Identification of Opioid Ligands Possessing Mixed μ Agonist/δ Antagonist Activity among Pyridomorphinans Derived from Naloxone, Oxymorphone, and Hydropmorphone

Article abstract of DOI:10.1021/jm030311v

A series of pyridomorphinans derived from naloxone, oxymorphone, and hydromorphone (7a-k) were synthesized and evaluated for binding affinity at the opioid δ/μ and κ receptors in brain membranes using radioligand binding assays and for functional activity in vitro using [ ³⁵S]GTP-γ-S binding assays in brain tissues and bioassays using guinea pig ileum (GPI) and mouse vas deferens (MVD) smooth muscle preparations. The pyridine ring unsubstituted pyridomorphinans possessing the oxymorphone and hydromorphone framework displayed nearly equal binding affinity at the μ and δ receptors. Their affinities at the κ site were nearly 10-fold less than their binding affinities at the μ and δ sites. Introduction of aryl substituents at the 5′-position on the pyridine ring improved the binding affinity at the δ site while decreasing the binding affinity at the μ site. Nearly all of the ligands possessing an N-methyl group at the 17-position with or without a hydroxyl group at the 14-position of the morphinan moiety displayed agonist activity at the μ receptor with varying potencies and efficacies. In the [ ³⁵S]GTP-γ-S binding assays, most of these pyridomorphinans were devoid of any significant agonist activity at the δ and κ receptors but displayed moderate to potent antagonist activity at the δ receptors. In antinociceptive evaluations using the warm-water tail-withdrawal assay in mice, the pyridomorphinans produced analgesic effects with varying potencies and efficacies when administered by the intracerebroventricular route. Among the ligands studied, the hydromorphone-derived 4-chlorophenylpyridomorphinan 7h was identified as a ligand possessing a promising profile of mixed μ agonist/δ antagonist activity in vitro and in vivo. In a repeated administration paradigm in which the standard μ agonist morphine produces significant tolerance, repeated administration of the μ agonist/δ antagonist ligand 7h produced no tolerance. These results indicate that appropriate molecular manipulations of the morphinan templates could provide ligands with mixed μ agonist/δ antagonist profiles and such ligands may have the potential of emerging as novel analgesic drugs devoid of tolerance, dependence, and related side effects.

Full text of DOI:10.1021/jm030311v

1400
J . Med. Chem. 2004, 47, 1400-1412
Id en tifica tion of Op ioid Liga n d s P ossessin g Mixed µ Agon ist/δ An ta gon ist
Activity a m on g P yr id om or p h in a n s Der ived fr om Na loxon e, Oxym or p h on e, a n d
Hyd r op m or p h on e
Subramaniam Ananthan,*,† Naveen K. Khare,† Surendra K. Saini,† Lainne E. Seitz,† J effrey L. Bartlett,#
Peg Davis,‡ Christina M. Dersch,§ Frank Porreca,‡ Richard B. Rothman,§ and Edward J . Bilsky#
Organic Chemistry Department, Southern Research Institute, Birmingham, Alabama 35255,
Clinical Psychopharmacology Section, IRP, National Institute on Drug Abuse, Baltimore, Maryland 21224,
Department of Pharmacology, College of Medicine, The University of Arizona Health Sciences Center, Tucson, Arizona 85724,
and Department of Pharmacology, University of New England College of Osteopathic Medicine, Biddeford, Maine 04005
Received J une 24, 2003
A series of pyridomorphinans derived from naloxone, oxymorphone, and hydromorphone (7a -
k ) were synthesized and evaluated for binding affinity at the opioid δ, µ, and κ receptors in
brain membranes using radioligand binding assays and for functional activity in vitro using
[35S]GTP-γ-S binding assays in brain tissues and bioassays using guinea pig ileum (GPI) and
mouse vas deferens (MVD) smooth muscle preparations. The pyridine ring unsubstituted
pyridomorphinans possessing the oxymorphone and hydromorphone framework displayed
nearly equal binding affinity at the µ and δ receptors. Their affinities at the κ site were nearly
10-fold less than their binding affinities at the µ and δ sites. Introduction of aryl substituents
at the 5′-position on the pyridine ring improved the binding affinity at the δ site while decreasing
the binding affinity at the µ site. Nearly all of the ligands possessing an N-methyl group at
the17-position with or without a hydroxyl group at the 14-position of the morphinan moiety
displayed agonist activity at the µ receptor with varying potencies and efficacies. In the
[35S]GTP-γ-S binding assays, most of these pyridomorphinans were devoid of any significant
agonist activity at the δ and κ receptors but displayed moderate to potent antagonist activity
at the δ receptors. In antinociceptive evaluations using the warm-water tail-withdrawal assay
in mice, the pyridomorphinans produced analgesic effects with varying potencies and efficacies
when administered by the intracerebroventricular route. Among the ligands studied, the
hydromorphone-derived 4-chlorophenylpyridomorphinan 7h was identified as a ligand pos-
sessing a promising profile of mixed µ agonist/δ antagonist activity in vitro and in vivo. In a
repeated administration paradigm in which the standard µ agonist morphine produces
significant tolerance, repeated administration of the µ agonist/δ antagonist ligand 7h produced
no tolerance. These results indicate that appropriate molecular manipulations of the morphinan
templates could provide ligands with mixed µ agonist/δ antagonist profiles and such ligands
may have the potential of emerging as novel analgesic drugs devoid of tolerance, dependence,
and related side effects.
In tr od u ction
central nervous system, and each has a role in the
mediation of pain. Morphine and related opioids cur-
rently prescribed as potent analgesics for the treatment
of pain produce their analgesic activity primarily through
their agonist action at the µ opioid receptors. The
general administration of these medications is limited
by significant side effects such as respiratory depression,
muscle rigidity, emesis, constipation, tolerance, and
physical dependence.5,6
Chronic pain represents a major health and economic
problem throughout the world. Despite significant ad-
vances in understanding the physiological and patho-
logical basis of pain, an ideal analgesic is yet to be
discovered. Among analgesic drugs, the opioid class of
compounds still remains the effective treatment agents
for severe and chronic pain.1,2 The opioid drugs produce
their biological effects through their interaction with the
opioid receptors, which belong to the family of seven
transmembrane G-protein-coupled receptors. The exist-
ence of three opioid receptor types µ, δ, and κ has been
clearly established and is confirmed by cloning of these
three receptors from mouse, rat, and human cDNAs.3,4
All three opioid receptor types are located in the human
A large body of evidence indicates the existence of
physical or functional interactions between µ and δ
receptors. Ligands with agonist or antagonist action at
the δ receptor, for example, have been shown to modu-
late the analgesic and adverse effects of µ agonists.7-11
Whereas agonist action at the δ receptors potentiate
µ-mediated analgesic effects, antagonist action at the δ
receptor suppresses the tolerance, physical dependence,
and related side effects of µ agonists without affecting
their analgesic activity. In a study using the non-peptide
ligand naltrindole (1, Chart 1), Abdelhamid and co-
* To whom correspondence should be addressed. Phone: (205) 581-
2822. Fax: (205) 581-2726. E-mail: ananthan@sri.org.
† Southern Research Institute.
#
University of New England College of Osteopathic Medicine.
‡ The University of Arizona Health Sciences Center.
§ National Institute on Drug Abuse.
10.1021/jm030311v CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/10/2004

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1401
Ch a r t 1
agonist/δ antagonist profile of activity in the smooth
muscle assays in vitro.19 In analgesic activity evalua-
tions, this compound displayed partial agonist activity
in the warm-water tail-withdrawal assay and a full
agonist activity in the acetic acid writhing assay after
icv or ip administration in mice, and it did not produce
tolerance to antinociceptive effects on repeated ip injec-
tions. Studies in mice with selective antagonists char-
acterized this compound as a partial µ agonist/δ an-
tagonist.21 Paradoxically, however, in the in vitro bio-
chemical assays using [35S]GTP-γ-S binding, compound
2c failed to display µ agonist activity in guinea pig
caudate membranes as well as in cloned cells expressing
human µ receptors.22 In view of the weak µ agonist
activity and the confounding in vitro/in vivo profile of
this compound, we decided to explore newer analogues
to identify compounds possessing a µ agonist/δ antago-
nist profile of activity in vitro and in vivo.
Toward this goal, we initially explored a series of
analogues of 2c in which the p-chlorophenyl substituent
on the pyridine ring was replaced with a number of
substituted aryl and heteroaryl groups and found that
none of these displayed any significant µ agonist activity
in the [35S]GTP-γ-S binding assay in guinea pig caudate
membranes even at the high concentration of 10 µM.23
These results indicated that naltrexone-derived pyrido-
morphinans possessing the cyclopropylmethyl (CPM)
group at the N-17 position of the morphinan unit, in
general, may bind to and stabilize the receptors in the
antagonist state. It is known that the nature of the
nitrogen substituent on the basic nitrogen plays an
important role in the agonist or antagonist functional
activity among morphinan and 6,7-benzomorphan ligands
possessing selectivity toward µ receptors. For example,
the 4,5-epoxymorphinan ketones naltrexone (3) and
naloxone (4) possessing the N-cyclopropylmethyl and
N-allyl groups interact as antagonists whereas oxymor-
phone (5) and hydromrophone (6) possessing the N-
methyl group display agonist activity at the µ recep-
tors.24-26 In contrast, among epoxymorphinan ligands
possessing selectivity toward δ receptors, the influence
of the N-substituent on the agonist or antagonist
functional activity remains unpredictable. For example,
the replacement of the N-CPM group in naltrindole (1)
with a methyl group yielded a ligand (oxymorphindole)
possessing only partial agonist activity at δ receptors.
Replacement with other alkyl groups, such as ethyl and
higher alkyl groups, produced ligands that displayed
only an antagonist profile of intrinsic activity.26-28 In
addition to the N-substituent, the presence or absence
of a hydroxyl group at the 14-position also influences
the agonist or antagonist potency of 4,5-epoxymorphi-
nan ligands.29 In view of these results, it was of interest
to synthesize and evaluate N-17 and C-14 modified
analogues of 2a -c to study the effect of such modifica-
tions on the binding affinity and functional activity of
the resulting compounds at the opioid µ, δ, and κ
receptors. Herein, we describe the results of this inves-
tigation that led to the identification of ligands with
varying intrinsic activity profiles including those that
displayed the desired mixed µ agonist/δ antagonist
profile of activity among the group of ligands generically
represented by 7.
workers demonstrated that the δ receptor antagonist
greatly reduced the development of morphine tolerance
and dependence in mice in both the acute and chronic
models without affecting the analgesic actions of mor-
phine.12 Fundytus and co-workers reported that con-
tinuous infusion of the δ selective antagonist TIPP[Ψ]
by the intracerbroventricular (icv) route in parallel with
continuous administration of morphine by the sub-
cutaneous route to rats attenuated the development of
morphine tolerance and dependence to a large extent.13
Schiller and co-workers found that the peptide ligand
DIPP-NH2[Ψ] displayed mixed µ agonist/δ antagonist
properties in vitro and that the compound given icv
produced an analgesic effect with no physical depend-
ence and less tolerance than morphine in rats.14,15
Studies with antisense oligonucleotides of the δ receptor
have demonstrated that reduction of δ receptor expres-
sion diminishes the development and/or expression of
morphine dependence without compromising antino-
ciception produced by µ agonists.16,17 Furthermore,
genetic deletion studies using δ receptor knockout mice
have shown that these mutant mice retain supraspinal
analgesia and do not develop analgesic tolerance to
morphine.18 These observations suggest that the devel-
opment of opioid ligands, especially non-peptide ligands
possessing mixed µ agonist/δ antagonist activity, may
provide a novel approach for the development of anal-
gesic agents with low propensity to produce tolerance,
physical dependence, and other side effects.
In our studies on naltrexone-derived heterocycle an-
nulated morphinan ligands, we found that the pyrido-
morphinan 2a displayed high-affinity binding at the
opioid receptors and that the binding affinity and
antagonist potency of the pyridomorphinans at the δ
receptors are modulated by the substituents placed at
the 5′-position on the pyridine moiety. For example, the
introduction of aromatic groups such as a phenyl group
(2b) or a 1-pyrrolyl group at this position gave ligands
with high binding affinity and improved δ antagonist
potency as determined in bioassays using mouse vas
deferens smooth muscle preparations.19,20 Interestingly,
among phenyl ring substituted analogues of 2b, the
p-chlorophenyl compound (2c) displayed a mixed µ

