB1 receptors in the living
human and animal brain using PET would provide the
ability to conduct receptor studies in normal physiologi-
cal conditions and in disease states. In addition, the
development of a useful positron emission tomography
(PET) radiotracer for CB1 receptors may assist in the
design and testing of promising pharmaceuticals target-
ing this receptor. Several analogues of 5 have been
labeled with positron emitting isotopes and tested in
animals,9,10 (Figure 2) but the search for a ligand
suitable for quantitative PET imaging and analysis of
the CB1 receptors11,12 is a difficult one. Existing CB1
radiotracers for in vivo imaging studies exhibit low
ratios of radioactivity accumulation in regions of high
versus low density of CB1 receptors (1.6-2.5) due to
their high lipophilicity and/or insufficient binding af-
finity.9,10,12,13 It is apparent that currently available CB1
ligands require optimization and improvement.
Efforts to develop PET and single photon emission
computed tomography (SPECT) radioligands (Figure 2)
have focused on finding new ligands with improved
binding affinity and reduced lipophilicity.9-11,14,15 Un-
fortunately, a recent structure-activity relationship
(SAR) study with analogues of 5 demonstrated that for
this subset of the 1,5-biarylpyrazole group, ligands with
higher affinity tended to be more lipophilic or vice
versa.14
Since ligands 6a-f are of limited use due to their low
affinity or high lipophilicity, other less lipophilic ligands
may be more useful, for example the AAI class of
ligands. Previous SAR studies with AAI have demon-
strated the importance of the aminoalkyl tail, the alkyl
chain connected to the nitrogen atom of the indole, for
the binding affinity.16 The highest affinity ligand within
the series was 3-naphthoyl-1-(N-methylpiperidin-2-yl-
methyl)indole, 8 (Ki ) 0.9 nM), with the N-methyl-
piperidin-2-ylmethyl tail (Figure 3).16 Modification of the
tail is also an important tool in controlling lipophilicity
of an AAI ligand.
determined in vitro), and Lipophilicity (cLogD7.4
)
a
ligand
Z
Ki, nM
cLogD7.4
816
9a
9b
9c
9d
9e
1-naphthyl
4-Me-1-naphthyl
4-F-1-naphthyl
0.9 ( 0.1
1.1 ( 0.1
0.7 ( 0.1
1.7 ( 0.2
2.3 ( 0.3
2.2 ( 0.3
1.1 ( 0.1
1.3 ( 0.1
4 ( 1
3.6
4.0
3.8b
4.5
3.6
3.3
3.3
2.7
2.8
4.0
3.0
3.0
2.1
3.1
3.5
4-Br-1-naphthyl
4-OMe-1-naphthyl
4-OH-1-naphthyl
4-CN-1-naphthyl
4-(2-hydroxyethyl)-1-naphthyl
6-OH-1-naphthyl
2-Me-1-naphthyl
2-iodophenyl
9f
9g
9h
9i
7.8 ( 0.6
1.2 ( 0.3
100 ( 7
12 ( 1
9j18,19
9k
9l
N-naphthylamine
N-decahydroquinoline
1-naphthyl
9mc
1.4 ( 0.1
9n18,19 4-NO2-1-naphthyl
12.418,19
a The dissociative partition coefficient clogD7.4 was calculated
using ACD/LogD Suite software. b The experimental logD7.4 value
as determined for [18F]9b using a counting technique19 in four
assays is 2.54 ( 0.27. c Indazole instead of indole.
> Me > OH > H, which is also the same order as the
length of the substituent R from long to short.
The goal of this research was to develop cannabinoid
ligands with lower lipophilicity (logD) and increased
affinity. The similarity of the structure of 7 R ) H to 8
suggested that similar substitutions may also improve
Ki. Therefore, a new series of derivatives 8, 9a-n (Table
1) was synthesized to test the effects of different
substituents on affinity and lipophilicity. The ligand
with the best combination of these characteristics was
then labeled with a positron emitting isotope to test its
potential as a PET radioligands for CB1 receptors.
In the present study we used compound 8 as a lead
for development of high affinity CB1 ligands with lower
lipophilicity. The structure of compound 8 is comparable
to the structure 7 (R ) H) with the only dissimilarity
being a different piperidine attached to the indole. A
comparison of the two structural motifs reveals that
both amines in the piperidine ring are tertiary and
separated from the indole nitrogen by two carbons
(shown in bold in Figure 3). In a SAR study by
Eissenstat et al. improvements in the Ki of 7 were shown
when small R groups were substituted for hydrogen.17
The order of affinity for substitutions on 7 was R ) OMe
Results and Discussion
Structure-Activity Relationships (SAR). Twelve
new AAI compounds with binding affinity and lipophi-
licity in the range of Ki ) 0.7 to 100 nM and logD7.4
2.1 to 4.5 were synthesized (Table 1). For all the
structures in Table 1 where Z ) 4-R-naphthyl (9a-g,
9n), only R ) F gave an improved Ki as compared with
the lead compound 8. The length of the R-group was
defined by measuring the distance from the naphthalene
ring to the far atom of the group when the system has
been minimized with an AM1 semiempirical calculation
)

Imaging the Neuronal Cannabinoid Receptor
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5815
C2,C1,indole C3,C2 (Figure 5) that is 10° wider than
those of 9b or 8. The observation of decreased affinity
is consistent with the idea of a directing effect of the
group ortho to the carbonyl group (e.g., the iodo group
in 9j, and the other benzyl group of the naphthalene in
8).
This SAR is a refinement of the SAR from Shim et
al. or Tetko et al., who examined many cannabinoid
ligands including 7 and 8, as mentioned in the Intro-
duction.17,21 These authors indicated that more negative
electrostatic potential on carbon 4 of the naphthalene
ring of 8 would increase affinity. The authors also
suggest that a sterically forbidden region surrounds C2
and C3 of the naphthyl region. For ligand set 9, only
9b with a fluorine atom was an improvement over 8.
This can be explained by understanding that the sub-
stituent is a source of negative electrostatic potential
that can change the size of the dipole with the strength
of the potential and its location. Figure 5 shows the
direction of the molecular dipole of 8 with the z
component of the dipole vector of the dipole being zero
due to the positioning of the molecule. The z axis is
perpendicular to the plane of the paper; therefore, the
vector shown in Figure 5 is parallel to the plane of the
paper. Adding fluorine on carbon 4, as found on 9b, will
decrease the y component of this dipole versus that
shown in Figure 5.
Figure 4. Correlation of the molecular dipole in log(Debye)
with binding affinity, log(1/Ki, nM) for 8, 9a-g.
Possible explanations for why the correlation between
dipole and affinity exist could come from the structure
of the binding site for this molecule. For example, if a
source of negative electrostatic potential on the receptor
binding site exists near carbon 4, 5, or 6 of the
naphthalene ring in a hypothetical binding site, large
electronegative substituents would interfere with bind-
ing at this potential site. Or conversely, if a source of
electropositive potential existed on the binding site near
carbon 2 or 3, electronegative substituents (e.g. fluorine)
nearer to the source of potential would increase poten-
tial. This hypothesis is consistent with the CoMFA
model proposed by Shim et al. as explained above. The
sterically forbidden region around C2 is seen in these
data; however, it is most likely due to the large confor-
mational changes caused by substitutions at C2 on the
naphthalene ring. On the other hand, the explanation
may come from an intrinsic affinity of the receptor for
apolar ligands (e.g. anandamide or ∆9-tetrahydro-
cannabinol). This explanation was made less likely due
to the low correlation between lipophilicity and affinity.
Other possible relationships include one between
lipophilicity and affinity cited by Katoch-Rouse et al.
for 1,5-biarylpyrazole ligands.14 This relationship indi-
cated that increased affinity for this set of ligands led
to increased lipophilicity. When the current set of
ligands was checked for evidence of this relationship,
none was found. From this it can be inferred that for
the AAI class of ligands lipophilicity is not intrinsic to
affinity.
