Peripherally selective diphenyl purine antagonist of the CB1 Receptor

Article abstract of DOI:10.1021/jm401129n

Antagonists of the CB1 receptor can be useful in the treatment of several important disorders. However, to date, the only clinically approved CB1 receptor antagonist, rimonabant, was withdrawn because of adverse central nervous system (CNS)-related side effects. Since rimonabant's withdrawal, several groups are pursuing peripherally selective CB1 antagonists. These compounds are expected to be devoid of undesirable CNS-related effects but maintain efficacy through antagonism of peripherally expressed CB1 receptors. Reported here are our latest results toward the development of a peripherally selective analog of the diphenyl purine CB1 antagonist otenabant 1. Compound 9 (N-{1-[8-(2-chlorophenyl) -9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-yl}pentanamide) is a potent, orally absorbed antagonist of the CB1 receptor that is >50-fold selective for CB1 over CB2, highly selective for the periphery in a rodent model, and without efficacy in a series of in vivo assays designed to evaluate its ability to mitigate the central effects of Δ⁹-tetrahydrocannabinol through the CB1 receptor.

Full text of DOI:10.1021/jm401129n

Article
pubs.acs.org/jmc
Peripherally Selective Diphenyl Purine Antagonist of the CB1
Receptor
Alan Fulp, Katherine Bortoff, Yanan Zhang, Rodney Snyder, Tim Fennell, Julie A. Marusich,
Jenny L. Wiley, Herbert Seltzman, and Rangan Maitra*
Discovery Sciences, Research Triangle Institute, 3040 Cornwallis Road, P.O. Box 12194, Research Triangle Park, North Carolina
27709, United States
S
* Supporting Information
ABSTRACT: Antagonists of the CB1 receptor can be useful
in the treatment of several important disorders. However, to
date, the only clinically approved CB1 receptor antagonist,
rimonabant, was withdrawn because of adverse central nervous
system (CNS)-related side effects. Since rimonabant’s with-
drawal, several groups are pursuing peripherally selective CB1
antagonists. These compounds are expected to be devoid of
undesirable CNS-related effects but maintain efficacy through
antagonism of peripherally expressed CB1 receptors. Reported here are our latest results toward the development of a
peripherally selective analog of the diphenyl purine CB1 antagonist otenabant 1. Compound 9 (N-{1-[8-(2-chlorophenyl)-9-(4-
chlorophenyl)-9H-purin-6-yl]piperidin-4-yl}pentanamide) is a potent, orally absorbed antagonist of the CB1 receptor that is
>50-fold selective for CB1 over CB2, highly selective for the periphery in a rodent model, and without efficacy in a series of in
vivo assays designed to evaluate its ability to mitigate the central effects of Δ9-tetrahydrocannabinol through the CB1 receptor.
INTRODUCTION
■
The cannabinoid receptors CB1 and CB2 are G-protein-
coupled receptors (GPCRs) that primarily function as
activators of inhibitory G proteins (Gi/o).1 CB1 receptors are
expressed widely throughout the body; however, they are most
heavily expressed in the central nervous system (CNS). CB2
receptors are found most abundantly in the periphery. Both
Figure 1. Otenabant and a peripherally selective analogue.
receptors belong to an endocannabinoid system (ECS) that
regulates many physiological systems and can be targeted to
treat several serious disorders. Currently, various components
of the ECS are under evaluation for the treatment of pain,
inflammation, obesity, liver disease, and diabetes.2,3
peripherally selective. In rodent studies, compounds that have
<5% brain penetration in unperfused tissue samples are
considered peripherally restricted because the volume of
blood in a rodent brain accounts for ∼3−6% of its total
weight.7,8 Efforts are underway to produce peripherally
restricted CB1 antagonists are ongoing.9 Our group has
focused on 1 as a starting point for further optimization.10
Despite being nontissue selective, 1 has several properties that
could be optimized to produce peripherally selective com-
pounds, including a formula weight of 510, a topological polar
surface area (TPSA) of 102, and 3 H-bond donors.11 However,
intramolecular H bonding effectively lowers the polarity of 1,
allowing for penetration into the CNS.11 In our initial
publication, we reported several diphenyl purines with formula
weights and TPSA similar to 1 but lacking the ability to form
intramolecular H bonds. To that end, we were able to discover
2, an analogue of 1 that has a formula weight of 593 and a
TPSA 101.10 However, unlike 1, 2 is a peripherally selective
For example, the antagonism of CB1 has been investigated
for the treatment of metabolic disorders such as diabetes and
obesity. Past translational research in this regard led to the
clinical approval of rimonabant (SR141716A) for weight loss in
Europe. Rimonabant is a potent and selective CB1 receptor
inverse agonist/antagonist. Regrettably, rimonabant produced
serious side effects such as depression, anxiety, and suicidal
ideation in a subset of patients. These CNS-related side effects
led to the withdrawal of rimonabant from European markets,
and the drug was not approved for use in the United States.4,5
Several other CB1 antagonists, such as otenabant (1) (Figure
1), taranabant, and ibipinabant, were pulled from clinical
development upon discovery of rimonabant’s CNS-related side
effects.6
A possible strategy to avoid CNS-related side effects is to
generate CB1 antagonists that do not cross the blood-brain
barrier (BBB) and have low brain penetration. Generally,
compounds that have <10% brain penetration are considered
Received: July 25, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
compound. Oral dosing of 2 at 10 mg/kg in Sprague−Dawley
(SD) rats showed good oral absorption (Cmax = 1653 ng/g) and
limited penetration into the CNS (brain-to-plasma ratios of
0.05−0.11 were observed).12
On the basis of these promising results with 2, we
investigated a series of analogues with substitutions at the 4-
amino position of 3 (Figure 2) in an attempt to further limit
Scheme 1. Synthesis of Compounds
Figure 2. General strategy for compound development.