1402 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
Sch em e 1
a
Reagents and reaction conditions: (i) AcONH4, AcOH, reflux, 18 h; (ii) BBr3, CH2Cl2, -20 °C, 4 h.
Sch em e 2
a
Reagents and reaction conditions: (i) AcONH4, AcOH, reflux, 18 h; (ii) 10, 11, or 12, AcONH4, AcOH, reflux, 18 h; (iii) BBr3, CH2Cl2,
-20 °C, 4 h.
Ch em istr y
dialdehyde 9 gave the pyridomorphinan 16, which was
then converted to the N-nor compound 17 by reaction
with vinyl chloroformate followed by hydrolysis of the
resulting carbamate intermediate. Alkylation of 17 with
cyclopropylmethyl bromide followed by removal of the
methyl group from the ether function gave the desired
target compound 7l.
The synthesis of the pyridomorphinan target com-
pounds was accomplished from commercially available
morphinan ketones using the pyridine annulation meth-
odolgy that we had earlier developed for the prototype
compounds.19,20,23 As depicted in Scheme 1, the conden-
sation of naloxone (4) or hydromorphone (6) with
4-chlorophenylmalondialdehyde (9) in the presence of
ammonium acetate in acetic acid gave the corresponding
pyridine compounds 7a and 7h , respectively. Since
oxymorphone (5) was not commercially available, we
utilized oxycodone (8) as the starting material for the
preparation of the target compounds possessing oxy-
morphone framework. Thus, the condensation of 8 with
the aldehyde 9 under the standard reaction conditions
gave the methyl ether 7o, which was then converted to
7d by phenolic O-demethylation using BBr3. The target
compounds 7f, 7g, and 7i-k were obtained by reacting
hydromorphone (6) with the enaminoaldehydes 10-14
and ammonium acetate (Scheme 2). Oxycodone (8) was
reacted with the aldehydes 10-12 to obtain the corre-
sponding methyl ethers 7m , 7n , and 7p , which were
then demethylated with BBr3 to yield the target com-
pounds 7b, 7c, and 7e, respectively. The 14-deoxy
analogue of 2c was synthesized by the sequence of
reactions shown in Scheme 3. The pyridine ring annu-
lation reaction of hydrocodone (15) with the malon-
Biology
Op ioid Recep tor Bin d in g. The binding affinities
of the target compounds for the opioid δ and µ recep-
tors were determined by inhibition of binding of
[3H]DADLE30 and [3H]DAMGO31 to rat brain mem-
branes. [3H]DADLE binding to µ receptors was blocked
using 100 nM DAMGO. The affinities of the compounds
for the κ receptors were determined by inhibition of
binding of [3H]U69,59332 to guinea pig brain membranes
using previously reported procedures.19,33 The δ, µ, and
κ opioid receptor binding affinities along with binding
selectivity ratios for the target compounds 7a -l are
given in Table 1. The phenolic methyl ether compounds
7m -q were prepared as intermediates leading to the
corresponding phenolic targets. These methyl ethers
were also evaluated for their binding affinities. The
affinity data for these ethers as well as the previously
reported data for the prototype compounds 2a -c are
also listed in Table 1.

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1403
Sch em e 3
a
Reagents and reaction conditions: (i) 9, AcONH4, AcOH, reflux, 18 h; (ii) vinyl chloroformate, K2CO3, 1,2-dichloroethane, reflux, 36
h; (iii) 2 N HCl, EtOH, reflux, 2 h; (iv) cyclopropylmethyl bromide, NaHCO3, EtOH, reflux, 14 h; (v) BBr3, CH2Cl2, -20 °C, 4 h.
Ta ble 1. Binding Affinities of the Pyridomorphinans at the Opioid δ, µ, and κ Receptors in Rodent Brain Membranes
Ki ( SEM (nM)
selectivity ratio
compd
R
X
R′
R′′
δa
µb
κc
µ/δ
κ/δ
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
7k
7l
7m
7n
allyl
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
CPM
Me
Me
Me
Me
CPM
CPM
CPM
CPM
OH
OH
OH
OH
OH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
H
4-chlorophenyl
H
phenyl
4-chlorophenyl
4-bromophenyl
H
8.2 ( 0.1
18 ( 1.4
2.9 ( 0.1
3.9 ( 0.2
4.0 ( 0.3
8.0 ( 0.8
1.9 ( 0.1
4.4 ( 0.2
5.0 ( 0.6
3.7 ( 0.1
1.1 ( 0.1
2.6 ( 0.1
143 ( 9
467 ( 19
7.9 ( 0.2
26 ( 1.0
230 ( 10
196 ( 4
75 ( 5
57
0.4
9
59
49
1.6
13
34
40
25
88
24
2.3
26
98
82
48
1.9
16
23
9
15
124
120
108
8
43
18
18
75
366
2.3
45
>294
>338
320
13
11
20
9
47
264 ( 18
360 ( 17
468 ( 17
432 ( 18
66 ( 2
13 ( 0.5
24 ( 2
phenyl
81 ( 5
4-chlorophenyl
4-bromophenyl
3,4-dichlorophenyl
2,4-dichlorophenyl
4-chlorophenyl
H
148 ( 9.5
200 ( 11
93 ( 4
97 ( 4
62 ( 3
78 ( 13
91 ( 6
278 ( 7
H
H
403 ( 9
6.0 ( 0.3
6400 ( 350
>10000
OH
OH
OH
OH
H
OH
OH
OH
325 ( 16
894 ( 19
2050 ( 95
1890 ( 70
1970 ( 50
1.5 ( 0.1
13.5 ( 1.0
51.0 ( 8.0
151 ( 14
8070 ( 930
6.1 ( 0.7
11.0 ( 1.0
7250 ( 660
2.4 ( 0.3
phenyl
34 ( 0.6
21 ( 2
7o
7p
4-chlorophenyl
4-bromophenyl
4-chlorophenyl
H
phenyl
4-chlorophenyl
>7100
23 ( 1.3
41 ( 3
7370 ( 520
539 ( 20
8.8 ( 0.7
17.6 ( 1.6
20.0 ( 1.0
75 ( 7
7q
2a d
0.78 ( 0.06
0.87 ( 0.07
2.2 ( 0.2
1.6 ( 0.1
5.6 ( 0.5
469 ( 39
157 ( 11
>5000
2bd
H
H
2cd
1, NTIe
SNC-80e
DAMGOe
morphinee
U69,593e
naltrexonee
94
8760 ( 710
5820 ( 540
188 ( 20
4.8 ( 0.5
2.2 ( 0.2
1440
0.01
0.07
<1.5
0.3
1560
12
1.2
<0.01
0.3
7.5 ( 1.8
a
b
Displacement of [3H]DADLE (1.3-2.0 nM) in rat brain membranes using 100 nM DAMGO to block binding to µ sites. Displacement
of [3H]DAMGO (1.4-3.0 nM) in rat brain membranes. c Displacement of [3H]U69,593 (1.2-2.2 nM) in guinea pig brain membranes. Data
d
from ref 19. e Data included for comparison.
[
35S]GTP -γ-S Bin d in g Assa ys. All compounds were
selective antagonist ligands used were CTAP (2 µM) to
block µ receptors, TIPP (1 µM) to block δ receptors, and
nor-BNI (6 nM) to block κ receptors.34 The antagonist
properties of the compounds were determined by meas-
uring the test compound’s ability to inhibit stimulation
of [35S]GTP-γ-S binding produced by the selective ago-
nists (10 µM): SNC-80 for δ receptor, DAMGO for µ
receptor, and U69,593 for κ receptor.35,36 Compounds
were selected for more detailed study, using concentra-
screened at a 10 µM concentration for agonist and
antagonist activity at µ, δ, and κ receptors in vitro using
[35S]GTP-γ-S binding assays in guinea pig caudate
membranes as described previously.34-36 Agonist activ-
ity was tested by measuring the stimulation of [35S]GTP-
γ-S binding by the compounds in the absence and
presence of fixed concentrations of selective antagonists
to block receptors other than the one being studied. The

1404 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
Ta ble 2. Antagonist and Agonist Functional Activity of Selected Compounds in [35S]GTP-γ-S Binding Assays in Guinea Pig Caudate
Membranes
antagonist activity apparent Ki ( SD (nM)
agonist activity EC50 ( SD (nM), Emax (%)
compd
7b
δa
µb
κc
δd
µe
κf
>1000
>1000
>1000
g
g
3000 ( 670
40 ( 2%
660 ( 190
27 ( 2%
g
g
g
7c
920 ( 240h
>1000
>1000
7d
7f
17 ( 2.6
595 ( 48
>1000
>1000
767 ( 54
>1000
g
g
g
g
500 ( 80
44 ( 1%
1030 ( 100
60 ( 1%
900 ( 170
48 ( 2%
3220 ( 570
59 ( 3%
1310 ( 240
63 ( 3%
225 ( 30
51 ( 5%
g
7g
7h
7i
>1000
>1000
i
>1000
g
g
g
g
g
g
g
g
g
g
10.9 ( 1.0
55 ( 16h
74 ( 14h
1.1 ( 0.1
333 ( 25
265 ( 29
4770 ( 1500h
308 ( 21
>1000
>1000
>1000
7j
7k
7l
2c
1, NTI
SNC-80
1.56 ( 0.14
0.18 ( 0.01
0.062 ( 0.006
naj
9.2 ( 0.86
7.8 ( 0.42
3.2 ( 0.2
naj
11.2 ( 0.6
11.2 ( 0.4
8.8 ( 0.8
naj
g
g
g
g
g
g
g
g
760 ( 130
100%
naj
naj
DAMGO
morphine
U69,593
naj
naj
naj
naj
naj
naj
naj
naj
naj
naj
414 ( 50
100%
290 ( 80
32 ( 2%
naj
naj
g
g
naj
380 ( 40
100%
a
b
SNC-80 (10 µM) was used as the agonist selective for the δ receptor. DAMGO (10 µM) was used as the agonist selective for the µ
receptor. c U69,593 (10 µM) was used as the agonist selective for the κ receptor. The µ and κ sites were blocked with the antagonists
d
CTAP (2 µM) and nor-BNI (6 nM). e The δ and κ sites were blocked with the antagonists TIPP (1 µM) and nor-BNI (6 nM). f The δ and
g
h
µ sites were blocked with TIPP (1 µM) and CTAP (2 µM). Not active as an agonist. IC50 values. Ki values could not be calculated
because of partial inhibition of agonist stimulated [35S]GTP-γ-S binding. Ki value could not be calculated because of agonist activity.j na
i
) not applicable.
Ta ble 3. Antagonist and Agonist Functional Activity of
tion-response curves, based on their binding Ki values
Selected Compounds in Mouse Vas Deferens (MVD) and Guinea
(Table 1) and their profile of agonist and antagonist
Pig Ileum (GPI) Smooth Muscle Assays
activity in the initial [35S]GTP-γ-S binding assay. The
agonist activity
agonist efficacy of the compounds is expressed as a
MVD (δ)
IC50 (nM)
or % max
responsec
GPI (µ)
IC50 (nM)
or % max
responsec
antagonist activity
percent of stimulation compared to that produced by the
standard agonist. The results are presented in Table 2.
Bioa ssa ys in Sm ooth Mu scle P r ep a r a tion s. The
functional activity profiles of selected ligands were also
determined in the mouse vas deferens (MVD) and
guinea pig ileum (GPI) smooth muscle preparations as
described previously.37,38 The agonist activity was de-
termined by the ability of the compound to inhibit
electrically stimulated contractions of the GPI and
MVD. The GPI is primarily a µ receptor preparation,
even though the ileum does also contain κ receptors. In
the MVD, the opioid effects are predominantly mediated
through δ receptors, but µ and κ receptors also exist in
this tissue. Testing for antagonist activity was carried
out by preincubating the muscle preparations with the
test compound 30 min prior to washing with buffer and
testing with the standard δ agonist DPDPE in the MVD
and the µ agonist PL-017 in the GPI. The antagonist
and agonist potencies of the tested compounds are listed
in Table 3.
MVD (δ)
Ke (nM)a
GPI (µ)
compd
7b
7c
7d
7f
7g
7h
7i
Ke (nM)b
d
d
d
d
d
d
d
d
d
d
d
d
49%
213 ( 56
58%
152 ( 26
48.4 ( 8.7
565 ( 13
44%
489 ( 164
28.3%
523 ( 95
211 ( 36
8.7%
38.2 ( 14.7
d
d
67.7 ( 10
98.7 ( 20.1
177 ( 41
666 ( 127
447 ( 163
724 ( 131
109 ( 27
0%
21.9 ( 2.14
20.0 ( 6.6
d
5.02 (1.56
6.00 ( 1.33
37.0 ( 1.0
3.7 ( 1.0
0.91 ( 0.48
7j
7k
7l
7.9%
2a e
2be
2ce
190 ( 65 0%
43 ( 6.6
d
4.7%
21%
16%
0%
163 ( 22
18%
1, NTIf
DPDPEf
PL-017f
0.53 ( 0.18 43 ( 3.0
d
d
d
d
5.81 ( 1.63 11600 ( 2990
243 ( 38 59.4 ( 3.6
a
Determined using DPDPE as the agonist ligand for the δ
b
receptor. Determined using PL-017 as the agonist ligand for the
µ receptor. c Partial agonist activity is expressed as the percentage
d
inhibition of contraction at 1 µM. The agonist effects precluded
the determination of antagonist effects. e Data from ref 19. f Data
included for comparison.
An a lgesic Testin g a n d Assessm en t of Toler a n ce
Developm en t. The analgesic activity of selected ligands
was tested in mice using the 55 °C warm-water tail-
withdrawal test as previously described.21 The test
compounds were administered by the intracerbro-
ventricular (icv) route. The analgesic effects of the
compounds that were evaluated are given in Table 4.
The A50 values were calculated for compounds that
produced full antinociceptive effects with minimal or no
toxicity. For those compounds for which the A50 values
could not be calculated, the percentage antinociception

