Design, synthesis, and structure-activity relationships of indole-3-heterocycles as agonists of the CB1 receptor

Article abstract of DOI:10.1016/j.bmcl.2010.10.093

Novel indole-3-heterocycles were designed and synthesized and found to be potent CB1 receptor agonists. Starting from a microsomally unstable lead 1, a bioisostere approach replacing a piperazine amide was undertaken. This was found to be a good strategy

Full text of DOI:10.1016/j.bmcl.2010.10.093

Bioorganic & Medicinal Chemistry Letters 21 (2011) 506–509
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and structure–activity relationships
of indole-3-heterocycles as agonists of the CB1 receptor
Angus J. Morrison a, , Julia M. Adam a, , James A. Baker b, Robert A. Campbell b, John K. Clark a,
Jean E. Cottney b, Maureen Deehan b, Anna-Marie Easson a, Ruth Fields a, Stuart Francis a,
Fiona Jeremiah a, Neil Keddie a, Takao Kiyoi a, Duncan R. McArthur a, Karsten Meyer a,
Paul D. Ratcliffe a, Jurgen Schulz a, Grant Wishart a, Kazuya Yoshiizumi a
⇑
⇑
a Department of Chemistry, Merck Research Laboratories, MSD Ltd, Newhouse, Lanarkshire ML1 5SH, UK
b Department of Pharmacology, Merck Research Laboratories, MSD Ltd, Newhouse, Lanarkshire ML1 5SH, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Novel indole-3-heterocycles were designed and synthesized and found to be potent CB1 receptor ago-
nists. Starting from a microsomally unstable lead 1, a bioisostere approach replacing a piperazine amide
was undertaken. This was found to be a good strategy for improving stability both in vitro and in vivo.
This led to the discovery of 24, which had an increased duration of action in the mouse tail flick test
in comparison to the lead 1.
Received 31 August 2010
Revised 18 October 2010
Accepted 19 October 2010
Available online 25 October 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Cannabinoid
CB1 receptor
GPCR agonist
The CB1 cannabinoid receptor is a member of the G-protein
coupled receptor (GPCR) superfamily, which is characterized by
seven transmembrane helices. The CB1 receptor is located primar-
ily in the central nervous system but is also expressed on periphe-
ral neurones. Recently, many studies have revealed that the CB1
receptor is a potential therapeutic target against pain and several
other diseases including glaucoma, traumatic brain injury, multi-
ple sclerosis, and obesity.1 Several lines of evidence have been
reported regarding the analgesic effects of CB1 agonists in both
experimental animal models and clinical studies. Moreover, CB1
by introducing conformational constraints and steric blocking
groups had failed to deliver any enhancements in metabolic stabil-
ity as evidenced by the mouse microsomal stabilities of 1 and 2,
Table 1.2 Therefore, alternative strategies were sought that may
lead to more stable analogs.
Replacement of the amide bond with appropriate bioisosteres is
a classical approach in medicinal chemistry and in particular the
replacement with five-membered ring heterocycles.3 A simple
superposition model of 2 and a generic five-membered heterocycle
showed a good overlap of key features.4 An appropriately posi-
tioned heteroatom can mimic the H-bond acceptor functionality
of the carbonyl, if required, whilst allowing the amines to occupy
the same region. This is exemplified in Figure 2 by the superposition
of compounds 1 and 24. Based on this observation a set of five-
membered ring heterocycles bearing two and three heteroatoms
agonists such as
D
9-THC, one of the major bioactive components
of cannabis, are used clinically as antiemetics in cancer chemother-
apy or appetite stimulants in AIDS patients. However, the classical
cannabinoids represented by
D
9-THC are highly lipophilic and the
administration methods are still limited. We have previously
described indole-3-carboxamide derivatives 1 and 2, Figure 1, as
CB1 agonists that displayed potent antinociceptive activity in the
mouse tail flick test after iv administration, both of which dis-
played a fast onset and short duration of action.2
In following up these compounds we were keen to explore
increasing the duration of action of these types of compounds in
the mouse model. Attempts at improving the metabolic stability
⇑
Corresponding authors. Fax: +44(0)1698 736187.
E-mail addresses: angus.morrison@merck.com (A.J. Morrison), julia.adam@
merck.com (J.M. Adam).