Figure 5. The global energy minimum conformation of
compound 8, and the direction of the molecular dipole as
determined by an AM1 calculation.
and was assumed to be independent of the amino alkyl
tail. Although the affinity appears to decrease for the
larger substituents, the correlation of R length and Ki
for 8, 9a-o was weak.
Comparisons of the molecular dipole versus binding
affinity with CB1 receptors showed a correlation within
the tested set of structures (Figure 4). The R correlation
factor was 0.82 for the relationship of log(Debye), a
measure of the overall dipole strength, versus log(1/Ki),
a measure of a ligand’s affinity to CB1 receptors. A
dipole was calculated for the global energy minimum
as determined by an AM1 calculation. The magnitude
of the dipoles in this study indicated the difference in
electron distribution of one side of the molecule versus
the other side of the molecule (Figure 5). The calculation
demonstrated that dipoles for 9b and 8 were lower than
those for 9d and 9h. The differences in the dipole of
studied structures are primarily due to the changes in
the properties of the substituent and the electron
distribution in the naphthalene ring resulting from
those changes.
Two points that do not seem to be adequately de-
scribed by this correlation are the indazole 9m and the
2-methylnaphthoyl derivative 9i. The outlier 9m sug-
gests there are other negative interactions at the
indazole nitrogen or that the effect is from a more
localized dipole. 9i has a more rigid structure than the
other ligands, and has a torsion angle for naphthyl
It is noteworthy that the global minimum energy
conformer of compounds 9g and 2 can be easily super-
imposed as shown by the six pairs of atoms indicated
in Figure 6. Kandemirli et al. uses a similar alignment
to compare a mixed series of cannabinoids including 8
and 2.22 The similarity of the two ligands is an indicator
that they may share some of the same van der Waals

5816 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Willis et al.
Scheme 2. Preparation of Indazole 9m by Route 2a
a Reagents: (a) refluxing CH2Cl2, 2.5 h; (b) DMF, NaH, RT, 2.5
d.
Scheme 3. Synthesis of Amide Derivatives 9k, l by
Route 3 with Acid Chloride 6a
a Reagents: (a) trichloroacetyl chloride, NEt3; NaOH; SOCl2;
(b)(for 9k) 2-aminonaphthalene; or (c)(for 9l) decahydroquinoline,
CH2Cl2, NEt3.
piperidin-2-ylmethyl)indole) under two different condi-
tions. This route was used to synthesize 8, 9a-j, n. 8,
9a-d, f, i, and n were prepared in the presence of Lewis
acid. Compounds 9e, h, and g were prepared in a
solution of ionic liquid using the basic Friedel-Crafts
reaction mechanism.23 The ionic liquid23 improved the
yield over dipolar aprotic solvents and allowed the use
of acid labile protecting groups, including benzyl and
triisopropylsilyl (TIPS) ethers. Compounds 9e and 9h
were synthesized under condition (b) with the corre-
sponding benzyl derivatives 19 and 20, and 9g via the
TIPS ether, 21, and deprotected in a normal manner
(Scheme 1).
Figure 6. “Flat” and 3-D schemes of the superimposition of
8 and 2 by the six pairs of atoms (a-f).
Scheme 1. Synthesis of 8, 9a-j, n by Friedel-Crafts
Acetylation by Route 1a
Route 2 (Scheme 2) was used for the synthesis of the
indazole, 9m. When 2-diazocarboxylic acid was heated,
benzyne was generated and underwent a 1,3-dipolar
cycloaddition with 1-naphthyldiazomethyl ketone to
generate 11. Intermediate 11 was N-alkylated at posi-
tion 1 on indazole with 2-chloromethyl-N-methylpiperi-
dine.24 Route 3 (Scheme 3) generated 9k and 9l from
the acid chloride, 12, and 2-aminonaphthalene, or
decahydroquinoline via amide-forming conditions. The
acid chloride 12 was synthesized from 10 by a proce-
dure25 used for a similar amine that involved an
acetylation at indole carbon 3 with trichorloacetyl
chloride, hydrolysis of the trichloroacetyl group, and
conversion to the acid chloride.
Scheme 4 shows the synthesis of the acids used to
make 9e, h, and g. The precursor, 15, for 9g was
generated from 1,4-dibromonaphthalene. This synthesis
used a monoselective lithiation of 1,4-dibromonaphtha-
lene. When 1 equiv of n-butyllithium was added to the
reaction at -78 °C, monoalkalation is the primary
product. The alcohol was then protected as the TIPS
ether, 14, and carboxylated to give the acid 15. The
synthesis of precursor 17, shown in Scheme 4, for the
synthesis of 9e starts with the para bromination of
1-naphthol. By using acetonitrile as the solvent, bro-
mination is directed to carbon 4, giving 16 in high
yield.26 The bromine could then be carboxylated to give
a Reagents: (a) (for 8, 9a-d, f, i, and n) CH2Cl2, substituted
naphthoyl chloride, AlCl3; (b) (for 9e, h, and h) [Bmin][PF6],
substituted naphthoyl chloride, heat; (c) (for 9e, h) 10% Pd/C,
ammonium formate; (d) (for 9g) TBAF.
interactions. Figure 6 shows why it is reasonable to
believe the two structures have conformational similari-
ties, and why substitution at C4 of the naphthyl ring
may give positive van der Waals interactions. Indole
ligands with an aminoalkyl tail (e.g., ethyl-N-piperidine
from 7, or (N-methylpiperid-2-yl)methyl from 8) or long
chain tail (e.g., n-pentane) both give high affinity
ligands, indicating the tails may have similar functions
despite different tails.
Chemistry. The synthesis of the above ligands was
accomplished through one of three routes. Route 1
(Scheme 1) consists of a Friedel-Crafts acetylation of
an acid chloride synthon and indole 1016 (1-(N-methyl-

Imaging the Neuronal Cannabinoid Receptor
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5817
Scheme 5. Radiosynthesis of [18F]9b-1, [18F]9b-2a
Scheme 4. Synthesis of a Precursor for
Aminoalkylindole 9g, e, ha
a (a) K222/K18F, DMSO, 160 °C; (b) Chiralcel OD HPLC separa-
tion.
a Reagents: (a) n-BuLi, THF 1h, -78 °C; ethylene oxide, -78
°C to RT; NaH, TIPSCl; (b) n-BuLi, THF, 3 h, -78; CO2(g), -78
°C to RT; (c) NBS, MeCN; BnBr, NaH, DMF; (d) KOH, DMF, BnBr;
THF, K2CO3, reflux.
17. 6-Benzoxynaphthoic acid, 18, a precursor for 9h, was
synthesized from 6-hydroxynaphthoic acid. 9n was
synthesized as described previously with minor modi-
fications.18
In Vitro Binding Assay. The binding affinities of
ligands 8, 9a-n for CB1 receptors were determined by
measuring the ability of each ligand to compete with
[3H]2 binding in membrane fractions prepared from rat
cerebellum (for data, see Table 1).
Lipophilicity. LogD7.4 is a partition coefficient for
charged species between octanol and water at pH 7.4.
Figure 7. Chromatogram of chiral separation of racemic
[18F]9b, showing UV (254 nm) and radioactivity profiles. Peaks
at 7.4 min (9b-1) and 11.7 min (9b-2) were approximately
equal in area. Other peaks were nonradioactive byproducts.
logDpH can be defined as: logDpH ) log(∑aiorg/∑aiwater
)
where aiorg is the concentration of the ith microspecies
in octanol and aiwater is the concentration of the ith
microspecies in water. The lipophilicity, clogD7.4, was
calculated with the ACD/LogD software for all the
compounds in this study. Lipophilicity was also deter-
mined experimentally20 for [18F]9b (see Table 1).
tion of the enantiomers, obtaining two nearly equal
doses of radiolabeled compounds.