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1405
Ta ble 4. Analgesic Activity of Selected Ligands in the Mouse
ing site of the δ receptor as opposed to unfavorable
interactions at the µ and κ receptors. Introduction of a
chlorine or bromine substituent at the para position of
the free-rotating phenyl ring in 7c or 7g brings about a
modest decrease in binding affinity at the δ site. Of the
two isomeric dichlorophenyl compounds 7j and 7k , the
2,4-dichlorophenyl compound 7k binds with higher
affinity at the δ site than the 3,4-dichlorophenyl com-
pound. This structure-affinity trend parallels that
observed among dichlorophenyl analogues of 2b.23 Among
phenolic compounds 7a -l, the 2,4-dichlorphenyl com-
pound 7k displayed the highest δ receptor binding
affinity (Ki ) 1.1 nM) and the highest δ receptor binding
selectivity with µ/δ and κ/δ selectivity ratios of 88 and
366, respectively.
Warm-Water Tail-Withdrawal Assaya
A50 or
% nociception
95% confidence
limits (nmol)
naloxone
compd
7b
7c
7d
7f
7g
7h
7i
7j
7k
7l
2c
sensitivityb
67.1 nmol
49.0-91.8
yes
no
partial
no
yes
64% @ 100 nmol
43% @ 300 nmol
75% @100 nmol
48% @ 100 nmol
42.8 nmol
c
c
c
c
30.6-59.8
31.3-71.2
c
c
c
c
yes
yes
47.2 nmol
44% @ 300 nmol
40% @ 60 nmole
18% @ 300 nmol
21% @ 100 nmolf
4.2 nmol
NDd
NDd
NDd
yes
morphine
3.0-6.9
yes
a
Compounds were administered icv with A50 values calculated
b
at time of peak drug effect. Compounds exhibiting greater than
A comparison of the affinities of compounds possess-
ing an N-CPM group (2a -c) with those possessing an
N-methyl group (7b-d ) indicates that replacing the
CPM group with a methyl group in general led to
reduction in affinities at all three receptors. The reduc-
tions in affinities at the κ sites are relatively larger than
the reductions in binding affinities at δ or µ sites.
Compared to the N-CPM compound 2c, the N-allyl
analogue 7a also displayed reduced affinities at the δ,
µ, and κ receptors with greater reduction in affinity at
the µ site (9-fold) than at δ (4-fold) or κ sites (4-fold).
Comparison of the affinities of 7b-e with 7f-i and
those of 2c with 7l indicates that the replacement of
the 14-hydroxyl group with a hydrogen atom brings
about only a modest change in the affinity at the δ and
µ sites (less than 3-fold change in affinity). At the κ site,
however, the deoxy compounds displayed 3- to 6-fold
higher affinity than their 14-hydroxy counterparts. The
presence of a free phenolic hydroxyl group is usually
considered essential for high-affinity binding at opioid
receptors. It was not surprising, therefore, to find that
the affinities of the phenolic methyl ethers 7m -q were
lower than the affinities of their corresponding phenolic
compounds 7b-e and 7l at all three binding sites. The
magnitude of reduction in affinity at the δ site, in
general, was less (5- to 16-fold) than the reduction in
affinity at the µ (9- to 41-fold) or κ sites (15- to 90-fold).
One of the primary goals of the present investigation
was to identify ligands with a mixed profile of agonist
activity at the µ receptor and antagonist activity at the
δ receptor. As shown by the functional activity data in
the [35S]GTP-γ-S assays (Table 2), most of the com-
pounds examined in the present study in general
displayed the desired profile of µ agonist/δ antagonist
activity. Compounds 7c, 7i, and 7j were devoid of
agonist activity at the δ receptors but failed to inhibit
SNC-80 stimulated binding of [35S]GTP-γ-S to 100% in
the antagonist assays at the δ receptors. The maximum
percentage inhibition displayed by 7c, 7i, and 7j were
61 ( 3%, 69 ( 3%, and 58 ( 3%, respectively. A similar
partial inhibition profile (maximum inhibition of 66 (
6%) was also observed for 7j at the κ receptors. The
partial inhibition profile displayed by these ligands is
exemplified by the concentration-response curve for 7j
shown in Figure 1. For these compounds, the calculated
IC50 values instead of Ki values are listed in Table 2.
The intriguing observation of partial antagonist inhibi-
tion is currently under investigation and will be the
subject of a separate report.
80% reduction in the antinociceptive effect to a fixed dose of
naloxone are designated “yes” for naloxone sensitivity. c 95%
d
confidence levels could not be calculated. ND ) not determined.
e Doses of 100-600 nmol produced less than 40% MPE. f Doses of
300 and 600 nmol produced less than 10% MPE.
at the given dose is listed in the table. To determine
whether the analgesic activity of the tested compounds
is mediated through opioid receptors, the blockade of
antinociceptive activity by pretreatment with naloxone
was carried out. The analgesic activity was considered
as naloxone-sensitive if greater than 80% reduction in
the antinociceptive response was observed. Selected
compounds were also tested for antinociception in mice
pretreated with the µ selective antagonist â-FNA (19
nmol, icv, -24 h). The development of tolerance to
analgesic effect was evaluated using a repeated admin-
istration procedure as previously described.21
Resu lts a n d Discu ssion
An examination of the affinities of the phenolic target
compounds 7a -l reveals that, with the exception of 7b,
all of the ligands display high-affinity binding at the δ
site with Ki values less than 10 nM and are δ selective,
their binding affinities at δ site being higher than their
affinities at the µ and κ sites. Compounds 7b and 7f
possess the basic morphinan unit present in oxymor-
phone and hydromorphone, respectively, and do not
carry any substituent on the fused pyridine ring system.
These two compounds displayed a relatively nonselec-
tive binding profile between µ and δ receptors [Ki(µ)/
Ki(δ) ) 0.4 for 7b; Ki(µ)/Ki(δ) ) 1.6 for 7f]. Their
affinities at the κ site were significantly lower than their
affinities at δ and µ sites. The introduction of a phenyl
group at the 5′-position on these two templates gave
compounds 7c and 7g, which displayed 4- to 6-fold
enhanced affinity at the δ site in comparison to the
parent compounds. This improvement in the binding
affinity of the phenyl-substituted analogues at the δ site
is accompanied by a decrease in affinity at µ and κ sites,
thus leading to an enhancement in δ selectivity profile
of these compounds. These improvements in the binding
affinity and binding selectivity brought about by the
introduction of an aryl group at the 5′-position of the
pyridomorphinan is in conformity with the structure-
affinity relationships observed among closely related 5′-
aryl analogues of 2a .19,20,23 Thus, it appears that an aryl
group placed at the 5′-position of the pyridomorphinan
templates encounters favorable interactions at the bind-

1406 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
N-CPM group with a methyl group brings about a
significant reduction in the δ antagonist potency but
without altering the innate antagonist profile of these
ligands at the δ receptor. Interestingly, while the
introduction of a second chlorine atom at the meta
position of the chlorophenyl ring of 7h (compound 7j)
did not significantly change the binding affinity or the
antagonist potency at the δ receptor, the introduction
of the chlorine atom at the ortho position (compound
7k ) provided a 4-fold enhancement in binding affinity
and a 10-fold improvement in antagonist potency at the
δ receptor. Among the ligands studied, the 2,4-dichloro-
phenyl compound 7k is not only the most potent δ
antagonist but also the most potent µ agonist, thus
making it the best mixed µ agonist/δ antagonist ligand
in vitro. The profile of 7h is similar to that of 7k but
with somewhat weaker antagonist and agonist potencies
at the δ and µ receptors, respectively, in vitro. As
observed in an earlier work with κ receptor antago-
nists,35 some of the δ antagonist ligands (7b, 7c, 7f, 7g)
displayed a marked discrepancy between the binding
Ki values (Table 1) and the corresponding functional Ki
values (Table 2). The reason for this is unclear.
F igu r e 1. Concentration-response curve for inhibition of
SNC-80 stimulated [35S]GTP-γ-S binding by 7j.
The functional activity results obtained for the se-
lected compounds in the smooth muscle assays (Table
3) were somewhat similar to that obtained in the
[35S]GTP-γ-S assays. All of the ligands that displayed
agonist activity in the [35S]GTP-γ-S assays at the µ site
also displayed agonist activity in the GPI. One signifi-
cant exception is the activity of 7l, which was a potent
agonist in the GPI (IC50 ) 109 nM) but was found to be
an antagonist at the µ site in the [35S]GTP-γ-S assays.
This intriguing activity profile of 7l is similar to that
displayed by its 14-hydroxy analogue 2c studied ear-
lier.19,22 While none of the compounds displayed any
significant agonist activity at the δ site in the [35S]GTP-
γ-S assays, a few compounds, 7c, 7f, 7g, 7h , and 7j,
displayed agonist activity in the MVD smooth muscle
preparations. It appears that the agonist activity dis-
played by these compounds in the MVD may be due to
their agonist effects at the µ receptors. Although not
investigated in depth, we found that the agonist activity
of 7f and 7h in the MVD was blocked by the nonselective
antagonist naloxone but not by the δ selective antago-
nist ICI-174,864. Among the compounds studied, δ
antagonist Ke values could be determined for 7d , 7h ,
7i, 7k , and 7l in the MVD. The antagonist Ke values
for these compounds were in the range of 5 nM (7k ) to
38 nM (7d ). Compounds 7h and 7k were the two
compounds that displayed an in vitro µ agonist/δ
antagonist profile of activity in both the [35S]GTP-γ-S
and the smooth muscle assay systems.
Among the compounds studied, only two compounds,
7d and 7l, failed to display agonist activity at the µ
receptor. Compound 7l is the 14-deoxy analogue of 2c
and carries the N-CPM group at the 17-position. The
lack of µ agonist activity for 7l is therefore not surpris-
ing because removal of the hydroxyl function from the
14-position by itself may not be sufficient for inducing
µ agonist interactions for the morphinan ligands car-
rying the N-CPM group. With the exception of 7d , all
of the ligands carrying an N-methyl group displayed µ
agonist activity with varying potencies. Among the
ligands that displayed µ agonist activity, compound 7k
was the most potent with an EC50 value of 225 nM,
which is comparable to the EC50 values of morphine (290
nM) and DAMGO (414 nM). The rank order of potencies
for the agonist ligands was 7k > 7f > 7c > 7h > 7g >
7j > 7b > 7i, and there appears to be no strict
correlation between the agonist potency and the binding
affinity of these ligands at the µ receptor. This is not
surprising not only in view of the substantial differences
in assay conditions between the binding and functional
[35S]GTP-γ-S binding assays but also because the EC50
of stimulation of [35S]GTP-γ-S binding is a function of
binding affinity and intrinsic efficacy.39 The agonist
efficacies of these ligands, as indicated by their percent-
age maximum stimulation (Emax) values, were in the
range of 27% (7c) to 63% (7j). With the exception of 7c,
all of these ligands were more efficacious than morphine
(Emax ) 32%) but less efficacious than DAMGO (Emax
)
The structure-activity relationships observed in the
present study, together with those observed in our
earlier investigations, suggest that fusion of an unsub-
stituted pyridine ring at the 6,7-position of 4,5-epoxy-
morphinans has the effect of increasing the binding
affinity at the δ receptor and decreasing the binding
affinity at the µ receptor, thus leading to templates that
bind with nearly equal affinity to the δ and µ receptors.
These pyridomorphinans appear to have a basic ten-
dency to interact with δ receptors as antagonists ir-
respective of the nature of the alkyl substituent (CPM
or methyl) on the morphinan nitrogen. Their functional
100%). With regard to antagonist activity, the N-CPM
compound 7l displayed significant antagonist potency
at all three receptors. All of the N-methyl compounds
examined displayed no or only weak antagonist activity
at the µ and κ receptors. Interestingly, however, most
of the compounds displayed moderate antagonist poten-
cies at the δ receptor. The pyridomorphinans 7d and
7h possessing the 4-chlorophenyl substituent are N-
methyl analogues of the N-CPM compound 2c. These
two compounds 7d and 7h displayed δ antagonist Ki
values of 17 and 10.9 nM, respectively, compared to the
Ki value of 0.18 nM for 2c. Thus, the exchange of the