CNS penetration. Although the 4-amino-4-phenyl piperidine
portion of 2 was the most potent amino piperidine tested in our
previous studies, these efforts focused on the 4-amino
piperidine moiety of structure 3 because of its lower formula
weight; 3 allows us to add steric bulk while maintaining the
formula weight <600.
a
Reagents and conditions: (a) appropriate chloroformate, triethyl-
amine (TEA), tetrahydrofuran (THF); (b) appropriate carboxylic acid,
benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluoro-
phosphate (BOP), TEA, THF; (c) appropriate amino acid, BOP,
TEA, THF; (d) sulfamide, dioxane, 80 °C; (e) trifluoromethane-
sulfonic anhydride, TEA, dichloromethane.
RESULTS
■
Ligand Design and Synthesis. Of the compounds
evaluated in our earlier studies that lacked additional
substitution on the piperidine ring, the 4-amino piperidine
compounds (represented by structure 3) were the most potent
analogues.10 However, only a limited number of R groups were
tested. In this study, a number of different R-group
substitutions were targeted to augment our understanding of
the existing structure−activity relationship (SAR). Carbamates,
amides, and sulfonamides were chosen for further examination
because they were previously found to be favored over ureas.10
Synthesis of the compounds was accomplished as described
in Scheme 1. Carbamates were made by reacting the previously
described amine 410 and the appropriate chloroformate. Amides
7−16 were synthesized from 4 using benzotriazol-1-yloxy)tris-
(dimethylamino)phosphonium hexafluorophosphate (BOP),
triethylamine, and the appropriate amine or acid. Sulfamide
17 was synthesized by heating 4 with sulfamide in dioxane.
Sulfonamide 18 was prepared by reacting 4, triethylamine, and
trifluoromethanesulfonic anhydride in dichloromethane.
Pharmacological Characterization. Compounds were
tested in vitro for antagonist activity at CB1 using a calcium
mobilization assay. This CB1-antagonist activity is expressed as
an apparent dissociation affinity constant (Ke).13 Compounds
that had Ke of <10 nM were assessed for their selective affinity
for CB1 versus CB2 receptors, as determined using a
[3H]CP55940 displacement assay. For each receptor, the
values were expressed as the equilibrium dissociation constant
(Ki).10 Selectivity was expressed as a ratio of these constants.
In general, excellent activity and potency were observed with
these compounds (Table 1). Ten of the 14 compounds had Ke
< 10 nM and Ki < 20 nM at the CB1 receptor. All compounds
tested were found to be selective for the CB1 versus CB2
receptor. Selectivity ranged from ∼22-fold for 11 to >4000-fold
for 7.
reasons. First, testing compounds with different functional
groups could maximize the chances of obtaining a compound
with a desirable PK profile. Therefore, one amide and one
carbamate were chosen. Second, of the compounds with Ke <
10 nM and selectivity of >50-fold in each series, 6 and 9 had
the lowest molecular weight and clogP value in their respective
series. Compound 18 was excluded because of its relatively high
formula weight. Finally, the shielding effect of bulk on the R
group on polarity of the amide or carbamate functionality was
taken into consideration. Excessive steric bulk adjacent to the
amide or carbamate could effectively minimize the interaction
of polar groups of the molecule with their surroundings,
thereby increasing penetration across the BBB.
Both 6 and 9 were initially screened for CNS penetration by
oral dosing at 10 mg/kg in Sprague−Dawley (SD) rats.
Unperfused brain and plasma samples were collected at 1, 2, 4,
8, and 24 h post dose, and samples were analyzed by LC−MS
(data not shown). Brain-to-plasma ratios of 6 ranged from 0.14
to 0.46, representing significant brain penetration (data not
shown). Minimal to no penetration into the CNS was observed
with 9 (Table 2), with brain to plasma ratios ranging from 0.01
to 0.07. The blood volume of an unperfused rodent brain is
∼3−6%.7,8 Therefore, these PK data indicate that 9 has little to
no penetration into the CNS and is peripherally selective.
In Vivo Evaluation of Central CB1 Antagonism.
Compound 9 was also evaluated for its ability to attenuate
the central effects of a CB1 agonist in a series of in vivo tests. In
this battery of tests, known as the tetrad assay, a classical CB1
agonist such as Δ9-THC produces the suppression of
spontaneous activity, antinociception, hypothermia, and cata-
lepsy. These effects are believed to be mediated by CB1
Pharmacokinetic Studies of Select Compounds.
Pharmacokinetic (PK) studies were performed with com-
pounds 6 and 9. These compounds were chosen for several
B
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Pharmacological Assessment of Compounds
a
a
compound
R
Ke CB1 (nM)
pA2
Ki CB2 (nM)
Ki CB1 (nM)
selectivity (CB1/CB2)
5
Me
Et
17.0
2.3
7.77
8.64
8.40
8.48
8.31
8.68
8.27
9.05
8.13
8.54
6.10
6.85
5.56
8.98
6
3.57
2.55
3.35
19.47
1.93
4.01
1.54
2.67
2.02
426
119
4176
249
54
7
Ph
4.0
10649
835
8
cyclohexyl
n-Bu
3.3
9
4.9
1046
2360
89.5
60.2
215
10
11
12
13
14
15
16
17
18
cyclohexylmethyl
i-Bu
2.1
1223
22
5.4
cyclopentyl
3-methylbutyl
cyclopentylmethyl
(piperidin-1-yl)ethyl
Me2NCH2
NH2
0.89
7.4
39
81
2.9
569
282
789.0
140.0
2768.0
1.05
CF3
6.1
4501
738
a
Displacement was measured using [3H]CP55940.