In this study we did not decode the chiral structures
of the enantiomers [18F]9b-1 and [18F]9b-2. The two
enantiomers came off the HPLC column (Figure 7) in
the written order with [18F]9b-2 being the active isomer
(see below). Pure samples of the nonlabeled enantiomers
were prepared with a preparative chiral column and
were analyzed for specific rotation. 9b-2 gave (+)
rotation, the same sign as the R enantiomer for 8, and
9j.16;18
In Vivo Studies. Two separate studies were con-
ducted in mice; the first one used the racemate [18F]9b,
and the second one tested enantiomers [18F]9b-1 and
[18F]9b-2 separately. Racemic [18F]9b showed moderate
uptake into the whole mouse brain with maximum
radioactivity accumulation of 0.033% (ID/kg)/g, at 7 min
after administration (Figure 8). The maximum ac-
cumulation of radioactivity in the brain for racemic
[18F]9b is lower compared with some recent CB1 tracers
with antagonist properties, 6b-f. The maximum radio-
activity of compounds with similar maximum target to
nontarget tissue ratios of radioactivity was 0.11 for 6b
or 0.18%(ID/kg)/g for 6f.12
Radiochemistry. Compound 9b manifests the high-
est binding affinity with the series and moderate
lipophilicity. Therefore, it was chosen as a candidate for
radiolabeling with the PET isotope 18F. The presence
of a fluorine substituent in an activated position of the
naphthalene ring suggested a straightforward radio-
synthesis of [18F]9b via nucleophilic radiofluorination.
Aromatic nitro groups readily undergo substitution
reactions with the no-carrier-added (NCA) radiofluori-
nation method when they are positioned para or ortho
to an activating group.27 9n contains an aromatic nitro
substituent para to a carbonyl. With this in mind, the
[18F]9b was prepared with 10% radiochemical yield by
the reaction of 9n with Kryptofix222/K18F complex and
K2CO3 in DMSO followed by reverse phase semi-
preparative separation of the reaction mixture (Scheme
5). The specific activity of the final product was in the
range of 2400-3200 mCi/µmol when prepared with
455-530 mCi [18F]fluoride. Initially, the final product
[18F]9b was synthesized as racemic mixture. However,
because previous evidence16,18 demonstrated that only
one of the AAI enantiomers exhibited high binding
affinity at CB1, we also performed chiral HPLC separa-
The patterns of regional distribution beginning 20
min after injection of [18F]9b were consistent with the
known distribution of CB1 receptors.28,29 Thus, the
highest accumulation of radioactivity was consistently
observed in the hippocampus and the lowest was seen

5818 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Willis et al.
Figure 10. Inhibition of radioactivity accumulation of 18F-
labeled enantiomers, 9b-1 and 9b-2, by 5 in brain regions with
high CB1 receptor density and in the brainstem in mice. For
each enantiomer, 5 was administered 15 min before the
radiotracer to block the specific binding of radioligands to CB1.
In control studies there was no pretreatment with 5. Measure-
ments were taken 60 min after injection of radiotracer. Bars
represent mean + SEM (n ) 3). *P < 0.02, **P < 0.01, and
***P < 0.002 as determined by Student’s t-test indicate the
significant differences between the radioactivity values of a
control study with 9b-2 and the radioactivity values of the
other studies in the corresponding tissues.
Figure 8. Kinetics of the regional distribution of the radio-
activity of racemic [18F]9b in the mouse brain: (b) hippocam-
pus, ()) striatum, (2) total brain, (0) brain stem, (4) the ratio
of hippocampus to brain stem radioactivity. Radioactivity
accumulation is expressed as a percentage that is defined as
the ratio of a concentration of radioactivity (Ci/g tissue) per
injected dose radioactivity (Ci/kg body weight) times 100. Mice
used in this test had an average body weight of 0.025 kg.
Ratios shown represent target tissue (hippocampus) to non-
target tissue (brain stem). Each data point represents the
mean ( SEM (n ) 3).
receptors. The enantiomers [18F]9b-1 and [18F]9b-2 were
also tested. Preadministration of 5 did not have any
effect on accumulation of [18F]9b-1. In other words,
there was no displaceable radioactivity. In contrast, a
significant portion of the radioactivity of [18F]9b-2 was
displaceable. For [18F]9b-2 the displaceable portion of
the radioactivity nearly doubled compared to racemic
[18F]9b. Therefore, the results of pretreatment studies
with 5 indicate that while [18F]9b-2 is an active radio-
tracer for CB1 receptors, [18F]9b-1 is inactive.
The levels of radioactivity for the three regions with
high CB1 receptor density and the brainstem, shown in
Figure 10, were compared in studies with and without
5 pretreatment. The values of target (hippocampus,
striatum, and cerebellum) to nontarget (brainstem) ratio
measured at 60 min after radioligand administration
were (1.3, 1.2, 0.9) and (2.0, 1.8, 1.4) for 9b-1 and 9b-2,
respectively. The total radioactivity determined for the
control study of [18F]9b-1 was comparable to the radio-
activity determined for pretreatment studies with [18F]9b-
1 and [18F]9b-2. In contrast, the radioactivity deter-
mined for the control experiment of [18F]9b-2 was
significantly greater than the other experiments, sug-
gesting significant specific binding to CB1 receptors. The
results of these studies suggest that [18F]9b-2 has
potential as a PET radioligand for studying CB1,
whereas [18F]9b-1 is a useful indicator of the nondis-
placeable accumulation of radioactivity for [18F]9b-2.
Figure 9. Inhibition of the radioactivity accumulation for the
18F-labeled racemate 9b and enantiomers 9b-1 and 9b-2 in
mouse whole brain. Black bars indicate the control studies in
which there was no pretreatment with 5. Grey bars indicate
the studies in which 5 was administered 15 min before the
radiotracer to block the specific binding of radioligands to CB1
receptors. Measurements were taken 60 min after injection of
radiotracer for [18F]9b-1, [18F]9b-2, and [18F]9b. Bars represent
mean + SEM (n ) 3). **P < 0.01 as determined by Student’s
t-test indicate the significant differences of the radioactivity
in studies with 5 pretreatment and the radioactivity of the
control study.
in the brain stem. Following peak accumulation, a
gradual decline in radioactivity was observed through-
out the rest of the experiment. The hippocampus to
brain stem ratio of accumulated radioactivity of racemic
[18F]9b reached a maximum value of 1.6 approximately
80 min after administration of the radioligand (Figure
8). This ratio is comparable to that of the less lipophilic
radioligands for imaging CB1, 6d and 6f.11
Conclusion
A novel series of 3-naphthoyl substituted AAIs with
high affinity for CB1 receptors and lipophilicity (cLogD7.4
)
Preadministration of 5 (1 mg/kg), a selective CB1
antagonist, significantly reduced the radioactivity of
[18F]9b in the mouse whole brain compared to the
radioactivity in a control experiment without pre-
administration (Figure 9). This reduction, defined as the
displaceable radioactivity, was significant and indicated
that some fraction of the total radioactivity in the brain
belongs to the specific binding of radioligand to CB1
in the range of 2.7-4.5 was synthesized. A SAR of the
studied set of AAIs, demonstrated that the affinity is
inversely related to the size of the molecular dipole. This
relationship was sustained even when substitution was
not at naphthyl carbon four, i.e., compound 9h.
Radioligand [18F]9b specifically labels CB1 receptors
in mouse brain. Separation of and in vivo studies with

Imaging the Neuronal Cannabinoid Receptor
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5819
the two enantiomers of [18F]9b demonstrated that only
one of the enantiomers [18F]9b-2 exhibits specific bind-
ing to CB1 receptors and therefore has potential as a
PET radioligand. These results suggest that the inactive
enantiomer [18F]9b-1 could provide a reasonable mea-
surement of the nondisplaceable binding of the active
enantiomer [18F]9b-2.
second component 9b-2 with a flow rate of 10 mL/min.
Fractions were collected, dried, and analyzed on an analytical
ChiralPak OD column. Specific rotation for 9b-2 is [R]26
)
D
+30.47 (c 0.105 in ethanol) and 9b-1 is [R]27D ) -59.99 (c 0.130
in ethanol).