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1407
F igu r e 3. Antinociceptive dose- and time-response curve
for 7h (icv) with and without pretreatment with â-FNA (19
nmol, icv, -24 h).
F igu r e 2. Antinociceptive dose- and time-response curves
for 7h (icv) in the 55 °C warm-water tail-withdrawal assay.
activity at the µ receptor, however, appears to be
governed by the nature of the N-alkyl substituent, those
with the CPM group interacting as antagonists and
those with the N-methyl group interacting as agonists.
Further modulations in binding and functional activity
of the pyridomorphinans could be achieved through
introduction of appropriate substituents, particularly at
the 5′-position of the pyridine ring.
All of the compounds that were evaluated in the
functional assays in vitro were evaluated for antino-
ciceptive activity in mice using the warm-water tail-
withdrawal assay (Table 4). Among the compounds
tested, the antinociceptive A50 values could be deter-
mined for only three compounds: 7b, 7h , and 7i. Factors
that prevented us from determining the A50 values for
other compounds include lack of efficacy (7c, 7j, 7k , and
7l), lack of potency (7d ), insensitivity to naloxone (7f),
and toxicity (7g). All of the compounds examined in the
present study were found to be more efficacious in the
warm-water tail-withdrawal assay than the previously
studied compound 2c. From the results obtained in the
in vitro functional assays, 7k and 7h were identified as
compounds of interest as mixed µ agonist/δ antagonist
ligands. Of these two compounds, the δ antagonist/µ
agonist profile of 7k was superior to that of 7h in vitro.
In the antinociceptive evaluations, however, compound
7k was found to be not as efficacious as 7h . Compound
7h displayed full agonist efficacy with an A50 potency
value of 42.8 nmol in the warm-water tail-withdrawal
assay in mice (Figure 2). The antinociceptive activity
of this compound was completely blocked by the µ
selective antagonist â-FNA (Figure 3), confirming that
the analgesic activity of this compound is indeed medi-
ated through opioid µ receptors. From these studies, the
pyridomorphinan 7h emerged as a ligand possessing
mixed µ agonist/δ antagonist activity in vitro and in
vivo. It was gratifying to find that this compound when
tested in the tolerance development assays involving
repeated injections of the compound for 3 days induced
an insignificant shift in the antinociceptive potency (less
than 1.1-fold increase in A50 value), indicating very little
development of tolerance. This is in contrast to mor-
F igu r e 4. Antinociceptive dose-response curves for naive
control mice and mice injected repeatedly with A90 doses of
icv morphine or 7h given twice daily for 3 days.
phine, which in the same paradigm, produced a signifi-
cant 6.4-fold shift in the A50 values, indicating the
development of tolerance to its analgesic effects (Figure
4). The lack of tolerance displayed by this non-peptide
µ agonist/δ antagonist ligand 7h is an exciting finding
that supports the hypothesis that ligands with a mixed
µ agonist/δ antagonist profile of activity have the
potential of becoming therapeutically useful analgesic
agents devoid of tolerance and dependence development
commonly associated with pure µ agonist analgesics
such as morphine.
Con clu sion s
The fusion of an unsubstituted pyridine ring on the
oxymorphone and hydromorphone framework gave
pyridomorphinans that bind with nearly equal affinity
at the µ and δ receptors and with much less affinity at
the κ receptors. Introduction of aryl substituents at the
5′-position on these pyridomorphinan scaffolds in gen-

1408 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
C-2′′ H, C-3′′ H, C-5′′ H, C-6′′ H), 7.47 (d, 1H, J ) 2.1 Hz, C-4′
H), 8.69 (d, 1H, J ) 1.8 Hz, C-6′ H); MS m/z 473 (MH)+. Anal.
(C28H25ClN2O3‚0.1H2O) C, H, N.
eral improved the affinity and antagonist potency at the
δ receptor with retention of agonist activity at the µ
receptor, thus leading to mixed µ agonist/δ antagonist
ligands. The in vitro functional activity evaluations led
to the identification of hydromorphone-derived chloro-
phenylpyridomorphinan 7h and dichlorophenylpyrido-
morphinan 7k as ligands with the desired µ agonist/δ
antagonist properties in vitro. Antinociceptive evalua-
tions with 7h by the warm-water tail-withdrawal test
in mice demonstrated that the compound produces
antinociceptive effects without inducing analgesic toler-
ance on repeated administration. This finding is very
significant because it demonstrates the validity of the
hypothesis that opioid ligands with a mixed µ agonist/δ
antagonist profile of activity may have diminished
propensity to induce tolerance and therefore may have
therapeutic advantages over µ agonist analgesics for
long-term treatment of pain. Further studies are clearly
warranted to fully explore the potential of such ligands
as novel analgesic drugs superior to those currently
available.
6,7-Did eh yd r o-3,14-d ih yd r oxy-4,5r-ep oxy-17-m et h yl-
p yr id o[2′,3′:6,7]m or p h in a n (7b). Oxycodone hydrochloride
(2.0 g, 5.69 mmol), 3-(dimethylamino)acrolein (0.845 g, 8.52
mmol), ammonium acetate (1.31 g, 17.04 mmol), and acetic
acid (30 mL) was refluxed in an oil bath at 130-135 °C under
an atmosphere of argon for 18 h. Workup of the reaction
mixture and purification of the crude product as described
above for the preparation of 7a gave 6,7-didehydro-4,5R-epoxy-
14-hydroxy-3-methoxy-17-methylpyrido[2′,3′:6,7]morphinan (7m )
(0.792 g, 40%): mp 210-212 °C; TLC Rf ) 0.4 (CH2Cl2-
MeOH-NH4OH, 94.5:5:0.5); 1H NMR (CDCl3) δ 1.80-1.83 (m,
1H, C-15 H), 2.35-2.40 (m, 2H, C-15 H, C-16 H), 2.43 (s, 3H,
NCH3), 2.50-2.78 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.95 (d,
1H, J ) 6.5 Hz, C-9 H), 3.26 (d, 1H, J ) 18.7 Hz, C-10 H),
3.79 (s, 3H, OCH3), 4.5-5.8 (broad hump, 1H, C-14 OH), 5.53
(s, 1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H), 6.66 (d, 1H, J
) 8.1 Hz, C-1 H), 7.10 (dd, 1H, J ) 7.7 and 4.6 Hz, C-5′ H),
7.34 (d, 1H, J ) 7.7 Hz, C-4′ H), 8.56-8.58 (m, 1H, C-6′ H);
MS m/z 351 (MH)+. Anal. (C21H22N2O3‚0.2H2O) C, H, N.
A solution of 7m (0.67 g, 1.91 mmol) in CH2Cl2 (25 mL) was
cooled to -78 °C and treated dropwise with BBr3 (19.0 mL of
1 M solution in CH2Cl2, 19.0 mmol). After 30 min, the reaction
mixture was allowed to warm to -15 to - 20 °C and stir for 4
h. The mixture was then treated with Et2O (2 mL) and allowed
to warm to room temperature. After being stirred for an
additional 30 min, the mixture was diluted with water and
extracted twice with CH2Cl2. The organic layer was washed
with brine and dried (Na2SO4). The solvents were removed
under reduced pressure and the crude product obtained was
chromatographed over a column of silica using CH2Cl2-
MeOH-NH4OH (97.5:2:0.5) as the eluent to obtain 7b (0.179
g, 28%): mp >230 °C; TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH,
94.5:5:0.5); 1H NMR (DMSO-d6) δ 1.52-1.56 (m, 1H, C-15 H),
2.14-2.32 (m, 2H, C-15 H, C-16 H), 2.35 (s, 3H, NCH3), 2.44-
2.62 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.91 (d, 1H, J ) 6.1 Hz,
C-9 H), 3.16 (d, 1H, J ) 18.6 Hz, C-10 H), 4.75 (s, 1H, C-14
OH), 5.28 (s, 1H, C-5 H), 6.49-6.54 (m, 2H, C-1 H, C-2 H),
7.23 (dd, 1H, J ) 7.7 and 4.7 Hz, C-5′ H), 7.46 (dd, 1H, J )
7.7 and 1.4 Hz, C-4′ H), 8.48 (dd, 1H, J ) 4.7 and 1.5 Hz, C-6′
H), 9.01 (s, 1H, C-3 OH); MS m/z 337 (MH)+. Anal. (C20H20N2O3‚
0.3H2O) C, H, N.
6,7-Did eh yd r o-3,14-d ih yd r oxy-4,5r-ep oxy-17-m et h yl-
5′-p h en ylp yr id o[2′,3′:6,7]m or p h in a n (7c). Oxycodone hy-
drochloride (2.0 g, 5.68 mmol), was reacted with 3-(dimethyl-
amino)-2-phenylacrolein40 (1.40 g, 8.52 mmol) and ammonium
acetate (1.31 g, 17.04 mmol) in acetic acid (20 mL) by the same
procedure as described for the preparation of 7a to obtain
6,7-didehydro-4,5R-epoxy-14-hydroxy-3-methoxy-17-methyl-5′-
phenylpyrido[2′,3′:6,7]morphinan (7n ) (1.325 g, 55%): mp
>250 °C; TLC Rf ) 0.5 (CH2Cl2-MeOH-NH4OH, 96.5:3:0.5);
1H NMR (CDCl3) δ 1.82-1.86 (m, 1H, C-15 H), 2.37-2.42 (m,
2H, C-15 H, C-16 H), 2.44 (s, 3H, NCH3), 2.52-2.84 (m, 4H,
C-8 H2, C-10 H, C-16 H), 2.99 (d, 1H, J ) 6.4 Hz, C-9 H), 3.28
(d, 1H, J ) 18.7 Hz, C-10 H), 3.82 (s, 3H, OCH3), 5.58 (s, 1H,
C-5 H), 6.63 (d, 1H, J ) 8.1 Hz, C-2 H), 6.68 (d, 1H, J ) 8.1
Hz, C-1 H), 7.37-7.53 (m, 6H, C-4′ H, C-5′ phenyl-H), 8.78-
8.79 (m, 1H, C-6′ H); MS m/z 427 (MH)+. Anal. (C27H26N2O3‚
0.2H2O) C, H, N.
The methyl ether 7n (0.189 g, 0.44 mmol) in CH2Cl2 (10 mL)
was reacted with BBr3 (4.4 mL of 1 M solution in CH2Cl2, 4.4
mmol) as described for the preparation of 7b from 7m to yield
0.112 g (61%) of 7c: mp 190-192 °C; TLC Rf ) 0.3 (CH2Cl2-
MeOH-NH4OH, 96.5:3:0.5); 1H NMR (CDCl3) δ 1.82-1.85 (m,
1H, C-15 H), 2.38-2.42 (m, 2H, C-15 H, C-16 H), 2.44 (s, 3H,
NCH3), 2.53-2.82 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.98 (d,
1H, J ) 6.4 Hz, C-9 H), 3.27 (d, 1H, J ) 18.7 Hz, C-10 H),
4.20-5.68 (broad hump, 2H, C-3 OH, C-14 OH), 5.59 (s, 1H,
C-5 H), 6.60 (d, 1H, J ) 8.1 Hz, C-2 H), 6.69 (d, 1H, J ) 8.1
Hz, C-1 H), 7.37-7.51 (m, 6H, C-4′ H, C-5′ phenyl-H), 8.72-
8.73 (m, 1H, C-6′ H); MS m/z 413 (MH)+. Anal. (C26H24N2O3‚
0.5H2O) C, H, N.
Exp er im en ta l Section
Gen er a l Meth od s. Melting points were determined in open
capillary tubes with a Mel-Temp melting point apparatus and
1
are uncorrected. H NMR spectra were recorded on a Nicolet
300NB spectrometer operating at 300.635 MHz. Chemical
shifts are expressed in parts per million downfield from
tetramethylsilane. Spectral assignments were supported by
proton decoupling. Mass spectra were recorded on a Varian
MAT 311A double-focusing mass spectrometer in the fast atom
bombardment (FAB) mode or on a Bruker BIOTOF II in
electrospray ionization (ESI) mode. Elemental analyses were
performed by Atlantic Microlab, Inc. (Atlanta, GA) or by the
Spectroscopic and Analytical Laboratory of Southern Research
Institute. Analytical results indicated by elemental symbols
were within (0.4% of the theoretical values. Thin-layer
chromatography (TLC) was performed on Analtech silica gel
GF 0.25 mm plates. Flash column chromatography was
performed with E. Merck silica gel 60 (230-400 mesh). Yields
are of purified compounds and were not optimized. Naloxone
hydrochloride, oxycodone hydrochloride, hydromorphone hy-
drochloride, and hydrocodone bitartrate were obtained from
Mallinckrodt. 2-(4-Chlorophenyl)malondialdehyde was pur-
chased from Acros Organics. All other reagents were obtained
from Aldrich.
17-(Allyl)-6,7-d id eh yd r o-3,14-d ih yd r oxy-4,5r-ep oxy-5′-
(4-ch lor op h en yl)p yr id o[2′,3′:6,7]m or p h in a n (7a ). A solu-
tion of naloxone hydrochloride (1.0 g, 2.74 mmol), 2-(4-
chlorophenyl)malondialdehdye (0.552 g, 3.02 mmol), and
ammonium acetate (0.421 g, 5.48 mmol) in acetic acid (20 mL)
was heated to reflux in an oil bath at 130-135 °C under an
argon atmosphere for 18 h. The reaction mixture was cooled
to room temperature, and the solvent was removed under
reduced pressure. The residue was treated with water, and
the pH of the mixture was adjusted to 8 with saturated
aqueous NaHCO3 solution. The solid that separated was
collected by filtration, dissolved in CH2Cl2, and washed with
brine. The organic layer was dried (Na2SO4) and filtered, and
the solvent was removed under reduced pressure. The crude
product was chromatographed over a column of silica using
CHCl3-MeOH-NH4OH (98.5:1:0.5) as the eluent to obtain
(0.385 g, 30%) the desired product 7a : mp 168-172 °C; TLC
Rf ) 0.2 (CH2Cl2-MeOH-NH4OH, 97:2.5:0.5); 1H NMR (CDCl3)
δ 1.82-1.85 (m, 1H, C-16 H), 2.31-2.43 (m, 2H, C-15 H, C-16
H), 2.62 (m, 4H, C-8 H2, C-10 H, C-15 H), 3.11-3.25 (m, 4H,
C-9 H, C-10 H, CH2 CHdCH2), 4.80-5.50 (broad hump, 2H,
C-3 OH, C-14 OH), 5.18-5.28 (m, 2H, CHdCH2), 5.59 (s, 1H,
C-5 H), 5.78-5.91 (m, 1H, CHdCH2), 6.59 (d, 1H, J ) 8.1 Hz,
C-1 H), 6.68 (d, 1H, J ) 8.1 Hz, C-2 H), 7.37-7.45 (m, 4H,

