Conversion of 4-cyanomethyl-pyrazole-3-carboxamides into CB1 antagonists with lowered propensity to pass the blood-brain-barrier

Article abstract of DOI:10.1016/j.bmcl.2009.12.003

A series of amides, amidines and amidoximes have been made from the corresponding nitrile compounds, to provide potent antagonists and inverse agonists for the CB1 receptor with considerably lower lipophiliciy, higher polar surface area and improved plasma/brain ratios compared to the centrally acting rimonabant. Extensive investigations of ADME and in vivo pharmacological properties led to selection of the amide series and specifically the 4-(4-fluorophenyl)piperidin-4-ol derivative D4. A clear improvement in the peripheral profile over rimonabant was seen, although some contribution of central effect on the pronounced weight reduction in obese mice cannot be ruled out.

Full text of DOI:10.1016/j.bmcl.2009.12.003

Bioorganic & Medicinal Chemistry Letters 20 (2010) 453–457
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Conversion of 4-cyanomethyl-pyrazole-3-carboxamides into CB1
antagonists with lowered propensity to pass the blood–brain-barrier
Jean-Marie Receveur, Anthony Murray, Jean-Michel Linget, Pia K. Nørregaard, Martin Cooper,
*
Emelie Bjurling, Peter Aadal Nielsen, Thomas Högberg
7TM Pharma A/S, Fremtidsvej 3, DK-2970 Hørsholm, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of amides, amidines and amidoximes have been made from the corresponding nitrile com-
pounds, to provide potent antagonists and inverse agonists for the CB1 receptor with considerably lower
lipophiliciy, higher polar surface area and improved plasma/brain ratios compared to the centrally acting
rimonabant. Extensive investigations of ADME and in vivo pharmacological properties led to selection of
the amide series and specifically the 4-(4-fluorophenyl)piperidin-4-ol derivative D4. A clear improve-
ment in the peripheral profile over rimonabant was seen, although some contribution of central effect
on the pronounced weight reduction in obese mice cannot be ruled out.
Received 24 October 2009
Revised 28 November 2009
Accepted 1 December 2009
Available online 4 December 2009
Keywords:
CB1 antagonists
CB1 inverse agonists
Obesity
Ó 2009 Elsevier Ltd. All rights reserved.
DIO model
CNS
Blood–brain-barrier
Plasma/brain ratio
The endocannabinoid signaling substances are derived from
arachidonic acid and act on two types of GPCR receptors, that is,
cannabinoid receptors CB1 and CB2.1,2 The CB1 receptor is widely
expressed in the brain and to a lower extent in several peripheral
organs such as liver, adipose tissue, gastrointestinal tract, vagal
nerves, pancreas and skeletal muscle.1–3 CB2 receptors exist
mainly in the immune system, although recent publications have
reported their existence also in the CNS.1–3 The endocannabinoids
play a pivotal role, through the CB1 receptor, in regulating food in-
take and energy expenditure in synergy with other anorexigenic
and orexigenic regulatory pathways.1,2 CB1 receptor-deficient mice
are resistant to diet-induced obesity,4 and CB1 antagonists/inverse
agonists, such as rimonabant and taranabant shown in Figure 1, in-
hibit food intake and reduce body weight in obese animals and hu-
mans.1,2,5 These effects are likely to be mediated by a combination
of central and peripheral target organ actions.2,6 Pharmacological
blockade of the peripheral CB1 system can elicit beneficial effects
on other metabolic and atherogenic risk factors that are indepen-
dent of eating behavior and body weight.2,6 Clinical trials with
rimonabant also showed improvement in glycemic control and li-
pid profile in type 2 diabetic patients (SERENADE) and loss of vis-
ceral and hepatic fat in abdominally obese patients (ADAGIO).6,7
These long term peripheral effects could arise from, for example,
reduced hepatic lipogenesis, enhanced lipid oxidation, adipose
lipolysis, and reduced insulin resistance. Recently, blockade of
the CB1 receptor has also been shown in rodents to hold promise
as a therapy for chronic liver diseases such as liver fibrosis.8
An issue associated with this drug class is the induction of cen-
tral nervous system (CNS) side-effects such as anxiety and depres-
sion2,9, effects which are evidently linked to functional CB1
receptors in brain areas such as the frontal cortex, hippocampus
and amygdala.10 A plausible working hypothesis to avoid the CNS
effects seen with rimonabant and taranabant, but still achieve ef-
fects on body weight and metabolic parameters, would be to de-
sign compounds unable to pass the blood–brain-barrier (BBB).2,11
One possibility to achieve this would be by making compounds
having considerably higher polar surface area (PSA) and lower
O
N
CN
N
H
O
N
N
O
N
N
Cl
H
Cl
CF3
Cl
rimonabant
taranabant
Cl
* Corresponding author. Tel.: +45 3925 7760; fax: +45 3925 7776.