9 followed 30 min later by an i.v. injection of 30 mg/kg of Δ9-
THC or vehicle. Spontaneous activity was measured in an
automated locomotor-activity chamber where beam breaks
recorded horizontal movement. Total amulatory counts for this
test are shown in Figure 3A. Antinociception was measured by
the latency of mice to withdraw their tails from a warm water
bath (55 °C). The results for this assay are expressed as the
percent of maximum possible effect (% MPE) and are shown in
Figure 3B. Hypothermia was measured by comparing the
differences in rectal temperature prior to dosing and post Δ9-
THC dosing; these results are reported as change of
temperature (Δ °C) and are shown in Figure 3C. Finally,
catalepsy was measured by placing mice on an elevated ring
apparatus and measuring the amount of time each mouse
remained motionless during a 5 min period following agonist
administration. These data are reported as the percent ring
immobility and are shown in Figure 3D. The 30 mg/kg dose of
Table 2. In Vivo Pharmacokinetic Evaluation of Compound
9
a
time (h) plasma conc. (ng/mL) brain conc. (ng/mL) brain/plasma ratio
1
730
24.3
71.5
103
0.03
0.04
0.07
0.03
0.01
2
1750
1565
1965
438
4
8
62.5
5.35
24
a
Compound 9 was dosed orally at 10 mg/kg in three SD rats per time-
point.
receptors in the brain and spinal cord14 and are blocked by
coadministration of rimonabant, a potent and CNS penetrant
CB1 antagonist.14,15
The effects of 9 on the tetrad were measured in adult ICR
mice that were orally dosed with either vehicle or 10 mg/kg of
Figure 3. Effects of vehicle and Δ9-THC (30 mg/kg) tested in combination with vehicle (unfilled bars) or compound 9 (filled bars) on spontaneous
activity (A), antinociception (B), rectal temperature (C), and catalepsy (D). Values represent the mean ( SEM) of 5 to 6 mice per group. Pound
signs (#) indicate a significant main effect (p < 0.05) of Δ9-THC (right side of each panel) compared to the vehicle (left side of each panel).
C
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Δ9-THC significantly suppressed locomotor activity (panel A)
and produced antinociception (panel B), hypothermia (panel
C), and catalepsy (panel D), regardless of whether or not 9 was
also administered (i.e., p < 0.05, main effect of Δ9-THC dose
for each measure). Hence, 9 did not reverse any of the
cannabimimetic effects of Δ9-THC. It also did not produce any
significant effects in these tests by itself when tested in
combination with vehicle.
Of the two compounds submitted for PK analyses, only 9
was found to have selectivity for the periphery with brain-to-
plasma levels ranging from 0.01 to 0.07. At only one time point
did the brain-to-plasma ratio exceed 0.04. A brain-to-plasma
ratio of ≤0.06 in an unperfused rodent brain indicates no
discernible penetration into the CNS.7,8 Thus, 9 is highly
selective for the periphery, and this lack of CNS penetration
was also confirmed in the tetrad assay wherein the centrally
mediated effects of Δ9-THC could not be reversed by 9.
Reported here are our latest efforts toward the development
of a peripherally selective CB1 antagonist for the treatment of a
wide range of clinical indications. To date, two peripherally
selective analogues of 1 have been identified (2 and 9), with 9
representing a significant improvement over 2 in the areas
plasma concentration and BBB penetration. Although both
compounds are slowly absorbed, plasma levels for 9 are greater
than plasma levels of 2 at the Cmax (1965 vs 1653 ng/mL) and
at the 8 h time point (1965 vs 914 ng/mL).10 Although 2 has
brain-to-plasma ratios that range from 0.05 to 0.11,10 9 has
brain-to-plasma ratios ranging from 0.01 to 0.07. These data,
along with the results from the tetrad, make a compelling case
for the peripheral selectivity of 9. Efficacy studies are currently
underway to test the utility of 9 in models of metabolic
syndrome and liver disease.
DISCUSSION
■
Previous reports indicate that peripherally selective CB1
receptor antagonists have value in treating important diseases
such as obesity, diabetes, and liver fibrosis. However, non-
tissue-selective compounds that can inhibit CB1 within the
CNS have serious adverse effects.16 Although CB1 is
abundantly expressed within the CNS, it is also expressed at
lower levels in several peripheral tissues.1 In contrast, CB2
receptors have limited expression within the CNS but are
highly expressed in immune cells and other peripheral tissues.1
Recent studies also indicate that although CB1 expression in
peripheral organs is lower than what is noted within the CNS,
this receptor plays an important role in the regulation of
metabolic processes. For example, peripheral CB1 receptor
inverse agonism reduces obesity by reversing leptin resistance,
and hepatic CB1 mediates diet-induced insulin resistance via
the inhibition of insulin signaling and clearance in mice.17,18
Therefore, antagonism of the peripheral CB1 receptor is an
attractive strategy to treat cardio−metabolic diseases even
though the expression of CB1 is relatively lower in peripheral
tissues compared to the CNS.
EXPERIMENTAL SECTION
■
Chemistry General. The purity and characterization of the
compounds were established by a combination of the HPLC, TLC,
and NMR analytical techniques described below. H and 13C NMR
1
spectra were recorded on a Bruker Avance DPX-300 (300 MHz)
spectrometer and were determined in CDCl3 or CD3OD with
tetramethylsilane (TMS) (0.00 ppm) or solvent peaks as the internal
reference unless otherwise noted. Chemical shifts are reported in ppm
relative to the solvent signal, and coupling constant (J) values are
reported in hertz (Hz). Thin-layer chromatography (TLC) was
performed on EMD precoated silica gel 60 F254 plates, and spots were
visualized with UV light or I2 detection. Low-resolution mass spectra
were obtained using a Waters Alliance HT/Micromass ZQ system
(ESI). All test compounds were at least 95% pure, as determined by
HPLC on an Agilent 1100 system using an Agilent Zorbax SB-Phenyl,
2.1 × 150 mm, 5 μm column with gradient elution using the mobile
phases (A) H2O containing 0.05% CF3COOH and (B) methanol. A
flow rate of 1.0 mL/min was used.