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1409
5′-(4-Ch lor oph en yl)-6,7-dideh ydr o-3,14-dih ydr oxy-4,5r-
ep oxy-17-m et h ylp yr id o[2′,3′:6,7]m or p h in a n (7d ). Oxy-
codone hydrochloride (1.0 g, 2.84 mmol) was reacted with 2-(4-
chlorophenyl)malondialdehdye (0.584 g, 3.13 mmol) and am-
monium acetate (0.438 g, 5.68 mmol) in acetic acid (20 mL)
by the same procedure as described for the preparation of 7a
to obtain 5′-(4-chlorophenyl)-6,7-didehydro-4,5R-epoxy-14-
hydroxy-3-methoxy-17-methylpyrido[2′,3′:6,7]morphinan (7o)
(0.527 g, 40%): mp 112-114 °C; TLC Rf ) 0.3 (CH2Cl2-
MeOH-NH4OH, 96.5:3:0.5); 1H NMR (CDCl3) δ 1.82-1.86 (m,
1H, C-15 H), 2.37-2.41 (m, 2H, C-15 H, C-16 H), 2.44 (s, 3H,
NCH3), 2.52-2.83 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.98 (d,
1H, J ) 6.4 Hz, C-9 H), 3.29 (d, 1H, J ) 18.7 Hz, C-10 H),
3.81 (s, 3H, OCH3), 4.7-4.8 (broad s, 1H, C-14 OH), 5.57 (s,
1H, C-5 H), 6.63 (d, 1H, J ) 8.1 Hz, C-2 H), 6.68 (d, 1H, J )
8.1 Hz, C-1 H), 7.34-7.47 (m, 5H, C-4′ H, C-2′′ H, C-3′′ H,
C-5′′ H, C-6′′ H), 8.74 (d, 1H, J ) 2.1 Hz, C-6′ H); MS m/z 461
(MH)+. Anal. (C27H25ClN2O3‚0.1H2O) C, H, N.
The methyl ether 7o (0.352 g, 0.76 mmol) in CH2Cl2 (15 mL)
was reacted with BBr3 (7.6 mL of 1 M solution in CH2Cl2, 7.6
mmol) as described for the preparation of 7b from 7m to yield
0.132 g (39%) of 7d : mp 196-198 °C; TLC Rf ) 0.3 (CH2Cl2-
MeOH-NH4OH, 94.5:5:0.5); 1H NMR (CDCl3) δ 1.80-1.88 (m,
1H, C-15 H), 2.34-2.40 (m, 2H, C-15 H, C-16 H), 2.44 (s, 3H,
NCH3), 2.49-2.81 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.98 (d,
1H, J ) 6.4 Hz, C-9 H), 3.27 (d, 1H, J ) 18.7 Hz, C-10 H),
4.8-5.7 (broad hump, 2H, C-3 OH, C-14 OH), 5.58 (s, 1H, C-5
H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H), 6.69 (d, 1H, J ) 8.1 Hz,
C-1 H), 7.34-7.44 (m, 4H, C-2′′ H, C-3′′ H, C-5′′ H, C-6′′ H),
7.47 (d, 1H, J ) 2.0 Hz, C-4′ H), 8.69 (d, 1H, J ) 2.0 Hz, C-6′
H); MS m/z 447 (MH)+. Anal. (C26H23ClN2O3‚0.5H2O) C, H, N.
6H, C-8 H2, C-10 H, C-14 H, C-16 H2), 2.69 (s, 3H, NCH3),
3.16 (d, 1H, J ) 18.7 Hz, C-10 H), 3.39-3.41 (m, 1H, C-9 H),
5.51 (s, 1H, C-5 H), 6.58 (d, 1H, J ) 8.1 Hz, C-2 H), 6.66 (d,
1H, J ) 8.1 Hz, C-1 H), 7.15 (dd, 1H, J ) 7.8 and 4.7 Hz, C-5′
H), 7.34 (m, 1H, C-4′ H), 8.52 (dd, 1H, J ) 4.7 and 1.1 Hz,
C-6′ H), 8.52-8.58 (br s, 1H, C-3 OH); MS m/z 321 (MH)+.
Anal. (C20H20N2O2‚0.6H2O) C, H, N.
6,7-Dideh ydr o-4,5r-epoxy-3-h ydr oxy-17-m eth yl-5′-ph en -
ylp yr id o[2′,3′:6,7]m or p h in a n (7g). Hydromorphone hydro-
chloride (1.0 g, 3.10 mmol) was reacted with 3-(dimethyl-
amino)-2-phenylacrolein40 (0.651 g, 3.72 mmol) and ammonium
acetate (0.477 g, 6.20 mmol) in acetic acid (20 mL) by the same
procedure as described for the preparation of 7a to obtain 7g
(0.54 g, 44%): mp 182-184 °C; TLC Rf ) 0.3 (CH2Cl2-MeOH-
NH4OH, 95:4.5:0.5); 1H NMR (CDCl3) δ 1.95-2.14 (m, 2H, C-15
H2), 2.31-2.67 (m, 6H, C-8 H2, C-10 H, C-14 H, C-16 H2), 2.46
(s, 3H, NCH3), 3.11 (d, 1H, J ) 18.7 Hz, C-10 H), 3.29-3.31
(m, 1H, C-9 H), 4.8-5.6 (broad hump, 1H, C-3 OH), 5.58 (s,
1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H), 6.68 (d, 1H, J )
8.1 Hz, C-1 H), 7.35-7.52 (m, 6H, C-4′ H, C-5′ phenyl-H), 8.74
(d, 1H, J ) 1.6 Hz, C-6′ H); MS m/z 397 (MH)+. Anal.
(C26H24N2O2‚0.6H2O) C, H, N.
5′-(4-Ch lor op h e n yl)-6,7-d id e h yd r o-4,5r-e p oxy-3-h y-
d r oxy-17-m eth ylp yr id o[2′,3′:6,7]m or p h in a n (7h ). Hydro-
morphone hydrochloride (10.0 g, 31.1 mmol) was reacted with
2-(4-chlorophenyl)malondialdehdye (6.81 g, 37.3 mmol) and
ammonium acetate (4.79 g, 62.2 mmol) in acetic acid (140 mL)
by the same procedure as described for the preparation of 7a
to obtain 7h (3.033 g, 23%): mp 188-190 °C; TLC Rf ) 0.35
(CH2Cl2-MeOH-NH4OH, 96.5:3:0.5); 1H NMR (CDCl3) δ
1.97-2.14 (m, 2H, C-15 H2), 2.30-2.48 (m, 3H, C-8 H, C-10
H, C-16 H), 2.46 (s, 3H, NCH3), 2.55-2.64 (m, 3H, C-8 H, C-14
H, C-16 H), 3.11 (d, 1H, J ) 18.6 Hz, C-10 H), 3.28-3.30 (m,
1H, C-9 H), 5.2-5.9 (broad hump, 1H, C-3 OH), 5.57 (s, 1H,
C-5 H), 6.60 (d, 1H, J ) 8.1 Hz, C-2 H), 6.68 (d, 1H, J ) 8.1
Hz, C-1 H), 7.40-7.46 (m, 5H, C-4′ H, C-2′′ H, C-3′′ H, C-5′′
H, C-6′′ H), 8.72 (d, 1H, J ) 2.1 Hz, C-6′ H); MS m/z 431 (MH)+.
Anal. (C26H23ClN2O2) C, H, N.
5′-(4-Br om oph en yl)-6,7-dideh ydr o-3,14-dih ydr oxy-4,5r-
ep oxy-17-m et h ylp yr id o[2′,3′:6,7]m or p h in a n (7e). Oxy-
codone hydrochloride (2.0 g, 5.68 mmol) was reacted with 2-(4-
bromophenyl)-3-(dimethylamino)acrolein41 (2.16 g, 8.52 mmol)
and ammonium acetate (1.31 g, 17.04 mmol) in acetic acid (30
mL) by the same procedure as described for the preparation
of 7a to obtain 5′-(4-bromophenyl)-6,7-didehydro-4,5R-epoxy-
14-hydroxy-3-methoxy-17-methylpyrido[2′,3′:6,7]morphinan (7p)
(0.79 g, 28%): mp >230 °C; TLC Rf ) 0.4 (CH2Cl2-MeOH-
NH4OH, 94.5:5:0.5); 1H NMR (CDCl3) δ 1.82-1.85 (m, 1H, C-15
H), 2.37-2.41 (m, 2H, C-15 H, C-16 H), 2.44 (s, 3H, NCH3),
2.51-2.83 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.98 (d, 1H, J )
6.4 Hz, C-9 H), 3.28 (d, 1H, J ) 18.7 Hz, C-10 H), 3.81 (s, 3H,
OCH3), 4.5-5.2 (broad hump, 1H, C-14 OH), 5.56 (s, 1H, C-5
H), 6.63 (d, 1H, J ) 8.1 Hz, C-2 H), 6.68 (d, 1H, J ) 8.1 Hz,
C-1 H), 7.36-7.40 (m, 2H, C-2′′ H, C-6′′ H), 7.46 (d, 1H, J )
2.1 Hz, C-4′ H), 7.54-7.59 (m, 2H, C-3′′ H, C-5′′ H), 8.74 (d,
A solution of the compound in EtOH was treated with a 2
M solution of hydrogen chloride in Et2O. Removal of the
solvent under reduced pressure and trituration with Et2O gave
the 7h ‚2HCl salt: mp 276-278 °C dec; MS m/z 431 (MH)+.
Anal. (C26H23ClN2O2‚2HCl‚2H2O) C, H, N.
5′-(4-Br om op h e n yl)-6,7-d id e h yd r o-4,5r-e p oxy-3-h y-
d r oxy-17-m eth ylp yr id o[2′,3′:6,7]m or p h in a n (7i). Hydro-
morphone hydrochloride (3.0 g, 9.32 mmol) was reacted with
2-(4-bromophenyl)-3-(dimethylamino)acrolein41 (2.60 g, 10.25
mmol) and ammonium acetate (1.45 g, 18.64 mmol) in acetic
acid (60 mL) by the same procedure as described for the
preparation of 7a to obtain 7i (0.93 g, 21%): mp 186-188 °C;
1H, J ) 2.1 Hz, C-6′ H); MS m/z 505 (MH)+. Anal. (C27H25
BrN2O3) C, H, N.
-
1
The methyl ether 7p (0.508 g, 1.0 mmol) in CH2Cl2 (20 mL)
was reacted with BBr3 (10.0 mL of 1 M solution in CH2Cl2,
10.0 mmol) as described for the preparation of 7b from 7m to
yield 0.198 g (40%) of 7e: mp 196-198 °C; TLC Rf ) 0.3
(CH2Cl2-MeOH-NH4OH, 94.5:5:0.5); 1H NMR (CDCl3) δ
1.82-1.85 (m, 1H, C-15 H), 2.38-2.41 (m, 2H, C-15 H, C-16
H), 2.44 (s, 3H, NCH3), 2.53-2.81 (m, 4H, C-8 H2, C-10 H,
C-16 H), 2.97 (d, 1H, J ) 6.4 Hz, C-9 H), 3.27 (d, 1H, J ) 18.7
Hz, C-10 H), 4.4-5.8 (broad hump, 2H, C-3 OH, C-14 OH),
5.57 (s, 1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H), 6.69 (d,
1H, J ) 8.1 Hz, C-1 H), 7.34-7.38 (m, 2H, C-2′′ H, C-6′′ H),
7.48 (d, 1H, J ) 2.0 Hz, C-4′ H), 7.54-7.58 (m, 2H, C-3′′ H,
C-5′′ H), 8.68 (d, 1H, J ) 2.0 Hz, C-6′ H); MS m/z 491 (MH)+.
Anal. (C26H23 BrN2O3‚0.25H2O) C, H, N.
6,7-Did eh yd r o-4,5r-ep oxy-3-h yd r oxy-17-m eth ylp yr id o-
[2′,3′:6,7]m or p h in a n (7f). A mixture of hydromorphone hy-
drochloride (1.0 g, 3.10 mmol), 3-(dimethylamino)acrolein
(0.369 g, 3.72 mmol), ammonium acetate (0.477 g, 6.20 mmol),
and acetic acid (20 mL) was refluxed in an oil bath at 130-
135 °C for 18 h. Workup of the reaction mixture and purifica-
tion of the crude product as described for the preparation of
7a gave the desired product 7f (0.215 g, 22%): mp 164-166
°C; TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH, 95:4.5:0.5); 1H
NMR (CDCl3) δ 2.01-2.34 (m, 2H, C-15 H2), 2.54-2.77 (m,
TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH, 94.5:5:0.5); H NMR
(CDCl3) δ 1.99-2.13 (m, 2H, C-15 H2), 2.27-2.44 (m, 1H,
C-16H), 2.46 (s, 3H, NCH3), 2.33-3.08 (broad hump, 2H, C-3
OH, C-14 H), 2.49-2.65 (m, 4H, C-8 H2, C-10 H, C-16 H), 3.11
(d, 1H, J ) 18.7 Hz, C-10 H), 3.31 (dd, 1H, J ) 5.8 and 2.4
Hz, C-9 H), 5.55 (s, 1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2
H), 6.68 (d, 1H, J ) 8.1 Hz, C-1 H), 7.31-7.35 (m, 2H, C-3′′ H,
C-5′′ H), 7.41 (d, 1H, J ) 2.0 Hz, C-4′ H), 7.53-7.57 (m, 2H,
C-2′′ H, C-6′′ H), 8.65 (d, 1H, J ) 2.0 Hz, C-6′ H); MS m/z 475
(MH)+. Anal. (C26H23BrN2O2‚0.5H2O) C, H, N.
5′-(3,4-Dich lor op h en yl)-6,7-d id eh yd r o-4,5r-ep oxy-3-h y-
d r oxy-17-m eth ylp yr id o[2′,3′:6,7]m or p h in a n (7j). Hydro-
morphone hydrochloride (1.00 g, 3.1 mmol) was reacted with
2-(3,4-dichlorophenyl)-3-(dimethylamino)acrolein40 (1.135 g,
4.65 mmol) and ammonium acetate (0.478 g, 6.2 mmol) in
acetic acid (20 mL) by the same procedure as described for
the preparation of 7a to obtain 7j (0.25 g, 17%): mp 184-186
°C; TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH, 94.5:5:0.5); 1H
NMR (CDCl3) δ 1.97-2.14 (m, 2H, C-15 H2), 2.30-2.48 (m,
3H, C-8 H, C-10 H, C-16 H), 2.46 (s, 3H, NCH3), 2.55-2.64
(m, 3H, C-8 H, C-14 H, C-16 H), 3.12 (d, 1H, J ) 18.7 Hz,
C-10 H), 3.27-3.30 (m, 1H, C-9 H), 4.8-5.6 (broad hump, 1H,
C-3 OH), 5.57 (s, 1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H),
6.68 (d, 1H, J ) 8.1 Hz, C-1 H), 7.34 (dd, 1H, J ) 8.3 and 2.1

