α2 Adrenoceptor agonists as potential analgesic agents. 2. Discovery of 4-(4-imidazo)-1,3-dimethyl-6,7-dihydrothianaphtheneas a high-affinity ligand for the α(2D) adrenergic receptor [1]

Full text of DOI:10.1021/jm990569e

© Copyright 2000 by the American Chemical Society
Volume 43, Number 5
March 9, 2000
Please click below to access the corrected version.
Communications to the Editor
Sch em e 1
r2 Ad r en ocep tor Agon ists a s P oten tia l
An a lgesic Agen ts. 2. Discover y of
4-(4-Im id a zo)-1,3-d im eth yl-6,7-d ih yd r o-
th ia n a p h th en e a s a High -Affin ity Liga n d
for th e r2D Ad r en er gic Recep tor
Tina M. Ross, Michele C. J etter, Mark E. McDonnell,
Robert E. Boyd, Charlene D. Connelly,
Rebecca P. Martinez, Martin A. Lewis,‡
Ellen E. Codd, Robert B. Raffa,† and Allen B. Reitz*
Drug Discovery Division,
The R. W. J ohnson Pharmaceutical Research Institute,
Spring House, Pennsylvania 19477
Received November 15, 1999
In tr od u ction . The treatment of pain continues to be
the subject of considerable pharmaceutical and clinical
research.1 R2 Adrenoceptor agonists are intriguing
because they produce clinically effective analgesia that
is devoid of opioid-related side effects.2 The R2 adren-
ergic agonist clonidine (1) is prescribed for the epidural
treatment of severe pain, and the R2 agonist dexme-
detomidine (2) is used in veterinary practice and for
humans in several countries outside the United States.3
The primary site of antinociceptive action for R2 adr-
energic analgesia is in the dorsal horn of the spinal
cord.2a R2 Adrenergic receptors are localized largely in
the presynaptic membrane, and activation of these
receptors inhibits the release of pain-inducing neu-
rotransmitters such as substance P, glutamic acid, and
CGRP. We have recently reported the potent R2 adren-
ergic receptor affinity and analgesic activity associated
with (imidazomethyl)oxazoles and -thiazoles, such as 3
(RWJ -37210).4 In addition, we have investigated a large
number of (imidazomethyl)thiophenes, such as 4, that
have a similar biological profile.5 Analyzing the pre-
ferred conformations of 4 by molecular modeling re-
vealed significant torsional freedom for rotation of the
two single bonds between the imidazole and thiophene.
To restrict the rotation of this molecule, we extended
and combined the 2-ethyl substituent with the bridging
carbon atom to obtain the (imidazo)thianaphthene
chemical series (e.g. 8 and 9, Scheme 1). These com-
pounds are potent R2 adrenergic receptor ligands, and
certain members of the series are very active in animal
models of antinociception. 4-(4-Imidazo)-6,7-dihydro-
thianaphthene (14) has a favorable biological profile in
vivo and is a prototype for this class of analgesic agents.
* To whom correspondence should be addressed: Allen B. Reitz.
Tel: 215-628-5615. Fax: 215-628-4985. E-mail: areitz@prius.jnj.com.
‡ Present affiliation: Bristol-Myers Squibb Pharmaceutical Research
Institute.
† Present affiliation: Department of Pharmacy, Temple University,
Philadelphia, PA.
10.1021/jm990569e CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/17/2000

766 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 5
Communications to the Editor
Sch em e 2
Sch em e 4
thetic convenience in the preparation of 20 in order to
direct the position of keto ring fusion.6 Ketone 20 was
condensed with the Grignard reagent derived from 6 to
yield 21 (94% yield). Dehydration of 21 was accom-
plished by preparing the acetate and elimination of the
acetoxy group with 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU). The trityl group was then removed under the
mild conditions of formic acid at room temperature10
producing target 22 (50% overall yield). Compound 21
was also reduced using catalytic hydrogenation with
palladium hydroxide to give saturated derivative 23
(39%).