E-mail address: th@7tm.com (T. Högberg).
Figure 1. Two representative CB1 antagonists progressed to clinical development.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.12.003

454
J.-M. Receveur et al. / Bioorg. Med. Chem. Lett. 20 (2010) 453–457
Table 1
log D than the centrally acting drugs. We have previously described
the synthesisand SAR features of a diverse library of 4-cyanomethyl-
1,5-diphenyl-1H-pyrazole-3-carboxamides B (Scheme 1).12 In this
report we present the synthesis and characterization, as CB1 antag-
onists, of subsequent compound series in which the nitrile has been
converted into more polar neutral (amides D and amidoximes F) or
positively charged (amidines E) functional groups.
Antagonism and inverse agonism (shown in brackets) of hCB1 by representative
4-substituted pyrazole-3-carboxamides
O
X
R
N
N
Cl
Cl
For the conversion of nitriles to amides, the standard procedure
was hydrolysis of the intermediate acid A to amide acid C by treat-
ment with a mixture of acetic acid/concentrated sulphuric acid
(2:1) at 100 °C for 30 min, followed by coupling with the amine
using standard conditions (Scheme 1). When salt forms of amines
were used the coupling was conducted in the presence of diisopro-
pylethylamine. The nitriles B were used directly in the conversion
to amidines E and amidoximes F. The former were obtained by
firstly reacting ammonium chloride with trimethylaluminium in
dry toluene, followed by addition of nitrile B and raising the tem-
perature to 160 °C in a microwave oven. The amidoximes F were
conveniently obtained by reaction of B with hydroxylamine in
methanol at 40 oC.
The compounds were tested against the human CB1 receptor
using a GTPcS antagonist assay with CP55940 as agonist and mem-
branes produced from CHO-K1 cells stably transfected with recom-
binant human CB1 receptor.13 Several compounds were also
investigated for inverse agonism, by running the same assay with-
out agonist, and for selectivity versus the CB2 subtype receptor in a
R
IC50 GTPc
Sa (nM)
Bb
D
E
F
N
N
N
1
2
7.5 [1.8] 1.4 [0.82] 0.93 [0.40] 1.8 [1.3]
H
N
H
O
71
nd
7.0
26
nd
51
N
3
4
0.51
0.54
F
OH
3.0
0.19
1.9
nd
N
N
F
N
N
F
N
5
6
7
3.8
11
0.71
nd
9.7
21
1.1
nd
N
H
1.5 [0.59]
1.2 [0.79]
GTPcS antagonist assay using CP55940 as agonist. Table 1 displays
a comparison of a small set of amides D, amidines E and amidox-
imes F in comparison with the corresponding nitriles B, described
earlier. Some general trends can be mentioned about the impact of
lowering chromatographically measured log D values14 and raising
polar surface areas of the new western chemotypes D, E and F. For
example, compared to the nitrile B1 (log D 3.6/ PSA 74 Å2) the fol-
lowing differences are observed for the corresponding amide D1
(À1.0/+19), amidine E1 (À1.7/+26) and amidoxime F1 (À0.7/+35).
The somewhat surprisingly small effect on log D resulting from
the conversion of nitriles to amidoximes F might be attributable
F
2.1
nd
N
H
a
Antagonist (using CP55940 as agonist) and inverse agonist activity in GTPcS
assay with membranes produced from CHO-K1 cells stably transfected with hCB1
receptor. All compounds displayed efficacy above 100% and all values are single or
mean of double determinations.13
Nitriles described in earlier publication.12
b
to the intra-molecular hydrogen bonding indicated in Scheme 1 –
a possibility that would make this series less attractive as it would
facilitate BBB penetration. We also observed an inappropriate brain
to plasma ratio for the amidoximes and for some amides (see
below).