1410 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
Hz, C-5 H), 7.44 (d, 1H, J ) 1.9 Hz, C-4 H′), 7.52 (d, 1H, J )
8.3 Hz, C-6′′ H), 7.59 (d, 1H, J ) 2.2 Hz, C-2′′ H), 8.70 (d, 1H,
J ) 1.8 Hz, C-6′ H); MS m/z 465 (MH)+. Anal. (C26H22Cl2N2O2‚
0.5H2O) C, H, N.
solvent under reduced pressure gave the crude product, which
was purified by chromatography over a column of silica using
CH2Cl2-MeOH-NH4OH (98.5:1:0.5) as the eluent to give 5′-
(4-chlorophenyl)-17-(cyclopropylmethyl)-6,7-didehydro-4,5R-
epoxy-3-methoxypyrido[2′,3′:6,7]morphinan (7q) (0.93 g, 76%):
mp 196-198 °C; TLC Rf ) 0.5 (CH2Cl2-MeOH-NH4OH, 96.5:
5′-(2,4-Dich lor op h en yl)-6,7-d id eh yd r o-4,5r-ep oxy-3-h y-
d r oxy-17-m eth ylp yr id o[2′,3′:6,7]m or p h in a n (7k ). Hydro-
morphone hydrochloride (1.00 g, 3.1 mmol) was reacted with
2-(2,4-dichlorophenyl)-3-(dimethylamino)acrolein42 (1.135 g,
4.65 mmol) and ammonium acetate (0.716 g, 9.3 mmol) in
acetic acid (20 mL) by the same procedure as described for
the preparation of 7a to obtain 7k (0.464 g, 32%): mp 198-
200 °C; TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH, 94.5:5:0.5);
1H NMR (CDCl3) δ 1.96-2.15 (m, 2H, C-15 H2), 2.30-2.71 (m,
6H, C-8 H2, C-10 H, C-14 H, C-16 H2), 2.46 (s, 3H, NCH3),
3.11 (d, 1H, J ) 18.7 Hz, C-10 H), 3.30-3.33 (m, 1H, C-9 H),
5.56 (s, 1H, C-5 H), 6.61 (d, 1H, J ) 8.1 Hz, C-2 H), 6.69 (d,
1H, J ) 8.1 Hz, C-1 H), 7.20 (d, 1H, J ) 8.4 Hz, C-6′′ H), 7.31
(dd, 1H, J ) 8.2 and 2.1 Hz, C- 5′′ H), 7.38 (d, 1H, J ) 2.0
Hz, C-4′ H), 7.49 (d, 1H, J ) 2.1 Hz, C-3′′ H), 8.58 (d, 1H, J )
1.8 Hz, C-6′ H); MS m/z 465 (MH)+. Anal. (C26H22Cl2N2O2‚
0.5H2O) C, H, N.
1
3:0.5); H NMR (CDCl3) δ 0.14-0.19 and 0.53-0.59 (2m, 4H,
cyclopropyl CH2CH2), 0.85-0.89 (m, 1H, cyclopropyl CH),
1.97-2.15 (m, 2H, C-15 H2), 2.28-2.66 (m, 7H, C-8 H2, C-10
H, C-14 H, C-16 H, and NCH2-cyclopropyl), 2.81-2.86 (m, 1H,
C-16 H), 3.02 (d, 1H, J ) 18.7 Hz, C-10 H), 3.57-3.60 (m, 1H,
C-9 H), 3.81 (s, 3H, OCH3), 5.56 (s, 1H, C-5 H), 6.61 (d, 1H, J
) 8.1 Hz, C-2 H), 6.67 (d, 1H, J ) 8.1 Hz, C-1 H), 7.39-7.46
(m, 5H, C-4′ H, C-2′′ H, C-3′′ H, C-5′′ H, C-6′′ H), 8.75 (d, 1H,
J ) 1.6 Hz, C-6′ H); MS m/z 485 (MH)+. Anal. (C30H29ClN2O2)
C, H, N.
A solution of the methyl ether 7q (0.83 g, 1.71 mmol) in
CH2Cl2 (20 mL) was reacted with BBr3 (17.0 mL of 1 M
solution in CH2Cl2, 17.0 mmol) as described for the preparation
of 7b from 7m to yield 0.227 g (28%) of 7l: mp 178-180 °C;
1
TLC Rf ) 0.3 (CH2Cl2-MeOH-NH4OH, 96.5:3:0.5); H NMR
(CDCl3) δ 0.13-0.19 and 0.51-0.57 (2m, 4H, cyclopropyl
CH2CH2), 0.85-0.93 (m, 1H, cyclopropyl CH), 1.96-2.16 (m,
2H, C-15 H2), 2.28-2.66 (m, 7H, C-8 H2, C-10 H, C-14 H, C-16
H, and NCH2-cyclopropyl), 2.82-2.87 (m, 1H, C-16 H), 3.00
(d, 1H, J ) 18.7 Hz, C-10 H), 3.60-3.62 (m, 1H, C-9 H), 5.2-
5.8 (broad hump, 1H, C-3 OH), 5.57 (s, 1H, C-5 H), 6.58 (d,
1H, J ) 8.1 Hz, C-2 H), 6.67 (d, 1H, J ) 8.1 Hz, C-1 H), 7.39-
7.45 (m, 5H, C-4′ H, C-2′′ H, C-3′′ H, C-5′′ H, C-6′′ H), 8.70 (d,
5′-(4-Ch lor op h en yl)-17-(cyclop r op ylm et h yl)-6,7-d id e-
h yd r o-4,5r-ep oxy-3-h yd r oxyp yr id o[2′,3′:6,7]m or p h in a n
(7l). Hydrocodone (5.838 g, 19.52 mmol), obtained from the
bitartrate salt by conventional methods, was reacted with 2-(4-
chlorophenyl)malondialdehdye (5.46 g, 29.28 mmol) and am-
monium acetate (4.51 g, 58.56 mmol) in acetic acid (100 mL)
by the same procedure as described for the preparation of 7a
to obtain 5′-(4-chlorophenyl)-6,7-didehydro-4,5R-epoxy-3-meth-
oxy-17-methylpyrido[2′,3′:6,7]morphinan (16) (4.70 g, 55%):
mp 234-237 °C; TLC Rf ) 0.57 (CHCl3-MeOH, 9:1); 1H NMR
(CDCl3) 1.99-2.13 (m, 2H, C-15 H, C-16 H), 2.32-2.54 (m 3H,
C-8 H, C-10 H, C-16 H), 2.46 (s, 3H, NCH3), 2.55-2.64 (m,
3H, C-8 H, C-10 H, C-16 H), 3.18 (d, 1H, J ) 18.5 Hz, C-10
H), 3.25-3.28 (m, 1H, C-9 H), 3.8 (s, 3H, OCH3), 5.5 (s, 1H,
C-5 H), 6.64 (d, 1H, J ) 8.2 Hz, C-1 H), 6.69 (d, 1H, J ) 8.2
Hz, C-2 H), 7.5-7.6 (m, 2H, C-3′′ H, C-5′′ H), 7.70-7.73 (m,
3H, C-3′ H, C-2′′ H, C-6′′ H), 8.74 (d, 1H, J ) 2.3 Hz, C-6′ H);
MS m/z 445 (MH)+. Anal. (C27H25ClN2O2‚0.2H2O) C, H, N.
To a solution of compound 16 (2.85 g, 6.61 mmol) in 1,2-
dichloroethane (50 mL) potassium carbonate (3.11 g, 22.5
mmol) was added. The mixture was stirred in an inert
atmosphere, and vinyl chloroformate (4.1 g, 38.52 mmol) was
added dropwise. The reaction mixture was refluxed for 36 h
and filtered. The filtrate was evaporated to dryness and the
residue was dissolved in ethanol (15 mL). To this solution was
added 2 N HCl (5.0 mL) and the mixture was refluxed for 2 h.
The solvent was removed under reduced pressure. The residue
was treated with water and the pH of the mixture was
adjusted to 7-8 by addition of saturated aqueous NaHCO3
solution. The mixture was then extracted with CHCl3 (4 × 100
mL). The extracts were combined, washed with brine, and
dried (Na2SO4). Filtration and removal of the solvent yielded
the crude product, which was purified by chromatography over
a column of silica using CH2Cl2-MeOH-NH4OH 98.5:1:0.5
as the eluent to obtain 5′-(4-chlorophenyl)-6,7-didehydro-
4,5R-epoxy-3-methoxypyrido[2′,3′:6,7]morphinan (17) (1.47 g,
53%): mp 246-248 °C; TLC Rf ) 0.44 (CHCl3-MeOH, 9:1);
1H NMR (DMSO-d6) δ 1.69-1.73 (m, 1H, C-15 H), 1.82-1.92
(m, 1H, C-15 H), 1.99-2.08 (m, 1H, C-16 H), 2.36-2.44 (m,
1H, C-14 H), 2.59-2.78 (m, 4H, C-8 H2, C-10 H, C-16 H), 2.87-
2.96 (m, 1H, C-10 H), 3.37-3.39 (m, 1H, C-9 H), 3.67 (s, 3H,
OCH3), 5.54 (s, 1H, C-5 H), 6.64 (d, 1H, J ) 8.2 Hz, C-1 H),
6.62 (d, 1H, J ) 8.1 Hz, C-2 H), 7.5-7.6 (m, 2H, C-3′′ H, C-5′′
H), 7.70-7.73 (m, 3H, C-3′ H, C-2′′ H, C-6′′ H), 8.8 (d, 1H, J )
2.1 Hz, C-6′ H); MS m/z 431 (MH)+. Anal. (C26H23ClN2O2‚
0.5H2O) C, H, N.
1H, J ) 2.0 Hz, C-6′ H); MS m/z 471 (MH)+. Anal. (C29H27
-
ClN2O2) C, H, N.
Biologica l Assa ys. Ra d ioliga n d Bin d in g Assa ys for µ,
δ, a n d K Recep tor s. The µ binding sites were labeled using
[3H]DAMGO (1-3 nM) and rat brain membranes as previously
described31 with several modifications. Rat membranes were
prepared each day using a partially thawed frozen rat brain
that was homogenized with a polytron in 10 mL/brain of ice-
cold 10 mM Tris-HCl, pH 7.0. Membranes were then centri-
fuged twice at 30000g for 10 min and resuspended with ice-
cold buffer following each centrifugation. After the second
centrifugation, the membranes were resuspended in 50 mM
Tris-HCl, pH 7.4 (50 mL/brain), at 25 °C. Incubations pro-
ceeded for 2 h at 25 °C in 50 mM Tris-HCl, pH 7.4, along with
a protease inhibitor cocktail (PIC). The nonspecific binding was
determined using 20 µM of levallorphan. The δ binding sites
were labeled using [3H]DADLE (2 nM) and rat brain mem-
branes as previously described,30 with several modifications.
Rat membranes were prepared each day using a partially
thawed frozen rat brain that was homogenized with a polytron
in 10 mL/brain of ice-cold 10 mM Tris-HCl, pH 7.0. Membranes
were then centrifuged twice at 30000g for 10 min and
resuspended with ice-cold buffer following each centrifugation.
After the second centrifugation, the membranes were resus-
pended in 50 mM Tris-HCl, pH 7.4 (50 mL/brain), at 25 °C.
Incubations proceeded for 2 h at 25 °C in 50 mM Tris-HCl,
pH 7.4, containing 100 mM choline chloride, 3 mM MnCl2, 100
nM DAMGO to block binding to µ sites, and PIC. Nonspecific
binding was determined using 20 µM levallorphan. The κ
binding sites were labeled using [3H]U69,593 (2 nM) as
previously described,32 with several modifications. Guinea pig
brain membranes were prepared each day using partially
thawed guinea pig brain that was homogenized with a polytron
in 10 mL/brain of ice-cold 10 mM Tris-HCl, pH 7.0. The
membranes were then centrifuged twice at 30000g for 10 min
and resuspended with ice-cold buffer following each centrifu-
gation. After the second centrifugation, the membranes were
resuspended in 50 mM Tris-HCl, pH 7.4 (75 mL/brain), at 25
°C. Incubations proceeded for 2 h at 25 °C in 50 mM Tris-
HCl, pH 7.4, containing 1 µg/mL of captopril and PIC.
Nonspecific binding was determined using 1 µM U69,593. Each
3H ligand was displaced by 8-10 concentrations of test drug,
two times. Compounds were prepared as 1 mM solution with
10 mM Tris buffer (pH 7.4) containing 10% DMSO before drug
dilution. All drug dilutions were done in 10 mM Tris-HCl, pH
Compound 17 (1.37 g, 3.19 mmol) was dissolved in ethanol
(70 mL), and NaHCO3 (5.33 g, 6.37 mmol) was added. To this
mixture was added cyclopropylmethyl bromide (2.16 g, 16.0
mmol), and the reaction mixture was refluxed under nitrogen
for 16 h. The mixture was then concentrated, and water (180
mL) was added to the residue. The mixture was extracted with
CHCl3 (4 × 100 mL) and dried over Na2CO3. Removal of the