Sch em e 3
P h a r m a cology. The R2 adrenergic receptors are
members of the seven-transmembrane G-protein-cou-
pled (GPCR) family and are unusual because they have
a very large third intracellular loop.11 There are three
R2 adrenoceptor subtypes that have been identified and
cloned in humans: R2A, R2B, and R2C. The pharmacologi-
cally defined R2D adrenergic receptor is the rat (and
mouse) molecular ortholog of the human R2A subtype
and is often referred to as rat R2A. Because R2D and R2A
can be differentiated pharmacologically, there is value
in testing both to better understand the predictive value
of rodent models for efficacy in the clinic. Pharmacologi-
cal ranking of R2 adrenergic agents comparing receptor
subtype affinity with in vivo activity in rats and mice,
as well as studies with knock-out mice in which the R2A
adrenergic receptor subtype has been inactivated, sug-
gests that the rodent R2A (R2D) adrenergic receptor is
the one primarily responsible for antinociception.12 The
functional roles of the R2B and R2C receptors are not
presently well-established.
Compounds 8, 9, 14, 15, 19, 22, and 23 were tested
for in vitro receptor binding and in vivo biological
activity in animal models predictive of analgesic activity
in humans (Table 1). The R2D adrenergic receptor
binding assay from rat was the primary in vitro test,
using p-aminoclonidine as the displaceable ligand.13 The
compounds were also evaluated at the rat R1 adrenergic
receptor13 and in cells containing native R2A, R2B, and
R2C adrenergic receptors using RX 821002 as the
ligand.14
Ch em ist r y. Several complementary methods were
used to prepare the (imidazo)thianaphthenes. Reaction
of thienyl ketone 56 with the Grignard reagent7 derived
from 4-iodo-1-tritylimidazole (6) gave alcohol 7 (84%;
Scheme 1). This material was either treated with HCl
in MeOH to yield unsaturated derivative 8 (20%) or
reduced to fully saturated 9 by hydrogenation over
palladium on carbon (29%). Although this method was
used for the preparation of 14 and 15, we also developed
a strategy involving vinyl triflate coupling as shown in
Scheme 2. Thienyl ketone 10 was converted to vinyl
triflate 11 in 90% yield by using triflic anhydride and a
hindered pyridine base. Coupling of either the 4-imidazo
zinc reagent 12 or the 4-(tributylstannyl)imidazole (not
shown) gave alkene 13 (78% when using zincate 12).8
Cleavage of the trityl group with HCl in MeOH provided
target 14, which was further hydrogenated reducing the
alkene to afford saturated congener 15 (88%).
Parent unsaturated thianaphthene 19 was prepared
starting with condensation of thienyl ketone 16 and
(dioxanyl)ethyl Grignard reagent 17 (Scheme 3).9 Re-
sulting alcohol 18 was obtained in 90% yield. Treatment
with trimethylsilyl triflate promoted a Friedel-Crafts
reaction and dehydration to give target 19 (40%).
To complete the series, so that all possible positions
of thiophene ring fusion were represented, we used
thiophene 20 as a starting ketone (Scheme 4). The 2,5-
methyl groups were present originally only as a syn-
The compounds were tested for potential antinoci-
ceptive activity initially at a screening dose of 30 mg/

Communications to the Editor
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 5 767
Ta ble 1. Biological Activity of (Imidazo)benzothiophene Derivatives and Reference Compounds
in vivo antinociceptive testingb
adrenergic receptor subtype affinity (Ki values, nM)a
MAIT, % inhib
at 30 mg/kg po
RAIT, ED50
(mg/kg, po)
compd
R2D
3.0
0.35
1.5
R2A
224
87%c
254
R2B
R2C
ND
54%c
ND
R1
R1/R2D
8
9
14
940
ND
621
399
91
443
133
260
295
60
73
ND
ND
100
15.1
[7.2, 40.3]
0.9
15
0.56
27
99
172
110
196
100
[0.54, 1.53]
ND
ND
19
22
23
2.6
0.0086
2.9
277
25
68
663
100
141
ND
26
205
343
110
48
132
11627
17
60
40
93
7.9
[3.9, 13.5]
0.10
[0.0012, 0.5]
0.12
1
2
0.39
853
160
116
5
297
333
0.05 mg/kgd
[0.09, 0.01]
0.09 mg/kgd
[0.04, 0.17]
0.015
[0.06, 0.21]
a
b
Receptor binding Ki’s were determined using 5-8 concentrations in triplicate. MAIT, mouse abdominal irritant test; RAIT, rat
abdominal irritant test; 95% confidence limits shown in brackets. Screening in the MAIT at 30 mg/kg po was conducted on groups of 15
animals. c Percent inhibition of binding at 1 µM. ED50 values with 95% confidence limits in brackets.
d
Ta ble 2. Profile of Compound 14 (RWJ -52353)
values varied between 17 (for 23) and >10 000 (for 22).