R1
O
O
OH
N
NC
NC
R2
When comparing the CB1 potency for the same set of compounds
(B1-F1), an encouraging trend of increased potency compared to the
nitriles, was observed for the three more polar derivatives. The cor-
responding morpholine derivatives (2) showed a reduced potency,
consistent with previous findings.15,16 The related polar members
of phenyl piperidines (3 and 4) and piperazine (5) displayed variable
potencies, however the amides D were consistently very potent,
which might indicate some cooperative binding interactions. The
homologous amidinesE6 and E7 also show pronounced antagonistic
and inverse agonistic activity. Even if the comparisons should not be
stretched too far, this small set of compounds demonstrated that
lowering log D and increasing PSA had no negative effect on potency,
which supported our current strategy to modulate polarity by mod-
ifications in this part of the molecules.
EDAC/HOBt
R1 R2
N
N
N
N
Cl
Cl
Cl
Cl
N
H
A
B
NH4Cl, AlMe3
AcOH
H2SO4
Toluene, Δ
NH2OH. HCl
Et3N
MeOH
Δ
O
H2N
R1
OH
O
H2N
HN
O
N
R2
N
N
N
N
Cl
Cl
Cl
Cl
C
E
Although, the amidines E had very good microsomal stability
they showed insufficient in vivo ADME properties. For example,
E6 remained unaffected after 30 min in rat or mouse microsomes,
but it showed a modest bioavailability of 19% in rat even after
intraperitoneal administration (no detectable levels were found
after po administration). However, the plasma/brain ratio of 3.2
of E6 was an improvement compared to rimonabant having a ratio
of 0.22 in rat (Table 3). In mouse, rimonabant displays plasma/
brain ratios in the range of 0.3 to 0.5 throughout the period of 2
and 13 h after oral administration. As mentioned the amidoximes,
exemplified by F1, showed a poor plasma/brain ratio of 1.2 after
H
N
R1
O
EDAC/HOBt
DMF
HO
H
N
R2
R1
R2
N
R1
O
N
N
H2N
O
N
H
N
R2
Cl
Cl
N
F
N
Cl
Cl
D
Scheme 1.

J.-M. Receveur et al. / Bioorg. Med. Chem. Lett. 20 (2010) 453–457
455
Table 2
Antagonism, inverse agonism and receptor binding of hCB1 by representative 4-carbamoylmethyl-pyrazole-3-carboxamides.
O
H2N
R
O
N
N
Cl
Cl
D
R
log Dc
Solubility (
l
M)
IC50 GTP
c
Sa Antagonism (nM)
IC50 GTP
2.9
c
Sa Inverse agonism (nM)
IC50 Bindb (nM)
rimonabant
4.7
10
4.5
5.1
1.8
1
2
N
H
N
N
2.6
1.8
4.5
nd
1.4
7.0
0.82
1.7
N
H
O
7.9
2.6
OH
4
5
2.8
2.6
75
0.19
0.71
0.10
0.30
N
N
F
N
N
N
1.1
1.4
3.7
O
N
H
N
8
2.3
80
nd
1.2
9
2.4
2.9
200
16
3.0
1.7
0.55
0.34
N
H
N
CN
10
0.66
0.49
N
H
N
N
CN
11
12
13
2.8
3.4
2.7
nd
1.2
1.4
4.6
N
CN
OH
OH
9.0
105
0.23
0.26
0.11
0.11
0.78
2.9
N
N
N
14
15
16
2.3
2.5
1.5
400
290
425
3.6
nd
16
2.0
2.1
10
1.9
5.5
23
N
N
OH
N
F
COOH
COOH
N
N
17
1.6
500
30
nd
nd
F
a
b
c
As in Table 1.
Receptor binding in COS7 (rimonabant, 10, 13, 15, 17) or CHO-K1 (other) cells using [3H]-CP55940 as tracer.13
Chromatographically determined log D values.14

456
J.-M. Receveur et al. / Bioorg. Med. Chem. Lett. 20 (2010) 453–457
Table 3
Plasma/brain ratios in lean rodents of selected compounds
Compd
Species
Admin.
route
Dose
(mg/kg)
Time (h)
Plasma/
braina
Rimonabant
Rimonabant
E6
F1
D4
D8
D9
D14
Mouse
Rat
Rat
Mouse
Mouse
Mouse
Rat
po
po
ip
po
po
po
po
po
10
5
6
6
6
3
6
6
6
6
0.42
0.22
3.2
1.2
10.2
6.1
10
10
10
10
5
1.2
10.2
Mouse
10
a
Plasma/brain ratios were calculated after subtraction of 1% of the plasma con-
centration from the brain levels to account for contaminations in homogenized
brain.18
oral administration in mouse and was accordingly not pursued fur-
ther. A compilation of plasma/brain ratios in lean rodents is shown
in Table 3.
