Discovery of 1-(2,4-dichlorophenyl)-4-ethyl-5-(5-(2-(4-(trifluoromethyl) phenyl)ethynyl)thiophen-2-yl)-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide as a potential peripheral cannabinoid-1 receptor inverse agonist

Article abstract of DOI:10.1002/cmdc.201000246

Fast food nation! The title compound (pictured) was identified as a potential peripheral CB1-acting antagonist based on its negative central effects in CB1R agonist-induced hypothermia and analgesia models, and low brain exposure as indicated by the blood-to-plasma ratio. This compound represents a potential lead in the treatment of metabolic disorders. (Chemical Equation Presented)

Full text of DOI:10.1002/cmdc.201000246

DOI: 10.1002/cmdc.201000246
Discovery of 1-(2,4-Dichlorophenyl)-4-ethyl-5-(5-(2-(4-
(trifluoromethyl)phenyl)ethynyl)thiophen-2-yl)-N-(piperidin-1-yl)-1H-
pyrazole-3-carboxamide as a Potential Peripheral Cannabinoid-1 Receptor
Inverse Agonist
Ming-Shiu Hung, Chun-Ping Chang, Ting-Chieh Li, Teng-Kuang Yeh, Jen-Shin Song, Yinchiu Lin, Chien-
Huang Wu, Po-Chu Kuo, Prashanth K. Amancha, Ying-Chieh Wong, Wen-Chi Hsiao, Yu-Sheng Chao, and
Kak-Shan Shia*[a]
Cannabinoid-1 receptor (CB1R) is one of the most abundant
neuroregulatory receptors in the brain, and it is involved in
regulating feeding and appetite.[1] In addition to expression in
brain, this receptor is also found in the peripheral organs, such
as adipose tissues, muscle, and liver.[2] In sharp contrast, the
structurally closely related cannabinoid-2 receptor (CB2R) is ex-
pressed almost exclusively in the immune system and is pri-
marily involved in immune regulation and neurodegenera-
tion.[3] The therapeutic potential of CB1R antagonists has been
extensively reviewed,[4] and at least one compound (1; also
called rimonabant or SR141716A) has shown clinical evidence
of weight reducing action. However, after its launch in 2006, it
was subsequently withdrawn (2008) in Europe due to severe
psychiatric effects including depression, anxiety and stress dis-
orders. Currently, only two drugs, orlistat and sibutramine, are
available for the long-term treatment of obesity; however,
both have met with moderate success because of their limited
weight-loss efficacy and many accompanying adverse effects,
including high blood pressure and flatulence.[5]
as observed with brain-acting CB1R antagonists (e.g., 1). This
hypothesis, along with increasing evidence that weight loss
and significant reduction in insulin and triglyceride levels
might be achieved as demonstrated with several probable pe-
ripheral CB1R antagonists fed in diet-induced obese (DIO)
mice,[6–10] made up the foundation of proposing peripheral
CB1R as a potential therapeutic molecular target. However,
more in-depth studies on the peripheral CB1R-ligand axis are
required before it can gain firm grounds as a therapeutic
target in treating obesity or other metabolic disorders such as
type II diabetes.[11] Toward this end, discovery of non-brain-
penetrating chemical entities with substantial CB1R binding af-
finity and potency will have a top priority.
Herein, we wish to report that based on the skeleton of a
proven antiobesity agent (3) with significant weight-loss effica-
cy in DIO mice reported previously,[12] a novel series of aryl al-
kynylthiophene derivatives has been developed, leading to the
identification of compound 8 as a highly promising CB1 pe-
ripheral inverse agonist with excellent potency (EC50 =8.5 nm)
and poor brain permeability as indicated by a low brain-to-
plasma ratio (B/P=1/33). The design, synthesis, and prelimina-
ry structure–activity relationship (SAR) data of the series are
described below.