Identification of Opioid Ligands
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1411
7.4, containing 1 mg/mL bovine serum albumin. All washes
were done with ice-cold 10 mM Tris-HCl, pH 7.4.
for 30 min. In the testing of compounds 7d , 7h , 7i, and 7k , a
concentration causing less than 20% inhibition of contraction
height was used in the 30 min incubation to allow for a more
accurate determination of the antagonist potency against
DPDPE in the MVD. After the 30 min incubation, the tissues
were washed with fresh buffer for 30 min, and the agonist
dose-response curve was repeated. Rightward shifts in the
dose-response curves were calculated by dividing the antago-
nized dose-response curve IC50 value by the unantagonized
IC50 value. IC50 values represent the mean of two to four
tissues. IC50 estimates and their associated standard errors
were determined by using a computerized nonlinear least-
squares method.44
An tin ocicep tive Stu d ies. Male ICR mice (Harlan) were
used for all evaluations. Mice were housed in a temperature
and humidity controlled vivarium on a (12 h):(12 h) light:dark
cycle with unlimited access to food and water prior to the
formal procedures. Graded doses of morphine or the test
compounds were injected intracerebroventricularly (icv) under
light ether anesthesia.21 Morphine sulfate was dissolved in
distilled water and injected in a volume of 5 µL. The dihydro-
chloride salt of 7h was dissolved in water and injected in a
volume of 5 µL. All other compounds were dissolved in 100%
DMSO and injected in a volume of 5 µL. Antinociceptive assays
were performed at various times after injection.
Wa r m -Wa ter Ta il-With d r a w a l Assa y. Naive mice were
baselined in the 55 °C warm-water tail-withdrawal test as
previously described.21,45 Doses of morphine or the test com-
pound were injected icv, and antinociception was assessed at
10, 20, 30, 45, 60, 80, 120, and 180 min postinjection. Percent
antinociception was calculated using the formula %MPE
(maximal possible effect) ) 100 × (test - control)/(cutoff -
control) where control is the predrug observation, test is the
postdrug observation, and cutoff is the maximal length of
stimulus allowed (10 s for 55 °C tail-withdrawal). Antino-
ciceptive A50 values and 95% confidence intervals were deter-
mined using linear regression software (FlashCalc). Opioid
activity of the test compounds was assessed by pretreating
animals with naloxone (10 mg/kg ip, -10 min) followed by an
icv injection of an approximate A90 dose of test compound. If a
compound did not produce a full agonist effect, then the dose
that produced the greatest antinociceptive effect was used.
Antinociception was assessed in the 55 °C warm-water tail-
withdrawal test at 10, 20, and 30 min. A positive response to
a fixed dose of naloxone was indicated when greater than 80%
reduction in the antinociceptive effect of the agonist was
observed.
[
35S]GTP -γ-S F u n ction a l Assa ys. The [35S]GTP-γ-S bind-
ing assays were performed according to the methods described
previously.35 Guinea pig caudate membranes (10-20 µg of
protein in 300 µL of 50 mM Tris-HCl, pH 7.4, with 1.67 mM
DTT and 0.15% BSA) were added to polystyrene 96-well plates
filled with 200 µL of a reaction buffer containing 50 mM Tris-
HCl, pH 7.4, 100 mM NaCl, 10 mM MgCl2, 1 mM EDTA, 100
µM GDP, 0.1% BSA, 0.05-0.1 nM [35S]GTP-γ-S, and varying
concentrations of drugs. The reaction mixture was incubated
for 3 h at 22 °C (equilibrium). The reaction was terminated
by the addition of 0.5 mL of ice-cold Tris-HCl, pH 7.4 (4 °C),
followed by rapid vacuum filtration through Whatman GF/B
filters previously soaked in ice-cold Tris-HCl, pH 7.4 (4 °C).
The filters were washed twice with 0.5 mL of ice-cold H2O (4
°C). Bound radioactivity was counted at an efficiency of 98%
by liquid scintillation spectroscopy. Nonspecific binding was
determined in the presence of 10 µM GTP-γ-S.
In initial screening experiments, each test agent was tested
to determine agonist and antagonist activity using a 10 µM
concentration in the absence and presence of selective antago-
nists (6000 nM CTAP, 6 nM nor-BNI or 20 nM NTI or 500
nM TIPP) and selective agonists (10 µM SNC-80, 10 µM
DAMGO, or 10 µM U69,593). Compounds showing significant
agonist activity were further characterized. In this case,
agonist dose-response curves (10 data points each) were
generated in the presence of selective antagonists using
previously defined “blocking” concentrations:34 µ receptors
(1000 nM TIPP, 6 nM nor-BNI), δ receptors (2000 nM CTAP,
6 nM nor-BNI), and κ receptors (1000 nM TIPP, 2000 nM
CTAP). Each curve was run with a 10 µM concentration of
the standard agonist (DAMGO, U69,593, or SNC-80). The data
were expressed as a percent of the stimulation produced by
the standard agonist. As described elsewhere,35,36 compounds
showing significant antagonist activity were further assessed
with full dose-response curves to determine the functional
Ki values for inhibition of agonist-stimulated [35S]GTP-γ-S
binding using 10 µM DAMGO, 10 µM SNC-80, or 10 µM
U69,593.
Da ta An a lysis. The data of the two separate experiments
(opioid binding assays) or three experiments ([35S]GTP-γ-S
assays) were pooled and fit by applying the nonlinear least-
squares curve-fitting program MLAB-PC (Civilized Software,
Bethesda, MD) to the two-parameter logistic equation43 for the
best-fit estimates of the IC50 and slope factor. The Ki values
were then calculated using the equation Ki ) IC50/(1 + [L]/Kd).
Toler a n ce Regim en . Mice were injected twice daily (8 a.m.
and 8 p.m.) with an approximate A90 dose of morphine or A90
doses of 7h for 3 days. Antinociceptive dose-response curves
in the warm-water tail-withdrawal assay were generated on
the morning of the fourth day using the procedures outlined
above.
GP I a n d MVD Bioa ssa ys.37,38 Electrically induced smooth
muscle contractions of mouse vas deferens and strips of guinea
pig ileum longitudinal muscle myenteric plexus were used.
Tissues came from male ICR mice weighing 25-40 g and male
Hartley guinea pigs weighing 250-500 g. The tissues were
tied to a gold chain with suture silk, suspended in 20 mL baths
containing 37 °C oxygenated (95% O2, 5% CO2) Krebs bicar-
bonate solution (magnesium free for the MVD), and allowed
to equilibrate for 15 min. The tissues were then stretched to
optimal length with previously determined 1 g tension (0.5 g
for MVD) and allowed to equilibrate for 15 min. The tissues
were stimulated transmurally between platinum wire elec-
trodes at 0.1 Hz with 0.4 ms pulses (2 ms pulses for MVD)
and supramaximal voltage. An initial dose-response curve of
DPDPE or PL-017 was constructed at the start of each assay
to establish tissue effects, allowing each tissue to be used as
its own control. Tissues not producing typical results were not
used. Experimental compounds were added to the baths in 14-
60 µL volumes. Succeeding doses of agonist were added
cumulatively to the bath at 3 min intervals to produce a
concentration-response curve. The tissues were then washed
extensively with fresh buffer until the original contraction
height was reestablished. Agonist effects of the compounds at
1 µM were measured as the percent inhibition of contraction
height 10 min after addition to the bath. Antagonist effects to
DPDPE and PL-017 were assayed after incubation of the
tissues with 1 µM concentration of the compound in the bath
Ack n ow led gm en t . This investigation was sup-
ported by the National Institute on Drug Abuse, Grant
No. DA08883. We thank Dr. J ames M. Riordan, Mr.
Mark D. Richardson, Ms. J oan C. Bearden, and Ms.
J ackie W. Truss for analytical and spectral data. We
are grateful to Dr. J ohn A. Secrist III and Dr. J ohn A.
Montgomery for their encouragement, valuable com-
ments, and suggestions during the course of this work.
Refer en ces
(1) Parrott, T. Using Opioid Analgesics To Manage Chronic Non-
cancer Pain in Primary Care. J . Am. Board Fam. Pract. 1999,
12, 293-306.
(2) Cherny, N. New Strategies in Opioid Therapy for Cancer Pain.
J . Oncol. Manage. 2000, 9, 8-15.
(3) Dhawan, B. N.; Cesselin, F.; Raghubir, R.; Reisine, T.; Bradley,
P. B.; Portoghese, P. S.; Hamon, M. International Union of
Pharmacology. XII. Classification of Opioid Receptors. Pharma-
col. Rev. 1996, 48, 567-592.
(4) Aldrich, J . V.; Vigil-Cruz, S. C. Narcotic Analgesics. In Burger’s
Medicinal Chemistry and Drug Discovery, Abraham, D. J ., Ed.;
J ohn Wiley & Sons: New York, 2003; Vol. 6, pp 329-481.