As binding to the R1 adrenoceptor may result in un-
wanted cardiovascular effects, we have tried to minimize
this interaction. Typical R1/R2D Ki values for these
compounds range between 130 and 300.
All of the compounds show at least some activity in
the MAIT at the oral screening dose of 30 mg/kg. The
most potent compounds in the MAIT (14, 15, and 23)
were then studied further in the RAIT. Unsaturated
derivative 14 had an ED50 of 15.1 mg/kg po in the RAIT,
and alkenes 15 and 23 were more potent (ED50 ) 0.9
and 7.9 mg/kg).
We chose 4-(4-imidazo)-6,7-dihydrothianaphthene 14
for further in vivo evaluation. It is a potent R2D adren-
ergic receptor ligand (Ki ) 1.5 nM) with an R1/R2D Ki
ratio of 295. Further, 14 is achiral, obviating the need
for the resolution of enantiomers or chiral synthesis. A
listing of the analgesic testing of 14 (RWJ -52353) is
given in Table 2. In addition to activity in the MAIT
and RAIT upon oral administration, it is also active in
more stringent tests such as the mouse hot-plate and
tail-flick tests, albeit at higher doses. There is no clear
correlation in the series between R2D receptor affinity
and analgesic activity. For example, 22 with high R2D
adrenergic receptor affinity shows relatively little an-
algesic activity, with 15 showing the greatest degree of
analgesia and being the third most potent at the R2D
adrenergic receptor. This type of discrepancy between
binding affinity at an in vitro receptor and biological
activity upon oral administration is often due to dif-
ferential pharmacokinetics of the compounds being
evaluated. In addition, the relative agonist activity of
the compounds was inferred from in vivo and functional
testing, so it is possible that some are not full R2D
adrenergic agonists.
R2D adrenergic receptor
R1 adrenergic receptor
mouse abdominal irritant
1.5 nM Ki
443 nM Ki
11.6 mg/kg poa
[6.8, 20.0]
rat abdominal irritant
mouse 48 °C hot plate
mouse tail flick
15.1 mg/kg po
[7.2, 40.3]
25.9 mg/kg po
[18.6, 38.2]
100.2 mg/kg po
[48.0, 223.6]
a
ED50 value [95% confidence limits].
kg po in the mouse abdominal irritant test (MAIT) using
acetylcholine bromide as the irritant.15 If a compound
showed >75% inhibition of the response, it was tested
further in the rat air-induced abdominal irritant test
(RAIT).16 Compound 14 has been examined in a variety
of additional animal models including the mouse hot-
plate17 and tail-flick assays (Table 2).18
Resu lts a n d Discu ssion . All of the compounds
tested showed excellent receptor affinity for the R2D
adrenergic receptor subtype. Saturated derivatives 9
(0.35 nM Ki) and 15 (0.56 nM Ki) were ca. 3-10 times
more potent than their singly unsaturated counterparts
8 and 14. Fully aromatic thianaphthene 19 also had
good affinity (2.6 nM Ki). Surprisingly, alkene 22 was
exquisitely active at the R2D adrenergic receptor, show-
ing a very high affinity of 0.0086 nM Ki. In this case,
the saturated derivative (23) was less active (2.9 nM
Ki). Alkenes 8, 14, and 22 are chemically stable over
prolonged periods (>3 months) at room temperature,
showing no signs of air oxidation to their aromatic
counterparts such as 19. The very high affinity of 22
may be due in part to a butressing effect imparted by
the methyl groups on the thiophene ring.
The compounds of the present study were tested in
cell lines natively expressing other R2 adrenergic recep-
tor subtypes. The compounds displayed higher affinity
for the R2A subtype, relative to the R2B, by a factor of
2-4-fold, and 15 and 23 had lower affinity for the R2C
adrenergic subtype relative to the R2B. Alkene 22,
however, had equal affinity at R2A and R2C and less at
2B. In general, there was only modest selectivity for
the R2 adrenergic subtypes found in humans.
These compounds bound the rat R1 less avidly than
the R2D adrenergic receptors. The ratio of R1/R2D Ki
Con clu sion . The (imidazo)thianaphthenes are a new
class of R2 adrenergic agents that display potent recep-
tor binding affinity and are active in animal models of
analgesia. 4-(4-Imidazo)-1,3-dimethyl-6,7-dihydrothia-
naphthene (22) is an exceedingly potent ligand for the
R2D adrenergic receptor (Ki ) 0.0086 nM) and is >10 000-
R
fold selective relative to the R1 adrenergic receptor. This
compound is more potent than dexmedetomidine at R2D
and is a useful pharmacological tool drug for studies
involving the R2D adrenergic receptor.