Aryl alkynylthiophenes 4–13 were synthesized in good to
high yield (65–92%) mainly through Pd-mediated coupling of
an appropriate alkyne with the common intermediate 2, readi-
ly prepared following the synthetic procedure reported in the
previous literature,[12] under modified Sonogashira coupling
conditions (Scheme 1).[13] As a typical example, a mixture of
compound 2 and 1-ethynyl-4-(trifluoromethyl)benzene dis-
solved in THF in the presence of PdCl2(PPh3)2/CuI and 2-etha-
nolamine was stirred in a sealed pressure vessel, immersed in
an oil bath at 1008C, overnight to afford product 8 in 86%
yield after chromatographic purification. Compounds thus ob-
tained were subjected to various biological evaluations toward
CB1R and CB2R, the results of which are compiled in Table 1.
As indicated, our synthetic efforts were mainly focused on
replacing the linear alkyl linker of 3 with an array of aryl moiet-
ies. An initial analogue 4 (IC50 =25.8 nm; EC50 =89.8 nm; B/P=
1/23), obtained by substituting the propyl side chain with a
benzene ring, was found to have fourfold improvement in
plasma exposure as compared to 3 (IC50 =6.1 nm; EC50 =
13.8 nm; B/P=1/6), suggesting that a significant impact on in-
creasing plasma exposure might be easily achieved through
appropriate chemical modifications. Along this line, the R
As such, there is still an urgent medical need for the safe
and effective treatment of obesity in modern society. It is
widely accepted that peripherally acting CB1R antagonists
might avoid adverse central nervous system (CNS) side effects,
[a] Dr. M.-S. Hung, Dr. C.-P. Chang, T.-C. Li, Dr. T.-K. Yeh, Dr. J.-S. Song, Y. Lin,
Dr. C.-H. Wu, P.-C. Kuo, Dr. P. K. Amancha, Dr. Y.-C. Wong, W.-C. Hsiao,
Dr. Y.-S. Chao, Dr. K.-S. Shia
Division of Biotechnology and Pharmaceutical Research
National Health Research Institutes
35, Keyan Road, Zhunan Town, Miaoli County 35053 (Taiwan, ROC)
Fax: (+886)37-586456
E-mail: ksshia@nhri.org.tw
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201000246.
ChemMedChem 2010, 5, 1439 – 1443
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1439

MED
Figure 1, is proposed around the pyrazole-5 substitu-
tion, presumably making it extremely sensitive to the
presence of the polar functionalities (i.e., ꢀCOOH).
Thus, approaches to increase polarity by introduc-
ing a polar group(s) on the 5-position were aban-
doned. More interestingly, when the benzene ring of
4 was extended by two methylene units, the result-
ing compound 12 (IC50 =4.4 nm; EC50 =10.3 nm;
CB2R/CB1R=395; B/P=1/31) possessed, not only ex-
cellent biological activities as with parent 3, but a de-
sirable plasma exposure and CB1R selectivity over
CB2R. Also noteworthy is the finding that the intrinsic
property appears to vary with the length of the linker
as indicated with 12 and its counterparts 4 and 11,
wherein the former behaves as a neutral antagonist
and the latter as inverse agonists (Figure 2).[12] In-
spired by favorable biological properties conferred
on 12, its structurally closely related derivatives,
which are being designed to become large enough
not to cross the BBB but remain small enough to
bind tightly to the CB1R, are under active pursuit in
Scheme 1. Reagents and conditions: a) various alkynes, PdCl2(PPh3)2, CuI, 2-ethanolamine,
THF in a sealed pressure vessel immersed in an oil bath at 1008C, 16 h, 65–92%.
group was changed to a p-methoxy benzene ring to provide 5
(IC50 =21.9 nm; EC50 =241.9 nm; B/P=1/26); again, the blood-
to-plasma ratio was greatly enhanced relative to parent 3.
Though analogues 4 and 5 resulted in losing their CB1 binding
affinity and potency by three- to fourfold and six- to 18-fold,
respectively, structural modifications directed toward enhanc-
ing plasma exposure by altering substituents on the benzene
ring appeared to be on the right track.