Figure 2. Blockade of CB1 agonist (WIN-55,212, 2.5 mg/kg, sc) induced hypothermia
in mouse of amide D4 (A:, 1, 3, 10 mg/kg, po) and rimonabant (B: 0.3, 1, 3 mg/kg, po).
Temperature measured 1 h after agonist dosing. The antagonists were administered
45 min prior to the agonist. V is denoting vehicle treatment. Student’s t test: *p < 0.05;
**p < 0.01; ***p < 0.001 compared with agonist alone.
With the limitations seen with the amidines and amidoximes,
we focused on the amide series D (Table 2), to capitalize on the
SAR information we gained earlier, with the intention to arrive at
compounds with log D values below 2.5 and good solubilities.
One of the first amides D1 showed very poor oral bioavailability
in rat (2%) which was attributed to a very low solubility, since both
PAMPA permeability and microsomal stability in rat were good
(72% remain after 30 min). Introduction of a cyclopropyl-carbox-
amide (D8) improved solubility over D1 while retaining pro-
nounced inverse agonism and receptor binding. The pyridyl
derivative D9 displayed good solubility, log D and CB1 potency.
This derivative D9, with a microsomal stability of 63% comparable
to D1, had on the other hand a reasonable oral bioavailability of
39% in rat, but penetrated into the brain over time (ratio ꢀ1.0 at
6 h). As the piperidinylamide D8 showed a plasma/brain ratio of
6.1, we decided to investigate larger compounds to determine if
they would have improved separation. The cyano substituted ben-
zylamine D10, which was the most polar and yet potent analogue
of this substance class identified earlier among the nitrile deriva-
tives B,12 did not show the anticipated improvement in log D and
solubility, so this series was not progressed further.
Disappointingly, the very potent piperazines D5 and D11 and
the 4-cyanopiperidine D12 did not have sufficiently good physico-
chemical properties to warrant further studies. However, the 4-hy-
droxy-4-phenylpiperidines D4 and D13 and the pyridyl analogues
D14 and D15 combined good potency with appropriate physico-
chemical properties. The pyridyl compounds improvement in log D
and solubility are only at the slight expense of inverse agonistic po-
tency. Further improvement of these physicochemical parameters
was achieved by the introduction of a 4-carboxyl moiety, however,
in this case with greater loss of CB1 potency (D16 and D17). The
acid D17 also showed a very poor oral bioavailability of 4% in
mouse and hence was not explored.
Encouragingly the 4-hydroxypiperidine D4 and the pyridyl ana-
logue D14 showed 10-fold higher concentrations in plasma than
brain, 6 h after oral administration to mouse (Table 3). Both com-
pounds showed excellent CB1 selectivity, with greater than 1000-
fold lower potency on CB2. As the phenylpiperidine D4 also showed
better ADME properties than the pyridyl compounds D14 and D15,
with high oral bioavailability in mouse (ꢀ100%) and rat (70%), we
continued in vivo investigations with D4 as tool compound. The
amide D4 was initially assessed for its central pharmacodynamic
effect in the hypothermia model, one of the four tests that consti-
tute the ‘tetrad’. CB1 agonists induce a reduction in body tempera-
ture by acting on CB1 receptors in the hypothalamus,17 an effect
that can be reversed by an antagonist. Notably, D4 gives a partial
reduction only in the highest dose of 10 mg/kg, whereas rimona-
bant already at 1 mg/kg gives a full blockade (Fig. 2).
The effect of D4 on food intake (data not shown) and body
weight was investigated in diet-induced obese (DIO) mice over
seven days (Fig. 3). The plasma/brain ratio of D4 was found to
be 13 in these obese animals after 6 h. A clear dose-dependant
47
46
45
44
**
***
43
42
41
40
39
38
37
36
***
***
***
***
***
***
***
***
***
***
***
***
***
-6 -5 -4 -3 -2 -1
0
1
2
3
4
5
6
7
8
9
Day
Figure 3. Effect on body weight after subchronic oral administration of D4 (N10, j
3 and ꢀ1 mg/kg) to DIO mice versus vehicle control animals (s). Data analyzed by
ANCOVA with body weight on day 1 as covariate followed by Williams’ test:
**
p < 0.01,
***
p < 0.001.