General Procedure for Making Carbamates from 1-[8-(2-
Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-
amine (4). To a solution of 1-[8-(2-chlorophenyl)-9-(4-chlorophen-
yl)-9H-purin-6-yl]piperidin-4-amine (4) (12.5 mg, 0.028 mmol, 1
equiv) in 2 mL of THF were added triethylamine (0.02 mL, 0.143
mmol, 5 equiv) and the appropriate chloroformate (2 equiv). The
reaction was stirred for 16 h. The reaction was concentrated in vacuo.
The crude material was purified by silica gel column chromatography
using 0−100% ethyl acetate/hexane to yield the pure compound.
Methyl N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-
yl]piperidin-4-yl}carbamate (5). The reaction proceeded in 79% yield.
1H NMR (300 MHz, CDCL3) δ 1.44−1.55 (m, 2H), 2.13 (d, J = 11.49
The goal of these studies was to produce a peripherally
selective CB1 antagonist based on 1 as an improvement over
our previously reported compound 2 that demonstrated ∼10%
brain penetration. To achieve this goal, the SAR around 1 was
expanded. Several potent (Ke CB1 <10 nM) and selective
(CB2/CB1 >50) amides were identified. Amides 8−14 with a
variety of alkyl groups appeared to have significant activity at
CB1. However, the presence of basic groups such as piperidine
and dimethyl amino was detrimental to activity at CB1, as
demonstrated by 15 and 16. Interestingly, affinity at the CB2
receptor could be modulated by changing the size and shape of
the alkyl substituent, as demonstrated by compounds 9−14. In
general, amides with cyclic substituents, such as 8, 10, and 14,
had excellent selectivity for CB1 over CB2.
Carbamates 5−7 all had significant activity at the CB1
receptor. However, an order of magnitude increase in activity
was observed with the addition of a single methylene to 5
(forming 6). Aromatic carbamate 7 and ethyl carbamate 6
possessed both good affinity and selectivity. However, phenyl
carbamate 7 was an order of magnitude more selective than
ethyl carbamate 6 for CB1 over CB2. This result is consistent
with the general trend observed for amides with cyclic
substituents discussed above.
Sulfamide 17 did not have significant activity in the calcium
mobilization assay. Sulfonamide 18 was very active at CB1, and
it was very selective for CB1 over CB2. The results for 18
augment our previously reported data for the methanesulfona-
mide (Ke CB1 = 71 nM; CB2/CB1 Ki = 37) and the
benzenesulfonamide (Ke CB1 = 0.3 nM; CB2/CB1 Ki = 4888)
analogues of 3.10 Data for all three sulfonamides of 3 indicated
that increased lipophilicity, and to a lesser extent mass, were
advantageous for both activity at CB1 and selectivity for CB1
over CB2.
Hz, 2H), 3.34 (t, J = 12.24 Hz, 2H), 3.68 (br s, 3H), 3.86 (br s, 1H),
4.60 (br s, 1H), 5.40 (br s, 2H), 7.13−7.45 (m, 7H), 7.51 (d, J = 6.88
Hz, 1H), 8.38 (s, 1H). [M + H]+ 497.8.
Ethyl N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-
yl]piperidin-4-yl}carbamate (6). The reaction proceeded in 55%
yield. 1H NMR (300 MHz, CDCL3) δ 1.25 (t, J = 6.92 Hz, 3H), 1.43−
1.55 (m, 2H), 2.13 (d, J = 11.21 Hz, 2H), 3.34 (t, J = 12.01 Hz, 2H),
3.86 (br s, 1H), 4.06−4.22 (m, 2H), 4.58 (br s, 1H), 5.39 (br s, 2H),
7.14−7.43 (m, 7H), 7.51 (d, J = 6.69 Hz, 1H), 8.38 (s, 1H). [M + H]+
511.3.
Phenyl N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-
yl]piperidin-4-yl}carbamate (7). The reaction proceeded in 50% yield.
D
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1H NMR (300 MHz, CDCL3) δ 1.59−1.72 (m, 2H), 2.22 (d, J = 11.40
Hz, 2H), 3.37 (t, J = 12.10 Hz, 2H), 3.84−4.02 (m, 1H), 4.99 (d, J =
7.72 Hz, 1H), 5.45 (br s, 2H), 7.02−7.45 (m, 12H), 7.52 (d, J = 6.88
Hz, 1H), 8.40 (s, 1H). [M + H]+ 559.8.
equiv). The reaction was stirred for 16 h. The reaction was
concentrated in vacuo. The crude material was purified by silica gel
column chromatography using 0−100% CMA 80/dichloromethane to
yield the compound. The product was purified further by dissolving in
ethyl acetate, washing with water, and precipitating with hexane. The
solid, pure compound was collected.
General Procedure for Making Amides from 1-[8-(2-
Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-
amine (4) Using Carboxylic Acids. To a solution of 1-[8-(2-
chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-amine
(4) (22 mg, 0.05 mmol, 1 equiv) in 2 mL of THF were added
triethylamine (0.02 mL, 0.143 mmol, 3 equiv), (benzotriazol-1-
yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP)
(22 mg, 0.05 mmol, 1 equiv), and the appropriate carboxylic acid (1
equiv). The reaction was stirred for 16 h. The reaction was
concentrated in vacuo. The crude material was purified by silica gel
column chromatography using 0−100% ethyl acetate/hexane to yield
the compound. The product was purified further by dissolving in ethyl
acetate and precipitating with hexane. The solid, pure compound was
collected.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-3-(piperidin-1-yl)propanamide (15). The reaction
1
proceeded in 41% yield. H NMR (300 MHz, CDCL3) δ 1.52−1.60
(m, 4H), 1.80 (br s, 3H), 1.98−2.14 (m, 3H), 2.76 (t, J = 5.93 Hz,
2H), 2.93 (br s, 2H), 3.30 (br s, 2H), 3.40 (br s, 2H), 3.56 (br s, 2H),
3.97−4.15 (m, 1H), 5.36−5.59 (m, 2H), 6.46−6.65 (m, 1H), 7.15−
7.42 (m, 7H), 7.54 (d, J = 7.63 Hz, 1H), 8.37 (s, 1H). [M + H]+ 580.6.