The introduction of an electron-withdrawing substituent was
Figure 1. The binding mode of rimonabant (1) in the proposed CB1R homol-
then carried out to afford the p-chloride derivative 6 (IC50 =
7.2 nm; EC50 =126.6 nm; B/P=1/18), in which a significant im-
provement in binding affinity and potency was immediately
observed as compared to 5 with a classical electron-donating
p-methoxy moiety. Encouraged by these findings, compounds
7–10, possessing a substituent(s) with stronger electron-with-
drawing ability on the benzene ring, were synthesized. Accord-
ingly, the halogenated analogues thus generated, without ex-
ception, exhibited strong binding affinities and potency
toward CB1R (IC50 =1.1–8.5 nm; EC50 =7.2–38.5 nm); however,
most of them incurred a certain degree of loss in CBR2/CB1R
selectivity, and this drawback will be addressed in the future
structural modifications. An attempt to install an extra carbox-
ylic group, which is supposed to increase polar surface area
(PSA) significantly and reduce blood–brain barrier (BBB) perme-
ability of the global molecule,[14] was also made on the potent
compound 7 (IC50 =6.6 nm; EC50 =38.5 nm; PSA=50.16 ꢁ2), of
which the benzene ring already bears a fluorine atom in the
para position. Unfortunately, the resultant carboxylic acid 13
(IC50 =2274.8 nm; EC50 =1407.6 nm; PSA=90.29 ꢁ2) exhibited a
complete loss of CB1R activity and potency, though the PSA
value, as expected, was practically doubled. This result appears
to be not too surprising in that, as manifested by many histori-
cal cases,[9,15] a hydrophobic pocket consisting of a series of ar-
omatic residues (Trp255/Tyr275/Phe278) as illustrated in
ogy model.
Figure 2. Eu-GTP binding assays of selected alkynylthiophenes 4, 8–12, and
a reference inverse agonist (1) and agonist (14) were conducted at a con-
centration of 10 mm (see section B3 in the Supporting Information). Data are
expressed as the mean ꢁSD of at least three experiments performed in du-
plicate. Statistical significance was assessed by unpaired two-tailed t-test
using the GraphPad Prism program (GraphPad Software, San Diego, CA,
USA). p <0.05 ( ) was considered significant. Compound 12 with the in-
*
duced Eu-GTP binding intensity around the basal level was assigned as a
neutral antagonist (NA); compounds 4 and 8–11 with a significant decrease
in the binding intensity relative to the basal level were defined as inverse
agonists (IA), as was reference 1.
1440
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1439 – 1443

Table 1. Biological evaluation of 5-(5-alkynyl-thiophen-2-yl)pyrazole derivatives on human CB1R and CB2R.
Compd
R
IC50 [nm][a,c]
EC50 [nm][b,c]
IC50 [nm][a,c]
CB2R/CB1R
IP[d]
B/P[e]
CB1R
CB1R
CB2R
1
3[f]
14.6ꢁ1.8
6.1ꢁ2.6
17.3ꢁ1.6
13.8ꢁ4.4
1939.8ꢁ131.1
919.6ꢁ53.1
130
151
IA
IA
3/1
1/6
4
5
25.8ꢁ10.0
89.8ꢁ22.1
1093.7ꢁ182.4
42
18
IA
IA
1/23
1/26
21.9ꢁ2.4
241.9ꢁ14.1
395.5ꢁ51.0
6
7.2ꢁ0.3
6.6ꢁ1.5
8.5ꢁ0.3
4.4ꢁ1.6
1.1ꢁ0.2
126.6ꢁ48.0
38.5ꢁ12
18.5ꢁ8.0
20.8ꢁ3.7
7.2ꢁ1.5
183.7ꢁ85.3
119.0ꢁ28.8
604.9ꢁ101.1
964.8ꢁ39.9
36.6ꢁ13.3
26
18
IA
IA
IA
IA
IA
1/18
1/12
1/33
1/12
1/22
7
8
71
9
219
33
10
11
12
6.6ꢁ1.3
4.4ꢁ0.5
54.0ꢁ0.6
10.3ꢁ3.6
536.4ꢁ150.9
1738.2ꢁ74.2
81
IA
1/15
1/31
395
NA
13
2274.8ꢁ356.5
1407.6ꢁ242.3
>10000
>4
ND[g]
ND[g]
[a] The binding affinity, determined by 50% inhibition of agonist [3H]-CP55940 ([3H]-14) binding to human CB1R or human CB2R, is expressed as the IC50
value. [b] Functional activity, determined by 50% inhibition of Eu-GTP binding to CB1R-overexpressing membrane, is expressed as the EC50 value. [c] Data
are expressed as the mean ꢁSD of at least three independent experiments. [d] Intrinsic property (IP) of each compound was identified at a concentration
of 10 mm as illustrated in Figure 2.[12] [e] The brain (B) to plasma (P) ratio was determined at a time point of 2 h after oral dosing (see section B4 in the Sup-
porting Information). [f] This compound has previously been structurally characterized.[12]. [g] Not determined (ND) because of poor CB1 activity.