Figure 1. Indole-3-carboxamide CB1 agonists.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.093

A. J. Morrison et al. / Bioorg. Med. Chem. Lett. 21 (2011) 506–509
507
Table 1
Table 2
CB1 agonist activities and mouse microsomal stabilities for indole-3-heterocycles
SAR of amine unit
MLMb
>270
a
Compound
R
pEC50
Compound
R
pEC50
MLMb
HLMc
a
1
2
—
—
7.6
7.9
>270
>270
>270
>270
1
7.6
14
7.7
71
89
20
21
7.4
8.7
53
41
41
64
2
7.9
>270
6
7
5.8
7.1
6.9
6.2
7.7
7.8
7.2
64
22
7.1
6.3
41
14
49
10
17
23
10
11
14
16
18
ntc
ntc
71
a
Values are means of at least three experiments and values varied by no more
than twofold.
b
c
MLM: mouse microsomal clearance (
HLM: human microsomal clearance (
l
l
L/min/mg).
L/min/mg).
Table 3
SAR of the indole 7-position
>120
128
a
Values are means of at least three experiments and values varied by no more
than twofold.
b
MLM: mouse microsomal clearance (lL/min/mg).
nt: not tested.
c
a
Compound
R
pEC50
MLMb
HLMc
1
2
21
24
25
26
—
—
OMe
Et
Cl
7.6
7.9
8.7
8.5
7.8
7.3
>270
>270
41
34
17
>270
>270
64
74
35
F
24
41
a
Values are means of at least three experiments and values varied by no more
than twofold.
b
MLM: mouse microsomal clearance (
HLM: human microsomal clearance (
l
l
L/min/mg).
L/min/mg).
c
prepared in a similar fashion using methylformyl chloroformate
as the coupling partner to furnish the ester derivatives, 12 and
13.6 Reduction, mesylation, and displacement with diethylamine
furnished the desired oxazole 10 and thiazole 11.
Figure 2. Superposition model of 1 (green) and 24 (orange).
and incorporating a simple tertiary amine were targeted for
synthesis.
Synthesis of the 1,2,4-thiadiazole 14 was achieved by initial for-
mation of the oxathiazolone, followed by microwave assisted [3+2]
cycloaddition in the presence of ethyl cyanoformate to provide the
ester derivative 15.7 Reduction, mesylation, and displacement with
diethylamine provided the thiadiazole 14. The 1,3,4-oxadiazole 16
was prepared by initial formation of the hydrazide followed by
dehydration with Burgess’ reagent to provide the chloromethyl
derivative 17,8 subsequent displacement with diethylamine pro-
vided the desired oxadiazole 16. Finally, the 1,2,4-oxadiazole 18
was synthesized by initial formation of the amidoxime 19, which
was achieved by initial dehydration of the primary amide 4 to
Compounds were synthesised as described in Scheme 1. Briefly,
starting from the previously described carboxylic acid 3,2a conver-
sion to the primary amide 4 was achieved via the acid chloride,
subsequent treatment with Lawesson’s reagent furnished the
thioamide 5. The oxazole 6 and thiazole 7 were readily prepared
from the requisite amide or thioamide by initially reacting with
1,3-dichloroacetone to deliver the chloromethyl oxazole 8 and
chloromethyl thiazole 9,5 followed by simple displacement with
diethylamine. The isomeric oxazole 10 and thiazole 11 were

508
A. J. Morrison et al. / Bioorg. Med. Chem. Lett. 21 (2011) 506–509
Scheme 1. Reagents and conditions: (a) (i) (COCl)2, DCM, (ii) NH3 (gas), DCM, 81%; (b) Lawesson’s reagent, toluene, rt, 74%; (c) 1,3-dichloroacetone, toluene, microwave
150 °C, 81% from 4 and 1,3-dichloroacetone, toluene, 40 °C, 61% from 5; (d) HNEt2, THF, microwave 150 °C, 84% from 8 and HNEt2, MeCN, microwave 150 °C, 51% from 9;
(e) methyl formylchloroacetate, DMA, microwave 150 °C, 76% from 4 and 49% from 5; (f) LiAlH4, THF, 0 °C, 42% from 12 and 94% from 13; (g) (i) CH3SO2Cl, iPr2NEt, DCM, 0 °C,
(ii) HNEt2, THF, microwave 150 °C, 20% for 10 over two steps and (i) CH3SO2Cl, Et3N, DCM, (ii) HNEt2, THF, MeCN, microwave 160 °C, 40% for 11 over two steps;
(h) chlorocarbonylsulfonyl chloride, CHCl3, reflux, 56%; (i) ethyl cyanoformate, m-xylene, microwave 200 °C, 71%; (j) NaBH4, MeOH, 86%; (k) CH3SO2Cl, NEt3, DCM, quant.;
(l) HNEt2, THF, microwave 150 °C, 61%; (m) (i) (COCl)2, DMF (cat.), DCM, (ii) chloroacetylhydrazideꢀHCl, Et3N, DCM, 30% over two steps; (n) N-(triethylammoniumsulfo-
nyl)carbamate, THF, microwave 150 °C, 71%; (o) HNEt2, THF, microwave 150 °C, 86%; (p) TFAA, NEt3, dioxane, 88%; (q) hydroxylamine hydrochloride, iPr2NEt, EtOH, 85 °C,
66%; (r) benzyl 2-(diethylamino)acetate, NaH, THF, rt to 90 °C, 20%.