3-(4-Bromonaphthoyl)-1-(N-methylpiperidin-2-ylmeth-
yl)indole (9c). 9c was prepared like that for 9a, using
4-bromonaphthoic acid30 (530.2 mg, 2.112 mmol), 6 mL SOCl2,
AlCl3 (500 mg, 3.75 mmol), and 10 (242.2 mg, 1.060 mmol).
The product was purified on silica with 3.75:1.75:0.5 hexane:
CHCl3:NEt3. Rf 0.19 in 6:4:1 hexane:CHCl3:NEt3. 9c was
recovered as an oily solid which was purified by washing with
hexane for a 67% yield. 1H NMR (300 MHz, CDCl3) δ 8.45 (m,
1H), 8.36 (d, J ) 8 Hz, 1H), 8.20 (d, J ) 8 Hz, 1H), 7.86 (d,
1H), 7.64 (td, 1H), 7.52 (m, 2H), 7.38 (m, 4H), 4.52 (dd, J ) 4,
14 Hz, 1H), 3.85 (dd, J ) 9, 14 Hz, 1H), 2.85 (d, J ) 12 Hz,
1H), 2.38 (s, 3H), 2.35 (m, 1H), 2.11 (dt, J ) 3, 12 Hz, 1H),
0.8-1.8 (m, 8H), Anal. (C26H25BrN2O) C, H, N, Br.
Experimental Section
Chemistry. Unless specifically indicated, all reagents were
purchased from Aldrich. All solvents used were ACS or HPLC
1
grade. H NMR spectra were recorded on a Bruker AM 300
(300 MHz) instrument; chemical shifts (δ) were recorded in
parts per million (ppm) downfield from TMS. High-perfor-
mance liquid chromatography (HPLC) analysis and purifica-
tion were performed with a Rheodyne 7725 injector, a Waters
515 HPLC pump, and an in-line Waters 2487 dual λ absorb-
ance detector (254 nm). HPLC chromatograms were recorded
by a Waters Pump Control Module connected to a HP computer
equipped with Waters Empower software program. Waters
Symmetry C-18 (3.5 µm, 4.6 mm × 75 mm), Chiracel OD (10
µm, 4.6 mm × 250 mm), and Waters Nova-Pak C-18 radial
compression (6 µm, 25 mm × 100 mm) columns were used in
the HPLC analyses and preparative separations. Analytical
thin-layer chromatography (TLC) was done on precoated plates
of silica gel (0.25 mm, F254, Alltech). Flash chromatography
was conducted using silica gel (230-400 mesh, Merck). Melting
points were determined on a Mel-Temp II apparatus. Elemen-
tal analyses were performed by Quantitative Technologies, Inc.
(QTI). The newly synthesized final compounds gave satisfac-
tory elemental analyses results (C, H, N, F (0.40%). The clog
D7.4 values were calculated using ACD/LogD Suite software
(Advanced Chemistry Development Inc., Toronto, Canada).
Compounds 8 (AM251),16 9j,18 and 9n18 were synthesized
using a procedure described elsewhere in the literature.
3-(4-Methynaphthoyl)-1-(N-methylpiperidin-2-ylmeth-
yl)indole (9a). 4-Methylnaphthoic acid (206.6 mg, 1.110
mmol) was heated in 1.5 mL SOCl2 at 70 °C for 2 h, evaporated
under vacuum, and used as is for the next step. AlCl3 (170.9
mg, 1.282 mmol) was stirred as a suspension in 2 mL CH2Cl2
for 15 min. The acid chloride was dissolved in 2 mL CH2Cl2,
added to the suspension of aluminum chloride and stirred for
0.5 h. 1-(N-Methylpiperidin-2-ylmethyl)indole16 (10) (129.6 mg,
0.567 mmol) in 1 mL CH2Cl2 was then added to the reaction
dropwise at RT. The reaction was then warmed to 40 °C for 1
h, worked up with K2CO3, and extracted into CH2Cl2. The
product was purified on silica with 3.75:1.75:0.5 hexane:CHCl3:
NEt3. Rf was 0.26 in 6:4:1 hexane:CHCl3:NEt3. An oily solid
was further purified by washing with hexane for an 82% yield
3-(4-Methoxynaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (9d). 9d was prepared like that for 9a, using
4-methoxynaphthoic acid (56.8 mg, 0.281 mmol), 0.5 mL SOCl2,
AlCl3 (64.1 mg, 0.481 mmol), 10 (31.9 mg, 0.140 mmol). The
reaction was purified on silica with 6:4:1 hexane:CHCl3:NEt3.
Rf 0.11 in 6:4:1 hexane:CHCl3:NEt3. 9d was recovered as an
oily solid which was purified by washing with hexane for a
1
77% yield. mp ) 143 °C, H NMR (300 MHz, CDCl3) δ 8.45
(m, 1H), 8.36 (d, J ) 8 Hz, 1H), 8.20 (d, J ) 8 Hz, 1H), 7.86 (d,
1H), 7.64 (td, 1H), 7.52 (m, 2H), 7.38 (m, 4H), 4.52 (dd, J ) 4,
14 Hz, 1H), 3.85 (dd, J ) 9, 14 Hz, 1H), 2.85 (d, J ) 12 Hz,
1H), 2.38 (s, 3H), 2.35 (m, 1H), 2.11 (dt, J ) 3, 12 Hz, 1H),
0.8-1.8 (m, 8H), Anal. (C27H28N2O2‚0.2EtOAc) C, H, N.
3-(4-Cyanonaphthoyl)-1-(N-methylpiperidin-2-ylmeth-
yl)indole (9f). 4-Cyanonaphthoic acid was made from
4-methylnaphthonitrile by a procedure from the cited litera-
ture31 by a radical bromination of the aryl methyl, then
conversion of the aryl bromide to an aldehyde by a nucleophilic
reaction of hexamine with the aryl bromide and a hydrolytic
workup. This aldehyde was converted to an acid with potas-
sium permanganate. 9f was prepared like that for 9a, using
the prepared 4-cyanonaphthoic acid (104.6 mg, 0.530 mmol),
2 mL SOCl2, AlCl3 (223.5 mg, 1.676 mmol), and 10 (73.9 mg,
0.320 mmol). Product was purified on silica with 6:4:1 then
5:5:1 hexane:CHCl3:NEt3 solvent system. Rf 0.16 in 6:4:1
hexane:CHCl3:NEt3. 9f was recovered as an oily solid, which
was purified by washing with hexane for a 35% yield. mp )
125 (dec) °C, 1H NMR (300 MHz, CDCl3) δ 8.42 (m, 1H), 8.36
(d, J ) 8 Hz, 1H), 8.41 (d, J ) 8 Hz, 1H), 8.12 (d, 1H), 8.00 (d,
J ) 7 Hz, 1H), 7.74 (td, J ) 7, 1 Hz, 1H), 7.67 (d, J ) 9 Hz,
1H), 7.59 (td, J ) 1, 8 Hz, 1H), 7.40 (m, 3H), 7.33 (s, 1H), 4.51
(dd, J ) 4, 14 Hz, 1H), 3.87 (dd, J ) 9, 14 Hz, 1H), 2.83 (dt, J
) 14, 7 Hz, 2H), 2.39 (s, 3H), 2.36 (m, 1H), 2.11 (dt, J ) 3, 12
Hz, 1H), 0.8-1.8 (m, 11H), Anal. (C27H25N3O‚0.2EtOAc) C, H,
N.
3-(2-Methylnaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (9i). 9i was prepared like that for 9a, using
2-methylnaphthoic acid (216.0 mg, 1.160 mmol), 2 mL SOCl2,
AlCl3 (223.5 mg, 1.676 mmol), and 5 (73.9 mg, 0.320 mmol).