our laboratories. Nevertheless, a structurally metabolic liability
for 12, such as benzylic oxidation which could occur easily and
exclude it from BBB penetration, could not be ignored.
ature by acting on CB1R in the hypothalamus; this effect can
be reversed by addition of brain-acting CB1R antagonists (e.g.,
inverse agonist 1). As indicated in Figure 3, a negative central
effect was observed for 8 at a dose of 20 mgkgꢀ1 after treat-
ment with agonist 14 (1 mgkgꢀ1), and a slight reversal, but not
a significant change, of the body temperature reduction was
detected at a dose of up to 50 mgkgꢀ1, suggesting that com-
pound 8 remains primarily in the circulating system and has
much poorer BBB permeability than the positive control 1, in
which a significant reversal of the temperature lowering effect
was observed at a dose as low as 2 mgkgꢀ1. Similar conclu-
Compound 8 (IC50 =8.5 nm; EC50 =18.5 nm; CB2R/CB1R=71;
B/P=1/33), which behaves as an inverse agonist with strong
binding affinity and high potency as well as a low blood-to-
plasma ratio, appears to be a good candidate for testing the
peripheral CB1R-ligand interaction. As such, 8 was initially eval-
uated for its central effects in the CB1R agonist-induced hypo-
thermia model, one of four tests known as “tetrad respons-
es”.[16] Accordingly, treatment with a CB1R agonist, such as
CP55940 (14), alone would induce a reduction in body temper-
ChemMedChem 2010, 5, 1439 – 1443
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1441

MED
sions could also be derived
through another tetrad-response
test, a CB1R agonist-induced an-
algesia model.
Table 2. Pharmacokinetic studies of compounds 1 and 8 in fasted male Sprague Dawley rats.[a]
Compd
Route
Dose
[mgkgꢀ1
tmax
Cmax
[ngmLꢀ1
t1/2
AUC(0–1)
BA
[%]
]
[h]
]
[h]
[hꢂngmLꢀ1
]
As illustrated in Figure 4, com-
pound 8 was administered at a
dose of 20 and 50 mgkgꢀ1, and
was found unable to counteract
the analgesic effect induced by
1
8
oral
oral
10
10
2.8ꢁ1.5
1.3ꢁ0.6
230ꢁ52
5.3ꢁ0.6
6.8ꢁ0.4
801ꢁ389
24
28
3000ꢁ920
12810ꢁ4330
[a] Values indicate the mean ꢁSD (n=3). The experimental protocol is detailed in section B4 of the Supporting
Information. Abbreviations: maximum plasma concentration of drug (Cmax); time taken to reach Cmax (tmax); half-
life (t1/2); area under the curve (AUC); oral bioavailability (BA).
ence in plasma exposure as evidenced by Cmax and AUC values,
thus rendering them ideal compounds, for comparison purpos-
es, to address the peripheral CB1R-acting hypothesis.