46
45
44
43
42
41
40
39
38
37
36
35
***
***
***
***
***
***
***
-6 -5 -4 -3 -2 -1
0
1
2
3
4
5
6
7
8
9
Day
Figure 4. Effect on body weight after subchronic oral administration of rimonabant
(N10 mg/kg) to DIO mice versus vehicle control animals (s). Statistics as in Figure 3.

J.-M. Receveur et al. / Bioorg. Med. Chem. Lett. 20 (2010) 453–457
457
7. (a) Rosenstock, J.; Hollander, P.; Chevalier, S.; Iranmanesh, A. Diabetes Care
2008, 31, 2169; (b) Després, J. P. Arterioscler. Thromb. Vasc. Biol. 2009, 29, 416;
(c) Després, J. P. Curr. Pharm. Des. 2009, 15, 553.
8. Caraceni, P. Best Pract. Res. Clin. Endocrinol. Metab. 2009, 23, 65.
9. (a) Moreira, F. A.; Grieb, M.; Lutz, B. Best Pract. Res. Clin. Endocrinol. Metab. 2009,
23, 133; (b) Viveros, M. P.; de Fonseca, F. R.; Bermudez-Silva, F. J.; McPartland, J.
M. Endocrinol. Metab. Immune. Disord. Drug Targets 2008, 8, 220.
10. Viveros, M. P.; Marco, E. M.; Llorente, R.; López-Gallardo, M. Neural Plast. 2007,
2007, 52908.
11. Kunos, G.; Osei-Hyiaman, D.; Bátkai, S.; Sharkey, K. A.; Makriyannis, A. Trends
Pharmacol. Sci. 2009, 30, 1.
12. Cooper, M.; Receveur, J.-M.; Bjurling, E.; Nørregaard, P. K.; Aadal Nielsen, P.;
Sköld.; Högberg, T. Bioorg. Med. Chem. Lett. 2009. doi:10.1016/
j.bmcl.2009.11.047.
reduction of body weight was observed reaching 5.6% and 15.1%
at 3 and 10 mg/kg, respectively. Rimonabant at 10 mg/kg gives
comparable reductions in body weight (14.5%) as D4 (Fig. 4),
but exhibits more pronounced central effects already at 1/10 of
the dose (cf. Fig. 2).
In conclusion, we have developed potent antagonists and in-
verse agonists of the CB1 receptor, incorporating either an amide,
amidine or amidoxime group which gives rise to considerably low-
er lipophilicity, higher polar surface area and improved plasma/
brain ratios compared to the centrally acting rimonabant. Exten-
sive investigation of ADME and in vivo pharmacological properties
led to selection of the amide series and specifically the 4-(4-fluoro-
phenyl)piperidin-4-ol derivative D4. Although the exposure of D4
in brain is considerably lower than that of rimonabant, it neverthe-
less displays comparable efficacy in terms of weight reduction. This
observation supports the concept that efficacious drugs for the
treatment of obesity and associated cardiometabolic risk factors
are feasible without CNS liabilities, In our search for peripherally
acting CB1 antagonists, we will report further studies, with nega-
tively charged compounds in due course.
13. CB1 antagonism was assessed by measuring inhibition of CP55940-induced
[
35S]GTP
cS binding to membranes prepared from CHO-K1 cells expressing the
human CB1 receptor (Harrison and Traynor, Life Sci. 2003, 74, 489). Assessment
of inverse agonism was conducted analogously without addition of the agonist
CP55940. CB1 receptor binding (Bylund and Toews, Am J Physiol. 1993, 265,
L421-9.) was analyzed by measuring displacement of [3H]-CP55940 binding to
COS7 or CHO-K1 cells expressing the human CB1 receptor. CP55940 is a well
known non-selective CB1 and CB2 receptor agonist (e.g., Felder et al., Mol.