The final step was warmed to 70 °C for 3 h. 9i was recovered
as an oily solid, which was purified by crystallization with
hexane for a 65% yield. mp ) 160 °C, 1H NMR (300 MHz,
CDCl3) δ 8.69 (b, 1H), 7.84 (d, 2H), 7.72 (t, 1H), 7.53-7.31 (m,
6H), 7.05 (b, 1H), 4.47 (bd, J ) 14 Hz, 1H), 3.77 (b, 1H), 2.80
(b, 1H), 2.41 (s, 3H), 2.35 (bs, 4H), 2.08, Anal. (C27H28N2O‚
0.2EtOAc) C, H, N.
1
(184.5 mg 0.465 mmol). mp ) 123-4 °C, H NMR (300 MHz,
CDCl3) δ 8.45 (m, 1H), 8.27 (d, J ) 8 Hz, 1H), 8.19 (d, J ) 8
Hz, 1H), 7.55 (m, 3H), 7.37 (m, 5H), 4.53 (dd, J ) 4, 14 Hz,
1H), 3.87 (dd, J ) 9, 14 Hz, 1H), 2.85 (d, J ) 12 Hz, 1H), 2.78
(s, 3H) 2.41 (s, 3H), 2.37 (m, 1H), 2.12 (dt, J ) 3, 12 Hz, 1H),
0.8-1.8 (m, 11H), Anal. (C27H28N2O‚0.09EtOAc) C, H, N.
3-(4-Fluoronaphthoyl)-1-(N-methylpiperidin-2-ylmeth-
yl)indole (9b). 9b was prepared like that for 9a, using
4-flouronaphthoic acid (97.0 mg, 0.510 mmol), 0.5 mL SOCl2,
AlCl3 (78.3 mg, 0.587 mmol), and 10 (51.4 mg, 0.225 mmol).
The product was purified on silica with 3.25:1.75:0.5 hexane:
CHCl3:NEt3. Rf 0.19 in 6:4:1 hexane:CHCl3:NEt3. An oily solid
was recovered and washed with hexane for a 76% yield. mp )
1
137 °C, H NMR (300 MHz, CDCl3) δ 8.56 (m, 1H), 8.27 (d, J
) 8 Hz, 1H), 8.19 (d, J ) 8 Hz, 1H), 7.63 (m, 3H), 7.37 (m,
4H), 7.19 (t, 1H), 4.53 (dd, J ) 4, 14 Hz, 1H), 3.87 (dd, J ) 9,
14 Hz, 1H), 2.85 (d, J ) 12 Hz, 1H), 2.41 (s, 3H), 2.37 (m, 1H),
2.12 (dt, J ) 3, 12 Hz, 1H), 0.8-1.8 (m, 11H), Anal.
(C26H25N2OF) C, H, N, F.
Product was separated into its two enantiomers on a
ChiralPak OD HPLC column (semipreparative, 250 × 20 mm)
75:25:0.2 hexane:IPA:diethylamine eluent. Product was washed
with hexane and dried under vacuum. Retention times were
14.8 min for the first component 9b-1 and 25.6 min for the
3-(4-Benzyloxynaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (19). 4-Benzyloxy-1-bromonaphthalene (16)32
(200.1 mg, 0.6389 mmol) in 5 mL tetrahydrofuran was added
to a solution of n-butyllithium (440 µL, 1.6 M) in 5 mL at -15
°C for 10 min, then stirred for another 30 min. Dried CO2 was
bubbled through the reaction for another 30 min. The product
was acidified with NaHSO4 and extracted into ether. 4-Ben-
zyloxynaphthoic acid (17) was recovered pure by crystallization
with hexane and ether in 45% yield. 17 (247.6 mg, 0.8897
mmol) was dissolved in 0.2 mL of thionyl chloride and 7 µL

5820 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Willis et al.
DMF at room temperature for 2 h, then dried under vacuum.
The resulting acid chloride and 425 mg of the ionic liquid
1-butyl-3-methylimidazolium hexafluorophosphate ([bmin]-
[PF6]) were heated to 150 °C in a sealed vial, and the product
was purified by washing the ionic liquid with CHCl3 and 1 N
Na2CO3. 19 (360.4 mg, 0.7375 mmol) was obtained pure by
flash chromatography with the solvent gradient 8:2:1 to 6:4:1
hexane:CHCl3:NEt3. Rf 0.15. 1H NMR (300 MHz, CDCl3) δ 8.48
(m, 1H), 8.32 (m, 1H), 8.29 (m, 1H), 7.53-7.31(m, 7H), 6.69
(d, J ) 8 Hz, 1H), 4.60 (dd, J ) 4, 14 Hz, 1H), 3.94 (dd, J ) 9,
14 Hz, 1H), 2.94 (d, J ) 12 Hz, 1H), 2.49 (m, 1H), 2.44 (s, 3H),
2.20 (dt, J ) 5, 10 Hz, 1H), 0.98-1.71 (m, 11H)
3-(4-Hydroxynaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (9e). 19 (205.0 mg, 0.4195 mmol) was stirred
under argon with 417.6 mg 10% Pd/C and 279.4 mg am-
monium formate. After being stirred for 1 h, the mixture was
filtered, diluted with EtOAc, and washed with (aq) K2CO3 and
3× DI water. 9e was obtained as pale yellow oil in 78% yield.
1H NMR (300 MHz, CDCl3) δ 8.47 (m, 1H), 8.32 (m, 2H), 7.52
(d, J ) 8 Hz, 1H), 7.28-7.49 (m, 6H), 6.72 (d, J ) 8 Hz, 1H),
4.57 (dd, J ) 4, 14 Hz, 1H), 3.92 (dd, J ) 9, 14 Hz, 1H), 2.91
(d, J ) 12 Hz, 1H), 2.44 (s, 4H), 2.17 (m, 1H), 0.75-1.72 (m,
9H). Anal. (C26H26N2O2‚H2O), C, H, N.
3-(Naphthoyl)-1-(N-methylpiperidin-2-ylmethyl)inda-
zole (9m). 3-Naphthoyl indazole was made by a 1,3 dipolar
cycloaddition of naphthoyldiazomethane and the benzyne
equivalent, 2-diazobenzene carboxylic acid; physical data
matched the literature.33 Wet 2-diazocarboxylic acid was used
in the first step of the reaction as the reagent becomes
explosive upon drying. The indazole (47.9 mg, 0.176 mmol),
N-methyl-2-chloromethylpiperidine (42.9 mg, 0.291 mmol), and
NaH (11 mg) were stirred in anhydrous DMF at room
temperature for 2 days, evaporated under vacuum, worked up
with saturated NaHCO3, and extracted into ether. The residue
was purified by flash chromatography using 8:2:1 hexane:
CHCl3:NEt3, yielding 9m (17.3 mg, 0.0451 mmol) with an Rf
of 0.30 in 6:4:1 hexane:CHCl3:NEt3. mp ) 98-100 1H NMR
(300 MHz, CDCl3) δ 8.44 (d, 8 Hz, 1H), 8.34 (m, 1H), 8.03 (d,
8 Hz, 1H), 7.86-7.99 (m, 2H), 7.45-7.61 (m, 5H), 7.38 (m, 1H),
4.79 (dd, 5, 14 Hz, 1H), 4.32 (dd, 9, 14 Hz, 1H), 2.88 (d, 11 Hz,
1H), 2.66 (m, 1H), 2.41 (s, 3H), 2.19 (m, 1H), 0.8-1.8 (m, 11H).
Anal. (C25H25N3O‚0.5EtOAc) C, H, N.