In summary, a novel class of aryl alkynylthiophenes has been
designed to target peripheral CB1R to eliminate/minimize CNS
side effects as observed with 1, a typical brain CB1R-acting
agent. The titled compound, 1-(2,4-dichlorophenyl)-4-ethyl-5-
(5-(2-(4-(trifluoromethyl)phenyl)ethynyl)thiophen-2-yl)-N-(piperi-
din-1-yl)-1H-pyrazole-3-carboxamide (8), is tentatively recog-
nized as a peripheral inverse agonist based on its negative
central effects in CB1R agonist-induced hypothermia and anal-
gesia models, and low brain exposure (B/P=1/33) measured at
Figure 3. Compounds 8 (20 and 50 mgkgꢀ1) and 1 (2 mgkgꢀ1) were admin-
istered (po) 1 h prior to ip treatment with agonist 14 (1 mgkgꢀ1) in the CB1
agonist-induced hypothermia model. Body temperature was measured at 30
a time point of 2 h following oral administration. The chronic
study of 8 in DIO mice is currently underway. A full account of
the in vivo studies, with particular emphasis on addressing
issues of central drug accumulation and peripheral metabolic
benefits, as well as further SAR studies on the series, will be re-
ported in due course.
and 65 min after agonist dosing. p <0.05; vs 14. See section B5 of the
*
Supporting Information for full details.
Experimental Section
See the Supporting Information for experimental details on chemi-
cal syntheses, biological assays and pharmacokinetic studies.
All experiments involving animals were approved by the Institu-
tional Animal Care and Use Committee of National Health Research
Institutes (Taiwan) and performed accordingly.
Acknowledgements
Figure 4. Compounds 8 (20 and 50 mgkgꢀ1) and 1 (2 mgkgꢀ1) were admin-
istrated (po) 1 h prior to ip treatment with agonist 14 (1 mgkgꢀ1) in the CB1
agonist-induced analgesia model. Tail flick response was measured 35 min
We are grateful to the National Health Research Institutes and
National Science Council of the Republic of China (grant NSC 96-
2323-B-400-001) for financial support.
after agonist dosing. p <0.05; vs 14. See section B5 of the Supporting In-
*
formation for full details.
Keywords: blood–brain barrier
cannabinoid-1 receptor · central nervous system · diet-induced
obesity
· brain-to-plasma ratio ·
agonist 14 (1 mgkgꢀ1). On the contrary, a significant reversal of
the tail-flick response was seen with brain-acting 1 at a dose
of 2 mgkgꢀ1. Accordingly, compound 8 was assumed to be a
potential peripheral CB1R candidate, and its pharmacokinetic
study was conducted prior to the subsequent long-term effica-
cy studies, results of which are compiled in Table 2.
[1] R. I. Wilson, R. A. Nicoll, Science 2002, 296, 678–782.
[2] A. C. Howlett, F. Barth, T. I. Bonner, G. Cabral, P. Casellas, W. A. Devane,
C. C. Felder, M. Herkenham, K. Mackie, B. R. Martin, R. Mechoulam, R. G.
Pertwee, Pharmacol. Rev. 2002, 54, 161–202.
As listed, it was found that, in addition to displaying similar
in vitro biological activities as described previously, both in-
verse agonists 1 (t1/2 =5.3 h; BA=24%) and 8 (t1/2 =6.8 h; BA=
28%) also had very similar pharmacokinetic profiles such as
half-life (t1/2) and oral bioavailability except for a distinct differ-
[3] a) E. J. Carrier, C. S. Kearn, A. J. Barkmeier, N. M. Breese, W. Yang, K. Nithi-
patikom, S. L. Pfister, W. B. Campbell, C. J. Hillard, Mol. Pharmacol. 2004,
65, 999–1007; b) S. Munro, K. L. Thomas, M. Abu-Shaar, Nature 1993,
365, 61–65.
[4] a) P. Goya, N. Jagerovic, Expert Opin. Ther. Pat. 2000, 10, 1529–1538;
b) P. Cowley, J. Adam, Expert Opin. Ther. Pat. 2002, 12, 1475–1489;
1442
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1439 – 1443

c) D. R. Janero, A. Makriyannis, Expert Opin. Emerging Drugs 2009, 14,
43–65.
[5] a) H. Bays, C. Dujovne, Am. J. Cardiovasc. Drugs 2002, 2, 245–253; b) A.