N-{1-[8-(2-chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-2-(dimethylamino)acetamide (16). The reaction
1
proceeded in 11% yield. H NMR (300 MHz, CDCL3) δ 1.50−1.66
(m, 2H), 2.01−2.13 (m, 2H), 2.48 (s, 7H), 2.64 (d, J = 9.32 Hz, 1H),
3.36 (t, J = 11.82 Hz, 2H), 4.05−4.26 (m, 1H), 5.41 (br s, 2H), 7.08
(d, J = 8.19 Hz, 1H), 7.14−7.42 (m, 7H), 7.51 (d, J = 6.78 Hz, 1H),
8.37 (s, 1H). [M + H]+ 524.7.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}cyclohexanecarboxamide (8). The reaction proceeded
1
in >99% yield. H NMR (300 MHz, CDCL3) δ 1.15−2.17 (m, 14H),
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}aminosulfonamide (17). To a solution of 1-[8-(2-
chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-amine
(4) (36.6 mg, 0.083 mmol, 1 equiv) in 2 mL of dioxane was added
sulfamide (40 mg, 0.42 mmol, 5 equiv). The reaction was heated at 80
°C for 16 h. The reaction was concentrated in vacuo. The crude
material was purified by silica gel column chromatography using 0−
100% ethyl acetate/hexane to yield 34 mg (79%) of the desired
2.23−2.40 (m, 1H), 3.31 (t, J = 12.20 Hz, 2H), 4.05−4.25 (m, 1H),
5.22−5.61 (m, 3H), 7.14−7.42 (m, 7H), 7.51 (d, J = 6.69 Hz, 1H),
8.38 (s, 1H). [M + H]+ 549.5.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}pentanamide (9). The reaction proceeded in >99%
yield. 1H NMR (300 MHz, CDCL3) δ 0.84−0.98 (m, 3H), 1.27−1.75
(m, 5H), 2.05−2.23 (m, 3H), 2.26−2.44 (m, 1H), 3.31 (t, J = 12.24
Hz, 2H), 4.02−4.30 (m, 1H), 5.22−5.63 (m, 3H), 7.13−7.44 (m, 7H),
7.51 (d, J = 6.78 Hz, 1H), 8.38 (s, 1H). [M + H]+ 523.5.
1
compound. H NMR (300 MHz, MeOH-d4) δ 1.51−1.72 (m, 2H),
2.15 (br s, 2H), 3.38−3.52 (m, 2H), 3.54−3.65 (m, 1H), 5.18−5.41
(m, 2H), 7.24−7.50 (m, 7H), 7.54−7.67 (m, 1H), 8.23 (s, 1H). [M +
H]+ 518.5.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-2-cyclohexylacetamide (10). The reaction proceeded
1
in >99% yield. H NMR (300 MHz, CDCL3) δ 0.86−1.00 (m, 3H),
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl,}-1,1,1-trifluoromethanesulfonamide (18). To a sol-
ution of 1-[8-(2-chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-amine (4) (27.9 mg, 0.064 mmol, 1 equiv) in 2 mL of
dichloromethane were added triethylamine (0.027 mL, 0.191 mmol, 3
equiv) and the trifluoromethanesulfonic anhydride (0.01 mL, 0.069
mmol, 1 equiv). The reaction was stirred for 16 h. The reaction was
concentrated in vacuo. The crude material was purified by silica gel
column chromatography using 0−100% ethyl acetate/hexane to yield
9 mg (25%) of the desired compound. 1H NMR (300 MHz, CDCL3)
δ 1.60−1.76 (m, 2H), 2.20 (d, J = 12.62 Hz, 2H), 3.30 (t, J = 12.57
Hz, 2H), 3.71−3.92 (m, 1H), 4.95 (d, J = 8.48 Hz, 1H), 5.50 (d, J =
12.34 Hz, 2H), 7.14−7.43 (m, 7H), 7.50 (d, J = 6.97 Hz, 1H), 8.39 (s,
1H). [M + H]+ 571.7.
Calcium Mobilization and Radioligand Displacement As-
says. Each compound was pharmacologically characterized using a
functional fluorescent CB1-activated Gαq16-coupled intracellular
calcium mobilization assay in CHO-K1 cells, as has been described
in our previous publication, and the apparent antagonist dissociation
equilibrium constant (Ke) values were determined.13 Briefly, CHO-K1
cells were engineered to coexpress human CB1 and Gqα16. Activation
of CB1 by an agonist then leads to the generation of inositol
phospahatase 3 (IP3) and activation of IP3 receptors, which leads to
the mobilization of intracellular calcium. Calcium flux was monitored
in a 96-well format using the fluorescent dye Calcein-4 AM in an
automated plate reader (Flexstation, Molecular Devices). The Ke of a
test compound was measured by its ability to shift the concentration
response curve of the synthetic CB1 agonist CP55940 to the right
using the equation
1.05−1.87 (m, 10H), 2.02−2.18 (m, 4H), 3.31 (t, J = 12.10 Hz, 2H),
4.10−4.28 (m, 1H), 5.22−5.62 (m, 3H), 7.14−7.43 (m, 7H), 7.51 (d, J
= 6.78 Hz, 1H), 8.38 (s, 1H). [M + H]+ 562.2.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-3-methylbutanamide (11). The reaction proceeded in
1
>99% yield. H NMR (300 MHz, CDCL3) δ 0.97 (dd, J = 9.94, 6.45
Hz, 6H), 1.50 (qd, J = 11.85, 3.81 Hz, 1H), 2.01−2.17 (m, 4H), 2.19−
2.27 (m, 2H), 3.31 (t, J = 12.24 Hz, 2H), 4.10−4.27 (m, 1H), 5.39 (d,
J = 8.01 Hz, 3H), 7.14−7.42 (m, 7H), 7.51 (d, J = 6.78 Hz, 1H) 8.38
(s, 1H). [M + H]+ 523.3.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}cyclopentanecarboxamide (12). The reaction pro-
1
ceeded in 86% yield. H NMR (300 MHz, CDCL3) δ 1.39−1.99 (m,
9H), 2.06−2.21 (m, 3H), 2.41−2.58 (m, 1H), 3.31 (t, J = 12.24 Hz,
2H), 4.08−4.27 (m, 1H), 5.40 (d, J = 7.91 Hz, 3H), 7.12−7.42 (m,
7H), 7.51 (d, J = 6.78 Hz, 1H), 8.38 (s, 1H). [M + H]+ 535.5.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-4-methylpentanamide (13). The reaction proceeded
1
in 82% yield. H NMR (300 MHz, CDCL3) δ 0.85−0.95 (m, 6H),
1.40−1.62 (m, 5H), 2.03−2.23 (m, 3H), 2.34 (t, J = 7.54 Hz, 1H),
3.31 (t, J = 12.24 Hz, 2H), 4.08−4.29 (m, 1H), 5.42 (d, J = 7.91 Hz,
3H), 7.14−7.43 (m, 7H), 7.51 (d, J = 7.16 Hz, 1H), 8.38 (s, 1H). [M
+ H]+ 537.5.