10.17 mmol), imidazole (2.788 g, 40.95 mmol), triisopropylsilyl
chloride (3.091 g, 20.51 mmol), and 2 mL DMF. The reaction
was complete after 2 h slowly heating to 85 °C. After dilution
in hexane and washing with DI water, 14 was recovered pure
in 99% yield. 2.5 M n-butyllithium (1.0 mL, 2.5 mmol) was
diluted in 10 mL THF. 14 (1.001 g, 2.458 mmol) and 15 mL
anhydrous THF were added to the solution of n-butyllithium
at -78 °C dropwise over 30 min. The mixture was stirred for
3 h at -78 °C and then dried. CO2 was allowed to condense
over the reaction. Reaction was continued for 15 min with CO2
flow, warmed to room temperature, and diluted with ether,
and then 1 M HCl was added until acidic. The reaction mixture
was extracted with ether 2 × 20 mL, brine added, and the
mixture extracted again with 20 mL ether, dried with mag-
nesium sulfate, filtered, and evaporated under vacuum. Prod-
uct (15) (451 mg, 1.21 mmol) was purified with flash chroma-
tography using 70:30 hexane:EtOAc. 1H NMR (300 MHz,
CDCl3) δ 10.7(vb, 1H), 9.13 (dd, 1.3, 9.0 Hz, 1H), 8.31 (d, 7.5
Hz, 1H), 8.13 (d, 8.1 Hz, 1H) 7.54-7.68 (m, 2H), 7.46 (d, 7.5
Hz, 1H), 7.27-7.38 (m, 5H), 4.55(s, 2H), 3.87(t, 7.2 Hz, 2H),
3.48 (t, 7.2 Hz, 2H).
3-(4-(2-Triisopropylsiloxyethyl)naphthoyl)-1-(N-meth-
ylpiperidin-2-ylmethyl)indole (21). 21 was obtained by the
same method as 19 using 15 (248.9 mg, 0.6681 mmol) and 0.14
mL oxalyl chloride. Then 10 (173.1 mg, 0.7578 mmol) and
([bmin][PF6])23 (228 mg) were heated to 100 °C for 14 h. 21
(124.4 mg, 0.2134 mmol) was isolated pure by flash chroma-
tography using 8:2:1 hexane:CHCl3:NEt3. 1H NMR (300 MHz,
CDCl3) δ 8.48 (dd, 3, 6 Hz, 1H), 8.25 (d, 8 Hz, 1H), 8.11 (d, 8
Hz, 1H) 7.2-7.6 (m, 13H), 4.58 (s, 2H), 4.52 (m, 1H) 3.89 (t, 7
Hz, 2H), 3.82 (dd, 9, 14 Hz, 1H), 3.48 (t, 7 Hz, 2H), 2.83 (d,
12H, 1H), 2.39 (s, 3H), 2.45 (b, 1H), 2.09 (dt, 3, 11 Hz, 1H),
0.8-1.7 (m, 9H).
3-(4-(2-Hydroxyethyl)naphthoyl)-1-(N-methylpiperidin-
2-ylmethyl)indole (9g). 21 (124.4 mg, 0.2134 mmol) was
stirred in anhydrous THF with tetrabutylammonium fluoride
(68 mg, 0.26 mmol) at 100 °C for 2 h. The pure product (60.4
mg, 0.103 mmol) was obtained by flash chromatography with
3:7:1 hexane:CHCl3:NEt3 solvent system. 1H NMR (300 MHz,
CDCl3) δ 8.31 (dd, 3, 6 Hz, 1H), 8.08 (d, 8 Hz, 1H), 7.97 (d, 8
Hz, 1H) 7.2-7.5 (m, 8H), 4.37 (d, 8 Hz, 1H), 3.90 (t, 2H) 3.68
(m, 1H), 3.36 (d, 14 Hz, 1H), 3.27 (t, 7 Hz, 2H), 2.68 (d, 12H,
1H), 2.24 (s, 3H), 2.45 (b, 1H), 2.09 (t, 11 Hz, 1H), 0.6-1.6 (m,
9H). Anal. (C28H30N2O2‚0.4EtOAc) C, H, N.
3-(6-Benzyloxynaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (20). 6-Hydroxynaphthoic acid was alkylated
with benzyl bromide and KOH in DMF at 100 °C for 2 h, giving
an unoptimized yield of 13% of 6-benzyloxynaphthoic acid
benzyl ester. 6-Benzyloxynaphthoic acid benzyl ester (246 mg,
0.714 mmol) was refluxed for 32 h in 5 mL THF and 1 g NaOH
in 5 mL water. The product was diluted with 25 mL water,
washed with 25 mL ether two times, acidified with HCl, and
extracted into ether. 6-Benzyloxynaphthoic acid (18) (135.7 mg,
0.488 mmol) was dried and determined to be pure. 20 was
obtained by the same method as for 19, using 18 (124.1 mg,
0.4459 mmol), 0.5 mL thionyl chloride, and then ([bmin][PF6])
(300 mg) and 10 (103.1 mg, 0.4514 mmol). The mixture was
heated to 155 °C for 30 min in a sealed vessel. 20 (42.2 mg,
0.1098 mmol) was purified by flash chromatography 8:2:1
hexane:CHCl3:NEt3. Rf 0.09. 1H NMR (300 MHz, CDCl3) δ 8.44
(m, 2H), 8.33 (m, 1H), 7.65 (d, 8 Hz, 1H) 7.3-7.6 (m, 11H),
5.33 (s, 2H), 4.53 (dd, 4, 14 Hz, 1H), 3.85 (dd, 9, 14 Hz, 1H),
2.85 (d, 12H, 1H), 2.41 (s, 3H), 2.36 (b, 1H), 2.12 (dt, 3, 12 Hz,
1H), 0.8-1.7 (m, 9H).
3-(Carboxy-1-aminonaphthyl)-1-(N-methylpiperidin-2-
ylmethyl)indole (9k). 3-(Carboxylic acid)-1-(N-methylpip-
eridin-2-ylmethyl)indole (92.0 mg, 0.332 mmol) was stirred at
room-temperature overnight with 74 µL thionyl chloride and
1 mL anhydrous CHCl3, evaporated, and stirred with 1 mL
methylene chloride, 60 µL NEt3, and 1-aminonaphthalene (47
mg, 0.33 mmol) for 1 h. Reaction was purified by silica with
the solvents 98:2 EtOAc:NEt3, then 90:10 EtOAc:NEt3. Product
was crystalized from warm EtOAc as brown crystals in 52%
1
yield. mp ) 165-6 °C, H NMR (300 MHz, CDCl3) δ 8.0-8.2
(m, 4H), 7.91 (dd, 1, 8 Hz, 1H), 7.87 (s, 1H), 7.74 (d, 8 Hz, 1H),
7.4-7.6 (m, 4H), 7.4 (m, 2H), 4.61 (dd, 4, 13 Hz, 1H), 4.00 (dd,
9, 14 Hz, 1H), 2.89 (d, 11 Hz, 1H), 2.50 (s, 3H), 2.44 (m, 1H),
2.18 (m, 1H), 1.0-1.8 (m, 11H). Anal. (C26H27N3O‚0.1EtOAc)
C, H, N.
3-(Carboxy-N-decahydroquinoline)-1-(N-methyl-
piperidin-2-ylmethyl)indole (9l). 3-(Carboxylic acid)-1-(N-
methylpiperidin-2-ylmethyl)indole (112.7 mg, 0.412 mmol) was
stirred at room-temperature overnight with 92 µL thionyl
chloride and 1 mL methylene chloride, evaporated, and stirred
with 1 mL methylene chloride, 62 µL NEt3, and decahydro-
quinoline (55.4 mg, 0.399 mmol) for 1 h. Reaction was purified
by silica with the solvents 98:2 EtOAc:NEt3, then 90:10 EtOAc:
1
NEt3. Product was recovered as a thick oil in 52% yield. H
3-(6-Hydroxynaphthoyl)-1-(N-methylpiperidin-2-yl-
methyl)indole (9h). 9h was obtained by the same method
as for 9e, using 20 (31.0 mg, 0.0806 mmol), 64 mg 10% Pd/C,
and 42 mg ammonium formate for 0.5 h. The product was
recovered as an oily solid in 90% yield. TLC conditions 5:5:1
hexane:CHCl3:NEt3 Rf 0.08. 1H NMR (300 MHz, CDCl3) δ 8.46
(dd, 3, 6 Hz, 1H), 8.14 (d, 9 Hz, 1H), 7.84 (d, 8 Hz, 1H) 7.2-
7.6 (m, 13H), 4.52 (dd, 4,14 Hz, 1H), 3.86 (dd, 9, 14 Hz, 1H),
2.57 (d, 12H, 1H), 2.40 (s, 3H), 2.35 (b, 1H), 2.09 (dt, 3,12 Hz,
NMR (300 MHz, CDCl3) δ 7.71 (dd, 5, 7 Hz, 1H), 7.44 (d, 2
Hz, 1H), 7.35 (d, 8 Hz, 1H), 7.15-7.25 (m, 2H), 4.55 (dd, 4, 14
Hz, 1H), 3.91 (dd, 9, 14 Hz, 1H), 3.4-3.8 (m, 3H), 2.89 (d, 11
Hz, 1H), 2.48 (s, 3H), 2.37 (m, 2H), 2.15 (dt, 4, 11 Hz, 1H),
0.8-1.9 (m, 24H). Anal. (C25H35N3O) C, H, N.