the nitrile and subsequent treatment with hydroxylamineꢀHCl.
With the amidoxime 19 in hand the 1,2,4-oxadiazole 18 was read-
ily prepared in one step from benzyl 2-(diethylamino) acetate un-
der basic conditions.
change was observed in both mouse and human liver microsomes
within this series. However, all had enhanced stability in compar-
ison to the original carboxamide compounds, Table 2. Having opti-
mized the heterocycle and amine portions of the molecule,
attention turned to the 7-position of the indole.
Synthesis of these compounds was readily achieved using the
method outlined in Scheme 1 and the appropriate 7-substituted in-
dole. Replacement of the 7-methoxy group for the similarly sized
ethyl was well tolerated with comparable potency and stability.
Replacement with smaller halogen groups resulted in a loss of po-
tency although microsomal stability was improved in both cases,
Table 3. With these results in hand, compound 24, was selected
The prepared compounds were tested for CB1 agonist activity
using CHO cells doubly transfected with human CB1 and luciferase
reporter gene.9 As shown in Table 1, the hypothesis that these het-
erocycles would provide bioisosteric replacements for the pipera-
zine amide was shown to be correct. The new compounds did
indeed retain CB1 agonist activity and in particular the 1,2,4-thia-
diazole 14 and 1,2,4-oxadiazole 16 had comparable potency. Fur-
thermore, and much to our delight, compounds 6, 7, and 14 had
enhanced metabolic stability in mouse liver microsomes in com-
parison to the original indole-3-carboxamides.
Table 4
DMPK profile of 24
With these results in hand, we focused our attention on the
1,2,4-thiadiazole motif which offered the best combination of po-
tency and metabolic stability within the series. An SAR study of
the amine portion of the molecule was carried out. Synthesis of
these compounds was readily achieved using the method outlined
in Scheme 1 and replacing the diethylamine with the appropriate
amine (scheme not shown). Replacement of the diethyl group for
the slightly smaller dimethylamine unit 20, resulted in a slight
drop in potency. Constraining the diethyl unit into a pyrrolidine
ring, 21 resulted in a marked 10-fold improvement in potency.
Increasing the ring size further to a piperidine 22 or N-ethyl piper-
azine 23 resulted in a drop in potency suggesting that SAR in this
region may be tight. With regard to metabolic stability, little
24
1
Microsomal stability, mouse CLint
PK (ICR mouse, 0.5
Vehicle
Plasma Cmax (ng/ml; t = 0.05 h)
AUCplasma, iv (h ng/ml)
Clearance (ml/min/kg)
T1/2 elimination (h)
Vss (L/kg)
(l
L/min/mg)
34
>270
lmol/kg, iv)
5% DMSO/saline
Water
593
253
77
0.32
2.0
667
124
28
1.2
1.4
Brain penetration (ICR mouse) same studies as above
Brain Cmax (ng/g)
Brain tmax (h)
32
1
0.05
1619
0.05
2.7
Brain:plasma Cmax ratio

A. J. Morrison et al. / Bioorg. Med. Chem. Lett. 21 (2011) 506–509
509
Table 5
Effects of 1 and 24 following intravenous dosing in the tail flick test in the mouse
Test compound
Dose (
l
mol/kg)
%MPEa
ED50
(l
mol/kg)
DofAb (min)
Vehicle (10% Tween 80)
Compound 1
Compound 1
10 ml/kg
0.1
0.3
2.9 2.8
14.5 3.7
72.3 9.7c
100 0c
Compound 1
1.0
Compound 1
0.2
50
Vehicle (10% Tween 80)
Compound 24
Compound 24
Compound 24
Compound 24
CP 55,940
10 ml/kg
0.35
1.17
2.01 4.9
29.3 8.3
54.1 15.4
100 0c
3.52
0.8
0.11
1.7
160
110
90
WIN 55,212-2
a
%MPE: %maximum possible effect, see Ref. 11.