768 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 5
Communications to the Editor
(6) Cagniant, P.; Guy, M.; Cagniant, D. Condensed Sulfur Hetero-
cycles. XLV. Synthesis of Some Semi-Aromatic Ketones Fused
with Thiophene Rings. Bull. Soc. Chim. Fr. 1970, 1, 322-331.
(7) Turner, R. M.; Lindel, S. D.; Ley, S. V. A Facile Route to
Imidazoyl-4-yl Anions and Their Reactions with Carbonyl
Compounds. J . Org. Chem. 1991, 56, 5739-5740.
(8) J etter, M. C.; Reitz, A. B. Synthesis of 4-Substituted Imidazoles
via Palladium-Catalyzed Cross-Coupling Reactions. Synthesis
1998, 829-831.
Ack n ow led gm en t. We thank Drs. Ellen W. Baxter,
J ohn R. Carson, and David A. Demeter for helpful
discussions.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails. This information is available free of charge via the
Internet at http://pubs.acs.org.
(9) Loozen, H. J . J .; Godefroi, E. F. Benzo[b]thiophenes from Thio-
phenes. A Facile Approach. J . Org. Chem. 1973, 38, 1056-1057.
(10) Commercon, A.; Ponsinet, G. Diastereoselective Chlorocyclo-
functionalization of N-Allylic Trichloroacetamides; Synthesis of
an Analogue and Potential Precursor of RP49532. Tetrahedron
Lett. 1990, 31, 3871-3874.
(11) (a) Docherty, J . R. Subtypes of Functional R1- and R2-Adreno-
ceptors. Eur. J . Pharmacol. 1998, 361, 1-15. (b) Hieble, J . P.;
Ruffolo, R. R., J r.; Sulpizio, A. C.; Naselsky, D. P.; Conway, T.
M.; Ellis, C.; Swift, A. M.; Ganguly, S.; Bergsma, D. J . Functional
Subclassification of R2-Adrenoceptors. Pharmacol. Commun.
1995, 6, 91-97.
Note Ad d ed a fter ASAP P ostin g. The version of
this paper that appears in the printed March 9 issue is
missing Scheme 4 and Tables 1 and 2. The ASAP
posting on February 17, 2000, and the current web
version are correct. An Addition and Correction for the
print version is available as manuscript number
J M000128R.
Refer en ces
(12) (a) Millan, M. J .; Bervoets, K.; Rivet, J .-M.; Widdowson, P.;
Renouard, A.; Le Marouille-Girardon, S.; Gobert, A. Multiple
Alpha-2 Adrenergic Receptor Subtypes. II. Evidence for a Role
of Rat Ralpha-2A Adrenergic Receptors in the Control of Nocicep-
tion, Motor Behavior, and Hippocampal Synthesis of Noradren-
aline. J . Pharmacol. Exp. Ther. 1994, 270, 958-972. (b) Stone,
L. A.; MacMillan, L. B.; Kitto, K. F.; Limbird, L. E.; Wilcox, G.
L. The R2a Adrenergic Receptor Subtype Mediates Spinal Anal-
gesia Evoked by R2 Agonists and Is Necessary for Spinal
Adrenergic-Opioid Synergy. J . Neurosci. 1997, 17, 7157-7165.
(13) Simmons, R. M. A.; J ones, D. J . Binding of [3H-]-Prazosin and
[3H-]-p-Aminoclonidine to R-Adrenoceptors in Rat Spinal Cord.
Brain Res. 1988, 445, 338-349.
(14) (a) Bylund, D. B.; Ray-Prenger, C.; Murphy, T. J . Alpha-2A and
Alpha-2B Adrenergic Receptor Subtypes: Antagonist Binding
in Tissues and Cell Lines Containing Only One Subtype. J .
Pharmacol. Exp. Ther. 1988, 245, 600-607. (b) Murphy, T. J .;
Bylund, D. B. Characterization of Alpha-2 Adrenergic Receptors
in the OK Cell, an Opossum Cell Line. J . Pharmacol. Exp. Ther.
1988, 244, 571-578.
(1) (a) Kowaluk, E. A.; Arneric, S. P. Novel Molecular Approaches
to Analgesia. Annu. Rep. Med. Chem. 1998, 33, 11-20. (b)
Millan, M. J . The Induction of Pain: An Integrative Review.