Kaya, N. Aydin, P. Topsever, Filiz, Oztꢃrk, Dag˘ar, E. KilinÅ, C. Ekmekcioglu,
Biomed. Pharmacother. 2004, 58, 582–587; c) D. W. Clark, M. Harrison-
Woolrych, BMJ 2004, 329, 1316.
453–457; b) 7TM Pharma (http://www.7tm.com/); 7TM Pharma an-
nounced the selection of a new preclinical development candidate,
TM38837, for the treatment of obesity and type II diabetes (http://
www.7tm.com/R-D/Metabolic_Disorders/TM38837.aspx; Last accessed:
July 8, 2010); however, the structure of TM38837 has not been released
yet.
[6] J. LoVerme, Duranti, Tontini, G. Spadoni, Mor, S. Rivara, N. Stella, C. Xu,
G. Tarzia, D. Piomelli, Bioorg. Med. Chem. Lett. 2009, 19, 639–643.
[7] a) J. F. Mcelroy, R. J. Chorvat, (Jenrin Discovery, West Chester, USA), WO/
2007/131219, 2007; b) J. F. Mcelroy, R. J. Chorvat, (Jenrin Discovery, Wil-
mington, USA), WO/2009/033125, 2009; c) Peripheral-acting CB1 antag-
onists discovered by Jenrin Discovery are reported to play a significant
role in the treatment of metabolic disorders with reducing risks of psy-
chiatric adverse effects; Non-brain penetrant CB1 antagonists from Jerin
Discovery show beneficial effects in treating diabetes and obesity (oral
presentation), 26th Annual Scientific Meeting of the Obesity Society,
October 3–7, 2008 (Phoenix, USA); d) JD-5006, a compound with low
brain exposure compared to previously developed CB1 blockers (e.g.,
1), may represent a safer alternative for the treatment of liver disease,
diabetes and related metabolic disorders, and as exemplified, signifi-
cant weight loss may not be required for CB1-mediated antidiabetic ef-
ficacy. The structure of JD-5006 has not yet been released; Peripheral
selective CB1-antagonists as metabolic disorder therapeutics devoid of
psychiatric liabilities (oral presentation), 7th International Symposium
for Chinese Medicinal Chemists, February 1–5, 2010 (Taiwan).
[9] H. K. Lee, E. B. Choi, C. S. Pak, Curr. Top. Med. Chem. 2009, 9, 482–503
and references therein.
[10] M. H. Son, H. D. Kim, Y. N. Chae, M. K. Kim, C. Y. Shin, G. J. Ahn, S. H.
Choi, E. K. Yang, K. J. Park, H. W. Chae, H. S. Moon, S. H. Kim, Y. G. Shin,
S. H. Yoon, Int. J. Obes. 2010, 34, 547–556.
[11] T. M. Fong, S. B. Heymsfield, Int. J. Obes. 2009, 33, 947–955.
[12] S. L. Tseng, M. S. Hung, C. P. Chang, J.-S. Song, C. L. Tai, H. H. Chiu, W. P.
Hsieh, Y. Lin, W. L. Chung, C. W. Kuo, C. H. Wu, C. M. Chu, Y. S. Tung, Y. S.
Chao, K. S. Shia, J. Med. Chem. 2008, 51, 5397–5412.
[13] K. Kobayashi, A. Sugie, M. Takahashi, K. Masui, A. Mori, Org. Lett. 2005,
7, 5083–5085 and references therein.
[14] S. A. Hitchcock, L. D. Pennington, J. Med. Chem. 2006, 49, 7559–7583
and references therein.
[15] J. H. M. Lange, C. G. Kruse, Drug Discovery Today 2005, 10, 693–702 and
references therein.
[16] S. M. Rawls, J. Cabassa, M. W. Adler, E. B. Geller, J. Pharmacol. Exp. Ther.
2002, 301, 963–968.
Received: June 10, 2010
[8] a) J. M. Receveur, A. Murray, J. M. Linget, P. K. Nørregaard, M. Cooper, E.
Bjurling, P. A. Nielsen, T. Hçgberg, Bioorg. Med. Chem. Lett. 2010, 20,
Published online on July 22, 2010
ChemMedChem 2010, 5, 1439 – 1443
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1443