Pharmacol. 1995, 48, 443). Experimentals: The [35S]GTP
c
S SPA (Scintillation
Proximity Assay; cf. Glickman et al., Assay Drug Dev. Technol. 2008, 6, 433)
binding assay was performed by incubating 5 g hCB1-membranes with 1 nM
35S]GTP
S in the presence of 3 nM of CP55940 (for antagonist assessment) and
l
[
c
various concentrations of the test compounds at room temperature for 1 h
0.4 mg/well SPA beads (PVT-WGA; RPNQ0001 Amersham Pharmacia Biotech)
were then added and the incubation continued for further 30 min on an orbital
shaker. The assay buffer contained 50 mM HEPES (pH 7.5), 50 mM NaCl,
Acknowledgements
2.5 mM MgCl2, 0.1% BSA, 1
were centrifuged at 1500 rpm for 5 min and radioactivity was read
immediately using Topcounter (PerkinElmer Life Sciences). The CB1
lM GDP and 100 lg/ml Saponin. Microtiter plates
The authors thank Rokhsana Andersen, Niklas Sköld, Kirsten
Nymann Petersen, Peter Andersen, Joan Gredal, Ann Christensen
and Helle Iversen, for excellent technical assistance.
a
receptor binding assay was performed by incubating 30,000 cells with
0.15 nM [3H]-CP55940 in the presence of various concentrations of the test
compounds at 4 °C for 3 h. The assay buffer contained 50 mM Tris (pH 7.5),
3 mM MgCl2, 1 mM EDTA and 0.5% BSA. The cells were washed twice and then
lysed for 30 min at room temperature with 0.1N NaOH. After overnight
incubation with scintillation fluid (Microscint 20) radioactivity was read using
a Topcounter. Data from assays were analyzed and IC50 values determined by
non-linear regression using the PRISM software (GRAPHPAD Software, San Diego).
14. Valko, K.; Du, C. M.; Bevan, C.; Reynolds, D. P.; Abraham, M. H. Curr. Med. Chem.
2001, 8, 1137.
15. Lan, R.; Liu, O.; Fan, P.; Lin, S.; Fernando, S. R.; McCallion, D.; Pertwee, R.;
Makriyannis, A. J. Med. Chem. 1999, 42, 769.
16. Lange, J. H. M.; van Stuivenberg, H. H.; Coolen, H. K. A. C.; Adolfs, T. J. P.;
McCreary, A. C.; Keizer, H. G.; Wals, H. C.; Veerman, W.; Borst, A. J. M.; de Looff,
W.; Verveer, P. C.; Kruse, C. G. J. Med. Chem. 2005, 48, 1823.
References and notes
1. (a) Di Marzo, V. Pharmacol. Res. 2009, 60, 77; (b) De Petrocellis, L.; Di Marzo, V.
Best Pract. Res. Clin. Endocrinol. Metab. 2009, 23, 1; (c) Di Marzo, V.; Ligresti, A.;
Cristino, L. Int. J. Obesity 2009, 33, S18.
2. (a) Janero, D. R.; Makriyannis, A. Expert Opin. Emerging Drugs 2009, 14, 43; (b) Di
Marzo, V. Drug Discov. Today 2008, 13, 1026; (c) Akbas, F.; Gasteyger, C.; Sjödin,
A.; Astrup, A.; Larsen, T. M. Obes. Rev. 2009, 10, 58.
3. (a) DiMarzo, V. Diabetologia 2008, 51, 1356; (b) Mackie, K. Neuroendocrinology
2008, 20, 10.
4. Ravinet Trillou, C.; Delgorge, C.; Menet, C.; Arnone, M.; Soubrie, P. Int. J. Obes.
2004, 28, 640.
17. (a) 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.
Pharmacol. Biochem. Behav. 1991, 40, 471; (b) Rawls, S. M.; Cabassa, J.; Adler, M.
W.; Geller, E. B. J. Pharmacol. Exp. Ther. 2002, 301, 963.
5. Medicinal chemistry reviews (a) Lee, H.-K.; Choi, E. B.; Pak, C. S. Curr. Topics
Med. Chem. 2009, 9, 482; (b) Jagerovic, N.; Fernandez-Fernandez, C.; Goya, P.
Curr. Topics Med. Chem. 2008, 8, 205.
6. (a) Scheen, A. J.; Paquot, N. Best Pract. Res. Clin. Endocrinol. Metab. 2009, 23, 103;
(b) Samaha, F. F.; Chou, C. M. Obesity 2008, 17, 220.
18. Pardridge, W. M.; Kang, Y. S.; Buciak, J. L. Pharm. Res. 1994, 11, 738.