N-{1-[8-(2-Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]-
piperidin-4-yl}-2-cyclopentylacetamide (14). The reaction proceeded
1
in 98% yield. H NMR (300 MHz, CDCL3) δ 1.06−1.23 (m, 3H),
1.44−1.69 (m, 5H), 1.74−1.93 (m, 3H), 2.06−2.28 (m, 3H), 2.30−
2.41 (m, 1H), 3.32 (t, J = 12.20 Hz, 2H), 4.08−4.29 (m, 1H), 5.41 (d,
J = 7.91 Hz, 3H), 7.14−7.42 (m, 7H), 7.51 (d, J = 6.78 Hz, 1H), 8.38
(s, 1H). [M + H]+ 549.5.
Ke = [ligand]/[DR − 1] where DR is the EC50 ratio of CP55940 in
the presence or absence of a test agent.
General Procedure for Making Amides from 1-[8-(2-
Chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-
amine (4) Using Amino Acids. To a solution of 1-[8-(2-
chlorophenyl)-9-(4-chlorophenyl)-9H-purin-6-yl]piperidin-4-amine
(4) (22 mg, 0.05 mmol, 1 equiv) in 2 mL of THF were added
triethylamine (0.02 mL, 0.143 mmol, 3 equiv), (benzotriazol-1-
yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP)
(22 mg, 0.05 mmol, 1 equiv), and the appropriate carboxylic acid (1
Typically, three different antagonist concentrations were used to
shift a five-point CP55940 agonist curve (10, 1, 0.1, 0.01, and 0.001
μM). The data were fitted using nonlinear regression analyses
(GraphPad Prism). The Ke values were used to calculate pA2, which
is the negative logarithm of Ke.19
Further characterization of select compounds was performed using
radioligand displacement of [3H]SR141716, and the equilibrium
E
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
dissociation constant (Ki) values were determined as described
previously.13 The selectivity of these compounds at CB1 versus CB2
was also determined by obtaining Ki values of either receptor using the
displacement of [3H]CP55940 in the membranes of CHO-K1cells
overexpressing either receptor. The data reported are the average
values from 3 to 6 measurements performed in duplicate. The standard
errors for most measurements were between 5 and 25% of the mean
and have been left out from the tables and figures for clarity.
effects:20 suppression of locomotor activity, antinociception, decreased
rectal temperature, and ring immobility. Prior to injection, rectal
temperature and baseline latency in the warm-water tail-withdrawal
test were measured in the mice. The latter procedure involved
immersing the mouse’s tail into a water bath maintained at 55 °C and
recording latency (s) for tail removal. Control latencies were 1−3 s. A
10 s maximal latency was used to avoid damage to the mouse’s tail.
After measurement of the temperature and baseline tail-withdrawal
latency, the mice were dosed with vehicle or with 10 mg/kg 9 via oral
gavage. Thirty minutes later, mice were injected i.v. in the tail vein with
vehicle or with 30 mg/kg Δ9-THC. After 5 min, they were placed into
individual activity chambers for 10 min. Spontaneous activity was
measured as the total number of beam interruptions during the entire
session. Tail-withdrawal latency was measured again at 20 min
postinjection. Rectal temperature was measured again at 30 min
postinjection. At 40 min postinjection, the mice were placed on the
elevated ring apparatus, and the amount of time that the animals
remained motionless during a 5 min period was recorded.
Rectal temperature values were expressed as the difference between
the control temperature (before injection) and the temperature
following drug administration (Δ °C). Spontaneous activity was
measured as total number of photocell beam interruptions during the
10 min session. Antinociception was expressed as the percent
maximum possible effect (MPE) using a 10 s maximum test latency
as follows: ((test − control)/(10 − control)) × 100. During the
assessment for catalepsy, the total amount of time (s) that the mouse
remained motionless on the ring apparatus (except for breathing and
whisker movement) was measured and was used as an indication of
catalepsy-like behavior. This value was divided by 300 s and multiplied
by 100 to obtain a percent immobility. Factorial ANOVAs (Δ9-THC
dose × 9 dose) were used to analyze the results of the tetrad tests.
Significant main effects and interactions were further analyzed with
Tukey post hoc tests (Δ = 0.05) as necessary.
Pharmacokinetic Studies. Male Sprague−Dawley (SD) rats, aged
9 weeks at the time of dosing, were acquired from Charles River
Laboratories and were dosed orally. Doses were formulated in corn oil
and administered by gavage. Animals were sacrificed at various time
points and analyzed. Samples were prepared and analyzed as follows.