Dofa: duration of action, time taken for the tail flick latency to return within the mean plus three standard deviations of the
b
baseline value following a twice ED50 dose.
c
Denotes that the effect was significantly greater than in vehicle treated mice (P <0.01, Dunn’s post-test)
D., Lawton, G., Eds.; Elsevier: Amsterdam, 2006; Vol. 44, pp 207–329;
(b)Pertwee, R. G., Ed.Handbook of Experimental Pharmacology; Springer:
Heidelberg, 2005; Vol. 168, (c) Huffman, J. W.; Padgett, L. W. Curr. Med.
Chem. 2005, 12, 1395.
for further profiling based on its overall potency and stability in
microsomes.
Compound 24 exhibited high affinity for both CB1 (pKi 8.2) and
CB2 (pKi 8.5) cannabinoid receptors, as determined by radioligand
competition binding assays using [3H]-CP 55,940 binding to either
hCB1 or hCB2 receptors expressed in insect Sf9 membranes. Fur-
thermore, compound 24 was screened at 10ꢁ5 M against a panel
of 63 unrelated molecular targets and showed no activity against
any other targets.
2. (a) Adam, J.; Cairns, J.; Caulfield, W.; Cowley, P.; Cumming, I.; Easson, M.;
Edwards, D.; Ferguson, M.; Goodwin, R.; Jeremiah, F.; Kiyoi, T.; Mistry, A.; Moir,
E.; Morphy, R.; Tierney, J.; York, M.; Baker, J.; Cottney, J.; Houghton, A.;
Westwood, P.; Walker, G. Med. Chem. Commun. 2010, 1, 54; (b) Moir, E. M.;
Yoshiizumi, K.; Cairns, J.; Cowley, P.; Ferguson, M.; Jeremiah, F.; Kiyoi, T.;
Morphy, R.; Tierney, J.; Wishart, G.; York, M.; Baker, J.; Cottney, J. E.; Houghton,
A. K.; McPhail, P.; Osprey, A.; Walker, G.; Adam, J. A. Bioorg. Med. Chem. Lett.
2010, 20, 7327.
The in vitro and in vivo DMPK profile of 24 and 1 (for compar-
ison) is summarized in Table 4. In comparison to the carboxamide,
24 is cleared less rapidly as was predicted from the in vitro micro-
somal stability results. Furthermore, 24 had a slightly lower vol-
ume of distribution but overall a longer half-life than 1. In stark
contrast, the brain levels were markedly different with 24 having
3. Wermuth, C. G.; Ciapetti, P.; Giethlen, B.; Bazzini, P. In Comprehensive Medicinal
Chemistry II; Taylor, J. B., Triggle, D. J., Eds.; Elsevier: Oxford, 2006; Vol. 2, pp
649–711.
4. Superposition analysis was carried out with SYBYL7.3 and GASP as distributed by
Tripos Inc., 1699 South Hanley Road, St. Louis, MO 63144-2913, USA. Image
created with PyMOL 0.99, DeLano, M. L. The PyMOL Molecular Graphics System
(2008) DeLano Scientific LLC, Palo Alto, CA, USA. http://www.pymol.org.
5. Einsiedel, J.; Thomas, C.; Hübner, H.; Gmeiner, P. Bioorg. Med. Chem. Lett. 2000,
10, 2041.
a much lower B:P ratio and a delayed brain tmax
.