Prog. Neurobiol. 1999, 57, 1-164.
(2) (a) Kingery, W. S.; Davies, M. F.; Maze, M. Molecular Mecha-
nisms for the Analgesic Properties of Alpha-2 Adrenergic Ago-
nists. In Molecular Neurobiology of Pain; Borsook, D., Ed.; IASP
Press: Seattle, 1997; pp 275-304. (b) Zhang, X.; Yao, X.-T.;
Dalton, J . T.; Shams, G.; Lei, L.; Patil, P. N.; Feller, D. R.; Hsu,
F.-L.; Cliff, G.; Miller, D. D. Medetomidine Analogues as
R2-Adrenergic Ligands. 2. Design, Synthesis, and Biological
Activity of Conformationally Restricted Naphthalene Derivatives
of Medetomidine. J . Med. Chem. 1996, 39, 3001-3013. (c) Munk,
S. A.; Harcourt, D.; Ambrus, G.; Denys, L.; Gluchowski, C.;
Burke, J . A.; Kharlamb, A. B.; Manlapaz, C. A.; Padillo, E. U.;
Runde, E.; Williams, L.; Wheeler, L. A.; Garst, M. E. Synthesis
and Evaluation of 2-[(5-Methylbenz-1-ox-4-azin-6-yl)imino]imi-
dazoline, a Potent, Peripherally Acting R2 Adrenoceptor Agonist.
J . Med. Chem. 1996, 39, 3533-3538. (d) J eon, Y. T.; Luo, C.;
Forray, C.; Vaysse, P. J .-J .; Branchek, T. A.; Gluchowski, C.
Pharmacological Evaluation of UK-14,304 Analogues at Cloned
Human R Adrenergic Receptors. Bioorg. Med. Chem. Lett. 1995,
5, 2255-2258.
(3) (a) Patel, S. S.; Dunn, C. J .; Bryson, H. M. Epidural Clonidine:
A Review of its Pharmacology and Efficacy in the Management
of Pain During Labour and Postoperative and Intractable Pain.
CNS Drugs 1996, 6, 474-497. (b) Duke, P.; Maze, M.; Morrison,
P. Dexmedetomidine: A General Overview. Int. Congr. Symp.
Ser. - R. Soc. Med. 1998, 221, 11-22.
(4) Boyd, R. E.; Press, J . B.; Rasmussen, C. R.; Raffa, R. B.; Codd,
E. E.; Connelly, C. D.; Bennett, D. J .; Kirifides, A. L.; Gardocki,
J . F.; Reynolds, B.; Hortenstein, J . T.; Reitz, A. B. R2 Adreno-
ceptor Agonists as Potential Analgesic Agents. 1. (Imidazom-
ethyl)oxazoles and thiazoles. J . Med. Chem. 1999, 42, 5064-
5071.
(15) Collier, H. O. J .; Dinneen, L. C.; J ohnson, C. A.; Schneider, C.
The Abdominal Irritant Response and its Suppression by
Analgesic Drugs in the Mouse. Br. J . Pharmacol. 1968, 32, 295-
310.
(16) VonVoigtlander, P. F.; Lewis, R. A. Air-Induced Writhing:
A
Rapid Broad Spectrum Assay for Analgesics. Drug Dev. Res.
1982, 577-581.
(17) (a) Eddy, N. B.; Leimbach, D. Synthetic Analgesics. II. Dithi-
enylbutenyl- and Dithienylbutylamines. J . Pharmacol. Exp.
Ther. 1953, 107, 385-395. (b) O′Callaghan, J . P.; Holtzman, S.
G. Quantification of the Analgesic Activity of Narcotic Antago-
nists by A Modified Hot-Plate Procedure. J . Pharmacol. Exp.
Ther. 1975, 192, 497-505.
(18) D′Amour, R. E.; Smith, D. L. A Method for Determining Loss of
Pain Sensation. J . Pharmacol. Exp. Ther. 1941, 72, 74-79.
(5) (a) Boyd, R. E.; Rasmussen, C. R.; Press, J . B. 4-[(Thien-3-yl)-
methyl]imidazole Analgesics. U.S. Patent 5,750,720, 1998. (b)
Boyd, R. E.; Rasmussen, C. R.; Press, J . B. 4-[(Thien-2-yl)methyl]-
imidazole Analgesics. U.S. Patent 5,621,113, 1997.
J M990569E