Plasma (50 μL) was mixed with 300 μL of acetonitrile, vortexed, and
centrifuged at 9000g for 5 min. The supernatant was transferred to
autosampler vials with low-volume inserts and injected without
dilution. The brain was homogenized with 50:50 ethanol/water (3:1,
v/v) using a Potter Elvehjem type homogenizer. The homogenate (50
μL) was mixed with 300 μL of acetonitrile, vortexed, and centrifuged
at 9000g for 5 min. Ths supernatant was transferred to autosampler
vials with low-volume inserts and injected without dilution. Standards
were prepared as above for each compound in blank plasma, blank
liver homogenate, and blank brain homogenate. The standards used
were within 15% of nominal except for 20% at LOQ. Compounds for
LC−MS/MS analyses were supplied at 1 mg/mL in methanol. The
stock solutions were further diluted to ∼100 ng/mL. The 100 ng/mL
solutions were used to optimize the mass spectrometer for MRM
transitions and mass spectrometer parameters. Infusion and flow
injection optimization were also performed. LC−MS/MS was
conducted using an Applied Biosystems API 4000 coupled with an
Agilent 1100 HPLC system. Chromatography was performed with a
Phenomenex Luna C18 (50 × 2 mm, 5 μm) column. Mobile phases
were 0.1% formic acid and 10 mM ammonium formate in water (A)
and 0.1% formic acid and 10 mM ammonium formate in methanol
(B). The initial conditions were 10% B and held for 1 min followed by
a linear gradient to 90% B over 5 min. Ninety percent B was held for 2
min before returning to the initial conditions. Compound 9 was
analyzed with multiple reactions and was monitored in the positive
mode with a transition of 522.922 → 368.0. The following parameters
were used, DP = 171, CE = 51, and CXP = 30.
ASSOCIATED CONTENT
* Supporting Information
■
S
HPLC and melting point data of target compounds 5−18. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Tetrad Test. Male adult ICR mice, obtained from Harlan
(Indianapolis, IN) and housed singly in polycarbonate mouse cages,
were used for the assessment of locomotor suppression, antinoci-
ception, hypothermia, and catalepsy. Separate mice were used for
testing each condition in this battery of procedures. All animals were
maintained in a temperature-controlled (20−22 °C) environment with
a 12 h light−dark cycle (lights on at 6 a.m.) and had free access to
food and water when in their home cages. The in vivo studies reported
in this Article were carried out in accordance with guidelines published
in the Guide for the Care and Use of Laboratory Animals (National
Research Council, 2011). Δ9-Tetrahydrocannabinol (Δ9-THC),
obtained from the National Institute on Drug Abuse (NIDA, Bethesda,
MD) through the NIDA Drug Supply Program, was mixed in a vehicle
of 7.8% polysorbate 80 N.F. (VWR, Marietta, GA) and 92.2% sterile
saline USP (Butler Schein, Dublin, OH) and was given by i.v.
injection. Compound 9 was dissolved in peanut oil and administered
via oral gavage at a volume of 5 mL/kg.
The measurement of spontaneous activity occurred in standard
Plexiglas locomotor activity chambers (33 × 51 × 23 cm3). Beam
breaks were recorded by San Diego Instruments Photobeam Activity
System software (model SDI: V-71215, San Diego, CA) on a
computer located in the experimental room. The apparatus contained
two 4-beam infrared arrays that measured horizontal movement. A
glass beaker filled with water heated at 55 °C was used for the warm-
water tail-withdrawal procedure. A digital thermometer (Physitemp
Instruments, Inc., Clifton, NJ) was used to measure rectal temperature.
The ring immobility device consisted of an elevated metal ring (5.5 cm
diameter, 16 cm height) attached to a metal stand.
AUTHOR INFORMATION
Corresponding Author
*Tel: 919-541-6795. Fax: 919-541-8868. E-mail: rmaitra@rti.
■
org.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Anne Gilliam for performing the radioligand
displacement assays. We express our gratitude to the NIDA
drug supply program for providing radiolabeled probes and to
Dr. Brian Thomas for supplying the CB1 cells. This research
was funded by research grants AA019740-01 to R.M. and DA-
003672 to J.W.
ABBREVIATIONS USED
■
CB1, cannabinoid receptor 1; CB2, cannabinoid receptor 2;
CMA 80, 80% chloroform, 18% methanol, and 2% ammonium
hydroxide; CNS, central nervous system; BBB, blood-brain
barrier; TPSA, topological polar surface area; ECS, endocanna-
binoid system; CBR, cannabinoid receptors; Ke, apparent
affinity constant; MDCK-mdr1, Madin-Darby canine kidney
cells transfected with the human MDR1 gene; A, apical; B,
basal; BOP, benzotriazole-1-yl-oxytris(dimethylamino)-
phosphonium hexafluorophosphate; CHO-K1, chinese hamster
Each mouse was tested in a battery of four tests in which
cannabinoid agonists produced a characteristic profile of in vivo
F
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
noid system for the treatment of obesity. Pharmacol. Rep. 2009, 61,
217−224.
ovary cells; IP3, inositol phospahatase 3; MRM, multiple
reaction monitoring; LOQ, below the limit of quantitation; NA,
not applicable
(17) Liu, J.; Zhou, L.; Xiong, K.; Godlewski, G.; Mukhopadhyay, B.;
Tam, J.; Yin, S.; Gao, P.; Shan, X.; Pickel, J.; Bataller, R.; O’Hare, J.;
Scherer, T.; Buettner, C.; Kunos, G. Hepatic cannabinoid receptor-1
mediates diet-induced insulin resistance via inhibition of insulin
signaling and clearance in mice. Gastroenterology 2012, 142, 1218−
1228.
REFERENCES
■
(1) Pertwee, R. G.; Howlett, A. C.; Abood, M. E.; Alexander, S. P.; Di
Marzo, V.; Elphick, M. R.; Greasley, P. J.; Hansen, H. S.; Kunos, G.;
Mackie, K.; Mechoulam, R.; Ross, R. A. International Union of Basic
and Clinical Pharmacology. LXXIX. Cannabinoid receptors and their
ligands: Beyond CB1 and CB2. Pharmacol. Rev. 2010, 62, 588−631.