6. Srivastava, P. C.; Pickering, M. V.; Allen, L. B.; Streeter, D. G.; Campbell, M. T.;
Witowski, J. T.; Sidwell, R. W.; Robins, R. K. J. Med. Chem. 1977, 20, 256.
7. (a) Morrison, A. J.; Paton, R. M.; Sharp, R. D. Synth. Commun. 2005, 35, 807; (b)
Fordyce, E. A. F.; Morrison, A. J.; Sharp, R. D.; Paton, R. M. Tetrahedron 2010, 66,
7192.
8. Brain, C. T.; Paul, J. M.; Loong, Y.; Oakley, P. J. Tetrahedron Lett. 1999, 40, 3275.
9. Price, M. R.; Baillie, G. L.; Thomas, A.; Stevenson, L. A.; Easson, M.; Goodwin, R.;
McLean, A.; McIntosh, L.; Goodwin, G.; Walker, G.; Westwood, P.; Marrs, J.;
Thomson, F.; Cowley, P.; Christopoulos, A.; Pertwee, R. G.; Ross, R. A. Mol.
Pharmacol. 2005, 68, 1484.
Based on its enhanced stability in vivo, 24 was progressed for test-
ing in the mouse tail flick test.10,11 Intravenous administration of 24
increased tail flick latency in a dose-dependent manner, Table 5,
but with a lower ED50 than the starting compound 1, this may be
due to the differences in CNS penetration between the two com-
pounds. Having demonstrated that the compound is efficacious in
this model the duration of action was measured, the time taken for
the tail flick latency to return within the mean plus three standard
deviations of the baseline value following a twice ED50 dose. Pleas-
ingly it was found that the duration of action of 24 was 160 min over
three times longer than the duration of action observed with 1, Table
5 and that our strategy of improving metabolic stability by replace-
ment of the carboxamide with an bioisosteric heterocycle had real-
ized the goal of improving duration of action in vivo.
In summary, a series of indole-3-heterocyclic derivatives were
synthesized and found to be agonists of the CB1 receptor. In addi-
tion, these heterocyclic derivatives had improved mouse micro-
somal stability when compared to our previously reported
indole-3-carboxamides. A systematic SAR study of the heterocycle,
amine, and 7-position of the indole identified a highly potent and
relatively stable CB1 receptor agonist 24. This compound was
found to have an improved duration of action after intravenous
administration in the mouse tail flick test, a preclinical model of
nociception.
10. Whiteside, G. T.; Gottshall, S. L.; Boulet, J. M.; Chaffer, S. M.; Harrison, J. E.;
Pearson, M. S.; Turchin, P. I.; Mark, L.; Garrison, A. E.; Valenzano, K. J. Eur. J.
Pharmacol. 2005, 528, 65.
11. Determination of tail flick latency in mice; male ICR mice weighing 20–34 g
were trained to sit still in a tail flick apparatus (Ugo Basile, Italy) whilst tail flick
latency was measured. The tail was exposed to a focused beam of radiant heat
at a point approximately 2.5 cm from the tip. Tail flick latency was defined as
the interval between the appliance of the thermal stimulus and withdrawal of
the tail. A 12 second cut-off was employed to prevent tissue damage. Four
groups of eight mice were treated with vehicle or one of three doses of the test
compound. Tail flick latency was measured before administration of the test
compound and at 20, 40, and 60 min after compound administration. Data
were plotted as mean SEM and expressed as % maximum possible effect
(%MPE).
ðpost drug latency ꢁ baseline latencyÞ
%MPE ¼
ꢂ 100
ðCut off latency ꢁ baseline latencyÞ
The time of maximum effect for each mouse in the two top dose groups was
determined and these values averaged to calculate the mean time of maximum
effect. For analytical purposes Tmax was defined as the time point closest to this
averaged value. Tmax data were then compared between groups using the Krus-
kal–Wallis one-way analysis of variance, a non-parametric statistical test. If sta-
tistical significance (P <0.05) was observed with this test, the vehicle group and
each of the treatment groups were compared using the non-parametric Dunn’s
test (Unistat 5.0 software).
References and notes
1. Recent reviews for cannabinoid receptor agonists: (a) Adam, J.; Cowley, P. M.;
Kiyoi, T.; Morrison, A. J.; Mort, C. J. W. In Progress in Medicinal Chemistry; King, F.