4-(2-Triisopropylsiloxyethyl)naphthoic Acid (15). 4-(2-
Triisopropylsiloxyethyl)-1-bromonaphthalene (14) was pre-
pared from 4-(2-hydroxyethyl)-1-bromonaphthalene34 (2.554 g,

Imaging the Neuronal Cannabinoid Receptor
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5821
1H), 0.8-1.6 (m, 9H). Anal. (C26H26N2O2‚1.5H2O‚1.4HCOOH)
efficiency). All assays were carried out in quadruplicate. The
Ki values were calculated by the Cheng-Prusoff equation
based on the measured IC50 values and Kd ) 0.5 nM for the
binding of [3H]2. The Kd value for [3H]2 was determined in
the same conditions using serial dilutions of [3H]2 with
nonradiactive 2.
C, H, N.
Radiochemistry. 3-(4-[18F]flouronaphthoyl)-1-(N-methyl-
piperidin-2-ylmethyl)indole ([18F]9b-1, [18F]9b-2). An aqueous
solution of [18F]fluoride, Kryptofix 222 (23 mg), and K2CO3 (4.3
mg) was added to a 10 mL vessel. Water was removed by
repeated azeotropic evaporations with MeCN at 120 °C and a
stream of argon. 4.0 mg of the nitro precursor 9n and 0.8 mL
DMSO were added to the vessel and heated to 160 °C for 18
min. The reaction was cooled, dissolved in 1 mL of the mobile
phase, injected onto a 7.8 × 300 mm Symmetry reverse phase
HPLC column, and eluted with 300:700:2 THF:H2O:TFA at a
flow rate of 4.5 mL/min. The radioactive peak was collected,
evaporated, diluted with 100 µL mobile phase, and then
injected on a 4.6 × 250 mm Chiralcel OD column with a 75:
25:0.2 hexane:IPA:diethylamine mobile phase at 1.1 mL/min.
This procedure separated the enantiomers with nearly 3 min
of baseline resolution between the two enantiomers. The two
enantiomers [18F]9b-1 and [18F]9b-2 came out at 7.4 and 11.7
min with the second peak being the active isomer.
Molecular Modeling. The dipole moment was determined
using the Chemdraw3D software package. The minimum
energy conformation of compound 8 was found by several
conformational energy searches. An initial conformational
energy search was performed, then Torsion 1 (indole C2, C3,
carbonyl carbon, and naphthyl C1) (numbering found in Figure
5) was rotated 180° and a conformational energy search was
repeated. The same was done with Torsion 2 (indole C3,
carbonyl carbon, naphthyl C1, and C2). The minimum energy
conformation for Torsion 1 and Torsion 2 was used to find the
minimum energy conformation for Torsion 3 (indole C2, N1,
tail CH2, and CH). Torsion 3 was rotated 60°, a conformational
energy search was performed and the process was repeated
three times. The lowest energy conformation of 8 was modified
to determine the rest of the ligands in this set.
Distribution of Radioactivity in the Mouse Brain. All
experiments with research animals were approved by the
Animal Care and Use Committee of the NIDA Intramural
Research Program. Male CD-1 mice weighing 25-35 g (Charles
River Laboratories, Wilmington, MA) received an injection of
a known amount of the radiotracer (5-7 µCi in 200 µL saline,
specific activity 2400-3200 Ci/mmol) into the lateral tail vein.
The mass of injected radioligands ranged from 0.06 to 0.07
nmol/kg. In the blocking studies, animals were administered
with nonradioactive 5 (1 mg/kg, sc) 15 min before radioligand
injection. Animals were euthanized by cervical dislocation at
predetermined intervals (n ) 3 for each time point) after
injection of the radioligand. The brains were removed quickly
from the skull and dissected. Samples were weighed and their
radioactivity assayed in a γ counter (Cobra, Packard Instru-
ments, Meriden, CT). The accumulation of radioactivity in the
brain samples was calculated as the amount of radioactivity
per gram of tissue and expressed as the percentage of the
administered radiotracer dose per kg of body weight.
Acknowledgment. The authors thank Drs. Elliot
Stein, Amy H. Newman, Alex Hoffman, and Reeti
Katoch-Rouse for their helpful discussions, and Ms.
Mary Pfeiffer for her kind editorial assistance.
Supporting Information Available: Elemental analysis
results and 1H NMR spectra for pure final compounds is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Howlett, A. C.; Barth, F.; Bonner, T. I.; Cabral, G.; Casellas, P.;
Devane, W. A.; Felder, C. C.; Herkenham, M.; Mackie, K.;
Martin, B. R.; Mechoulam, R.; Pertwee, R. G. International
Union of Pharmacology. XXVII. Classification of cannabinoid
receptors. Pharmacol. Rev. 2002, 54, 161-202.
(2) Cabral, G. A. Marijuana and the immune system; Humana
Press: Totowa, NJ, 1999.
(3) Martin, B. R.; Mechoulam, R.; Razdan, R. K. Discovery and
characterization of endogenous cannabinoids. Life Sci. 1999, 65,
573-595.
(4) Makriyannis, A.; Goutopoulos, A. Cannabinergics: old and new
theraputic possibilities. In Drug discovery strategies and meth-
ods; Makriyannis, A., Biegel, D., Eds.; Marcel Dekker: New
York, 2004; pp 89-128.
(5) Le Foll, B.; Goldberg, S. R. Cannabinoid CB1 Receptor Antago-
nists as Promising New Medications for Drug Dependence. J.
Pharmacol. Exp. Ther. 2005, 312, 875-883.
(6) Levitt, M. Cannabinoids as antiemetics in cancer chemotherapy.
In Cannabinoids as therapeutic agents; Mechoulam, R., Ed.; CRC
Press: Boca Raton, FL, 1986; pp 73-79.
(7) Arnone, M.; Maruani, J.; Chaperon, F. Selective inhibition of
sucrose and ethanol intake by SR141716a, an antagonist of
central cannabinoid receptors. Psychopharmacology 1998, 132,
104-106.
(8) Sieradzan, K. A.; Fox, S. H.; Hill, M.; Dick, J. P.; Crossman, A.
R.; Brotchie, J. M. Cannabinoids reduce levodopa-induced dys-
kinesia in Parkinson’s disease: A pilot study. Neurology 2001,
57, 2108-2111.
(9) Mathews, W. B.; Scheffel, U.; Finley, P.; Ravert, H. T.; Frank,
R. A.; Rinaldi-Carmona, M.; Barth, F.; Dannals, R. F. Biodis-
tribution of [18F] SR144385 and [18F] SR 147963: Selective
radioligands for positron emission tomographic studies of brain
cannabinoid receptors. Nucl. Med. Biol. 2000, 27, 757-762.
(10) Mathews, W. B.; Scheffel, U.; Rauseo, P. A.; Ravert, H. T.; Frank,
R. A.; Ellames, G. J.; Herbert, J. M.; Barth, F.; Rinaldi-Carmona,
M.; and Dannals, R. F. Carbon-11 labeled radioligands for
imaging brain cannabinoid receptors. Nucl. Med. Biol. 2002, 29,
671-677.