(2) Bouaboula, M.; Bianchini, L.; McKenzie, F. R.; Pouyssegur, J.;
Casellas, P. Cannabinoid receptor CB1 activates the Na+/H+
exchanger NHE-1 isoform via Gi-mediated mitogen activated protein
kinase signaling transduction pathways. FEBS Lett. 1999, 449, 61−65.
(3) Bouaboula, M.; Perrachon, S.; Milligan, L.; Canat, X.; Rinaldi-
Carmona, M.; Portier, M.; Barth, F.; Calandra, B.; Pecceu, F.; Lupker,
J.; Maffrand, J. P.; Le Fur, G.; Casellas, P. A selective inverse agonist
for central cannabinoid receptor inhibits mitogen-activated protein
kinase activation stimulated by insulin or insulin-like growth factor 1.
Evidence for a new model of receptor/ligand interactions. J. Biol.
Chem. 1997, 272, 22330−22339.
(18) Tam, J.; Cinar, R.; Liu, J.; Godlewski, G.; Wesley, D.; Jourdan,
T.; Szanda, G.; Mukhopadhyay, B.; Chedester, L.; Liow, J. S.; Innis, R.
B.; Cheng, K.; Rice, K. C.; Deschamps, J. R.; Chorvat, R. J.; McElroy, J.
F.; Kunos, G. Peripheral cannabinoid-1 receptor inverse agonism
reduces obesity by reversing leptin resistance. Cell Metab. 2012, 16,
167−179.
(19) Arunlakshana, O.; Schild, H. O. Some quantitative uses of drug
antagonists. Br. J. Pharmacol. Chemother. 1959, 14, 48−58.
(20) Martin, B. R.; Compton, D. R.; Thomas, B. F.; Prescott, W. R.;
Little, P. J.; Razdan, R. K.; Johnson, M. R.; Melvin, L. S.; Mechoulam,
R.; Ward, S. J. Behavioral, biochemical, and molecular modeling
evaluations of cannabinoid analogs. Pharmacol., Biochem. Behav. 1991,
40, 471−478.
(4) Janero, D. R.; Makriyannis, A. Cannabinoid receptor antagonists:
Pharmacological opportunities, clinical experience, and translational
prognosis. Expert Opin. Emerging Drugs 2009, 14, 43−65.
(5) Di Marzo, V. CB(1) receptor antagonism: Biological basis for
metabolic effects. Drug Discovery Today 2008, 13, 1026−1041.
(6) Lee, H. K.; Choi, E. B.; Pak, C. S. The current status and future
perspectives of studies of cannabinoid receptor 1 antagonists as anti-
obesity agents. Curr. Top. Med. Chem. 2009, 9, 482−503.
(7) Chugh, B. P.; Lerch, J. P.; Yu, L. X.; Pienkowski, M.; Harrison, R.
V.; Henkelman, R. M.; Sled, J. G. Measurement of cerebral blood
volume in mouse brain regions using micro-computed tomography.
NeuroImage 2009, 47, 1312−1318.
(8) Edvinsson, L.; Nielsen, K. C.; Owman, C. Circadian rhythm in
cerebral blood volume of mouse. Experientia 1973, 29, 432−3.
(9) Kunos, G.; Tam, J. The case for peripheral CB(1) receptor
blockade in the treatment of visceral obesity and its cardiometabolic
complications. Br. J. Pharmacol. 2011, 163, 1423−1431.
(10) Fulp, A.; Bortoff, K.; Zhang, Y.; Seltzman, H.; Mathews, J.;
Snyder, R.; Fennell, T.; Maitra, R. Diphenyl purine derivatives as
peripherally selective cannabinoid receptor 1 antagonists. J. Med.
Chem. 2012, 55, 10022−10032.
(11) Griffith, D. A.; Hadcock, J. R.; Black, S. C.; Iredale, P. A.;
Carpino, P. A.; DaSilva-Jardine, P.; Day, R.; DiBrino, J.; Dow, R. L.;
Landis, M. S.; O’Connor, R. E.; Scott, D. O. Discovery of 1-[9-(4-
chlorophenyl)-8-(2-chlorophenyl)-9H-purin-6-yl]-4-ethylaminopiperi-
dine-4-carboxylic acid amide hydrochloride (CP-945,598), a novel,
potent, and selective cannabinoid type 1 receptor antagonist. J. Med.
Chem. 2009, 52, 234−237.
(12) Fulp, A.; Bortoff, K.; Seltzman, H.; Zhang, Y.; Mathews, J.;
Snyder, R.; Fennell, T.; Maitra, R. Design and synthesis of cannabinoid
receptor 1 antagonists for peripheral selectivity. J. Med. Chem. 2012,
55, 2820−2834.
(13) Zhang, Y.; Gilliam, A.; Maitra, R.; Damaj, M. I.; Tajuba, J. M.;
Seltzman, H. H.; Thomas, B. F. Synthesis and biological evaluation of
bivalent ligands for the cannabinoid 1 receptor. J. Med. Chem. 2010,
53, 7048−7060.
(14) Wiley, J. L.; Martin, B. R. Cannabinoid pharmacological
properties common to other centrally acting drugs. Eur. J. Pharmacol.
2003, 471, 185−193.
(15) Compton, D. R.; Aceto, M. D.; Lowe, J.; Martin, B. R. In vivo
characterization of a specific cannabinoid receptor antagonist
(SR141716A): Inhibition of delta 9-tetrahydrocannabinol-induced
responses and apparent agonist activity. J. Pharmacol. Exp. Ther.
1996, 277, 586−594.
(16) Leite, C. E.; Mocelin, C. A.; Petersen, G. O.; Leal, M. B.;
Thiesen, F. V. Rimonabant: An antagonist drug of the endocannabi-
G
dx.doi.org/10.1021/jm401129n | J. Med. Chem. XXXX, XXX, XXX−XXX