(11) Gatley, S. J.; Lan, R.; Volkow, N. D.; Pappas, N.; King, P.; Wong,
C. T.; Gifford, A. N.; Pyatt, B.; Dewey, S. L.; Makriyannis, A.
Imaging the brain marijuana receptor: development of a radio-
ligand that binds to cannabinoid CB1 receptors in vivo. J.
Neurochem. 1998, 70, 417-423.
(12) Gatley, S. J.; Gifford, A. N.; Volkow, N. D.; Lan, R.; Makriyannis,
A. 123I-labeled AM251: a radioiodinated ligand which binds in
vivo to mouse brain cannabinoid CB1 receptors. Eur. J. Phar-
macol. 1996, 307, 331-338.
(13) Cosenza, M.; Gifford, A. N.; Gatley, S. J.; Pyratt, B.; Liu, Q.;
Makriyannis, A.; Volkow, N. D. Locomotor activity and oc-
cupancy of brain cannabinoid CB1 receptors by the antagonist/
inverse agonist AM281. Synapse 2000, 38, 477-482.
(14) Katoch-Rouse, R.; Pavlova, O. A.; Caulder, T.; Hoffman, A. F.;
Mukhin, A. G.; Horti, A. G. Synthesis structure-activity rela-
tionship, and evaluation of SR141716 analogues: Development
of central cannabinoid receptor ligands with lower lipophilicity.
J. Med. Chem. 2003, 46, 642-645.
(15) Lan, R.; Gatley, S. J.; Makriyannis, A. Preparation of iodine-
123 labeled AM251: a potential SPECT radioligand for the brain
cannabinoid CB1 receptors. J. Labelled Compd. Radiopharm.
1996, 38, 875-881.
(16) D′Ambra, T. E.; Eissenstat, M. A.; Abt, J.; Ackerman, J. H.;
Bacon, E. R.; Bell, M. R.; Carabateas, P. M.; Josef, K. A.; Kumar,
V.; Weaver, J. D.; Arnold, R.; Casiano, F. M.; Chippari, S. M.;
Haycock, D. A.; Kuster, J. E.; Luttinger, A.; Stevenson, J. I.;
Ward, S. J.; Hill, W. A.; Khanolkar, A.; Makriyannis, A.
C-Attached aminoalkylindoles: potent cannabinoid mimetics.
Bioorg. Med. Chem. Lett. 1996, 6, 17-22.
Radioligand in Vitro Binding Assays. Competition bind-
ing assays were carried out with [3H]2 (168 Ci/mmol, Perkin-
Elmer, MA) and rat cerebellum membrane fraction. The
membrane fraction was prepared from frozen Sprague-
Dawley rat cerebellum tissue (Pel-Freez Biologicals) as de-
scribed previously.14 On the day of assay, the pellet was
resuspended in ice-cold assay buffer (pH ) 7.4), containing
50 mM TrisHCl, 3 mM MgCl2, 1 mM EDTA, and 0.5% fat free
bovine serum albumin.
Aliquots of membrane fraction corresponding to 0.6 mg of
cerebellum tissue were incubated at 30 °C for 1 h with 0.5
nM [3H]2 and serial dilutions of test drugs in a total incu-
bation volume of 0.8 mL. All assays were performed in glass
test tubes. Nonspecific binding was determined in the presence
of 0.5 mM of 2 (Tocris Cookson Inc., UK). Binding was
terminated by vacuum filtration through GF/C filters pre-
treated in the rinse buffer. The rinse buffer used was 50 mM
Tris-HCl containing 0.5% of nonfat dry milk to reduce non-
specific binding to filter material. Radioactivity was measured
using a liquid scintillation counter (Beckman LS 6000, 37%

5822 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Willis et al.
(17) Tetko, I. V.; Kovalishyn, V. V.; Livingstone, D. J. Volume
learning algorithm artificial neural networks for 3D QSAR
studies. J. Med. Chem. 2001, 44, 2411-2420.
(18) Deng, H. Design and synthesis of selective cannabinoid receptor
ligands: Aminoalkylindole and other heterocyclic analogs. Ph.D.
Dissertation, University of Connecticut, Storrs, CT, 2000.
(19) Makriyannis, A.; Deng, H. Cannabimimetic indole derivatives.
WO 2001/028557.
(20) Wilson, A. A.; Jin, L.; Garcia, A.; DaSilva, J. N.; Houle, S. An
admonition when measuring the lipophilicity of radiotracers
using counting techniques. Appl. Radiat. Isot. 2001, 54, 203-
208.
(21) Shim, J.; Collantes, E. R.; Welsh, W. J. Three-dimensional
quantitative structure-activity relationship study of the
cannabimimetic (aminoalkyl)indoles using comparative molec-
ular field analysis. J. Med. Chem. 1998, 41, 4521-4532.
(22) Kandemirli, F. Structure-activity relationships investigation in
a mixed series of cannabiniods. Arzneim.-Forsch./Drug Res.
2002, 52, 731-739.
(23) Earle, M. J.; McCormac, P. B.; Seddon, K. R. The first high yield
route to a pharmaceutical in a room-temperature ionic liquid.
Green Chem. 2000, 2, 261-262.
(24) Makriyannis, A.; Liu, Q. Heteroindanes: a new class of potent
cannabimimetic ligands. WO 2003/035005.
(26) Carren˜o, M. C.; Ruano, J. L.; Sanz, G.; Toledo, M. A.; Urbano,
A. N-bromosuccinimide as a regioselective nuclear monobromi-
nating reagent for phenols and naphthols. Synlett 1997, 1241-
1242.
(27) Berridge, M. S.; Crouzel, C.; Comar, D. Aromatic fluorination
with n.c.a. 18F-fluoride: a comparative study. J. Labelled
Compd. Radiopharm. 1985, 22, 687-691.
(28) Herkenham, M.; Lynn, A. B.; Little, M. D.; Johnson, M. R.;
Melvin, L. S.; DeCosta, B. R.; Rice, K. C. Cannabinoid receptor
localization in brain. Proc. Natl. Acad. Sci. U.S.A. 1990, 87,
1932-1936.
(29) Herkenham, M.; Lynn, A. B.; Little, M. D.; Johnson, M. R.;
Melvin, L. S.; DeCosta, B. R.; Rice, K. C. Characterization and
localization of cannabinoid receptors in rat brain: a quantitative
in vitro autoradiographic study. J. Neurosci. 1991, 11, 563-583.
(30) Blicke, F. F.; Gomberg, M. Triphenyhnethyl. XXXIII.
Quinoidation in the triarylmethyls. J. Am. Chem. Soc. 1923, 45,
1765.
(31) Dewar, M. J.; Grisdale, P. J. Substituent Effects. The Preperation
of a series of substituted 1-Naphthoic acids. J. Chem. Soc. 1962,
3541-3546.
(32) Resch, J. F.; Lehr, G. S.; Wischik, C. M. Design and synthesis
of a potential affinity/cleaving reagent for beta-pleated sheet
protein structures. Bioorg. Med. Chem. Lett. 1991, 1, 519-522.
(33) Ried, W.; Scho¨n, M. Abfangreaktionen von Dehydrobenzol mit
Diazoketonen. Liebigs Ann. Chem. 1965, 141-144.
(34) Olah, G. A.; Singh, B. P.; Liang, G. Stable carbocations. 255.
R-Ethylenehaloarenium ions. J. Org. Chem. 1984, 49, 2922-2925.
(25) Leftheris, K.; Zhao, R.; Chen, B.; Kiener, P.; Wu, H.; Pandit, C.;
Wrobleski, S.; Chen, P.; Hynes, J.; Longphre, M.; Norris, D. J.;
Spergel, S.; Tokarski, J. Cannabinoid receptor modulators, their
processes of preparation, and use of cannabinoid receptor
modulators for treating respiratory and non-respiratory diseases.
WO 2001/058869.
JM0502743