1412 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 6
Ananthan et al.
(5) Duthie, D. J .; Nimmo, W. S. Adverse Effects of Opioid Analgesic
Drugs. Br. J . Anaesth. 1987, 59, 61-77.
(27) Portoghese, P. S.; Larson, D. L.; Sultana, M.; Takemori, A. E.
Opioid Agonist and Antagonist Activities of Morphindoles
Related to Naltrindole. J . Med. Chem. 1992, 35, 4325-4329.
(28) McLamore, S.; Ullrich, T.; Rothman, R. B.; Xu, H.; Dersch, C.;
Coop, A.; Davis, P.; Porreca, F.; J acobson, A. E.; Rice, K. C. Effect
of N-Alkyl and N-Alkenyl Substituents in Noroxymorphindole,
17-Substituted-6,7-dehydro-4,5R-epoxy-3,14-dihydroxy-6,7:2′,3′-
indolomorphinans, on Opioid Receptor Affinity, Selectivity, and
Efficacy. J . Med. Chem. 2001, 44, 1471-1474.
(29) Kshirsagar, T. A.; Fang, X.; Portoghese, P. S. 14-Desoxy Ana-
logues of Naltrindole and 7-Spiroindanyloxymorphone: The Role
of the 14-Hydroxy Group at δ Opioid Receptors. J . Med. Chem.
1998, 41, 2657-2660.
(30) Rothman, R. B.; Bykov, V.; Ofri, D.; Rice, K. C. LY164929: A
Highly Selective Ligand for the Lower Affinity [3H]D-Ala2-D-Leu5-
Enkephalin Binding Site. Neuropeptides 1988, 11, 13-16.
(31) Rothman, R. B.; Xu, H.; Seggel, M.; J acobson, A. E.; Rice, K. C.;
Brine, G. A.; Carroll, F. I. RTI-4614-4: An Analog of (+)-cis-
3-Methylfentanyl with a 27,000-Fold Binding Selectivity for
Mu Versus Delta Opioid Binding Sites. Life Sci. 1991, 48, 111-
116.
(6) van Ree, J . M.; Gerrits, M. A.; Vanderschuren, L. J . Opioids,
Reward and Addiction: An Encounter of Biology, Psychology,
and Medicine. Pharmacol. Rev. 1999, 51, 341-396.
(7) Traynor, J . R.; Elliott, J . δ-Opioid Receptor Subtypes and Cross-
Talk with µ-Receptors. Trends Pharmacol. Sci. 1993, 14, 84-
86.
(8) Rothman, R. B.; Holaday, J . W.; Porreca, F. Allosteric Coupling
among Opioid Receptors: Evidence for an Opioid Receptor
Complex. In Opioids I; Herz, A., Akil, H., Simon, E. J ., Eds;
Handbook of Experimental Pharmacology, Volume 104; Springer-
Verlag: Berlin, 1993; pp 217-237.
(9) J ordan, B. A.; Devi, L. A. G-Protein-Coupled Receptor Hetero-
dimerization Modulates Receptor Function. Nature 1999, 399,
697-700.
(10) George, S. R.; Fan, T.; Xie, Z.; Tse, R.; Tam, V.; Varghese, G.;
O’Dowd, B. F. Oligomerization of µ- and δ-Opioid Receptors. J .
Biol. Chem. 2000, 275, 26128-26135.
(11) Levac, B. A. R.; O’Dowd, B. F.; George, S. R. Oligomerization of
Opioid Receptors: Generation of Novel Signaling Units. Curr.
Opin. Pharmacol. 2002, 2, 76-81.
(12) Abdelhamid, E. E.; Sultana, M.; Portoghese, P. S.; Takemori,
A. E. Selective Blockage of Delta Opioid Receptors Prevents the
Development of Morphine Tolerance and Dependence in Mice.
J . Pharmacol. Exp. Ther. 1991, 258, 299-303.
(13) Fundytus, M. E.; Schiller, P. W.; Shapiro, M.; Weltrowska, G.;
Coderre, T. J . Attenuation of Morphine Tolerance and Depend-
ence with the Highly Selective δ-Opioid Receptor Antagonist
TIPP[ψ]. Eur. J . Pharmacol. 1995, 286, 105-108.
(32) Rothman, R. B.; Bykov, V.; de Costa, B. R.; J acobson, A. E.; Rice,
K. C.; Brady, L. S. Interaction of Endogenous Opioid Peptides
and Other Drugs with Four Kappa Opioid Binding Sites in
Guinea Pig Brain. Peptides 1990, 11, 311-331.
(33) Ananthan, S.; J ohnson, C. A.; Carter, R. L.; Clayton, S. D.; Rice,
K. C.; Xu, H.; Davis, P.; Porreca, F.; Rothman, R. B. Synthesis,
Opioid Receptor Binding, and Bioassay of Naltrindole Analogues
Substituted in the Indolic Benzene Moiety. J . Med. Chem. 1998,
41, 2872-2881.
(34) Thomas, J . B.; Atkinson, R. N.; Herault, X. M.; Rothman, R. B.;
Mascarella, S. W.; Dersch, C. M.; Xu, H.; Horel, R. B.; Carroll,
F. I. Optically Pure (-)-4-[(N-Allyl-3-methyl-4-piperidinyl)phenyl-
amino]-N,N-diethylbenzamide Displays Selective Binding and
Full Agonist Activity for the δ Opioid Receptor. Bioorg. Med.
Chem. Lett. 1999, 9, 3347-3350.
(35) Thomas, J . B.; Fall, M. J .; Cooper, J . B.; Rothman, R. B.;
Mascarella, S. W.; Xu, H.; Partilla, J . S.; Dersch, C. M.;
McCullough, K. B.; Cantrell, B. E.; Zimmerman, D. M.; Carroll,
F. I. Identification of an Opioid κ Receptor Subtype-Selective
N-Substituent for (+)-(3R,4R)-Dimethyl-4-(3-hydroxyphenyl)-
piperidine. J . Med. Chem. 1998, 41, 5188-5197.
(14) Schiller, P. W.; Weltrowska, G.; Schmidt, R.; Nguyen, T. M.-D.;
Berezowska, I.; Lemieux, C.; Chung, N. N.; Carpenter, K. A.;
Wilkes, B. C. Four Different Types of Opioid Peptides with Mixed
µ Agonist/δ Antagonist Properties. Analgesia 1995, 1, 703-706.
(15) Schiller, P. W.; Fundytus, M. E.; Merovitz, L.; Weltrowska, G.;
Nguyen, T. M.-D.; Lemieux, C.; Chung, N. N.; Coderre, T. J . The
Opioid µ Agonist/δ Antagonist DIPP-NH2[ψ] Produces a Potent
Analgesic Effect, No Physical Dependence, and Less Tolerance
Than Morphine in Rats. J . Med. Chem. 1999, 42, 3520-3526.
(16) Suzuki, T.; Ikeda, H.; Tsuji, M.; Misawa, M.; Narita, M.; Tseng,
L. F. Antisense Oligodeoxynucleotide to δ Opioid Receptors
Attenuates Morphine Dependence in Mice. Life Sci. 1997, 61,
165-170.
(17) Sanchez-Blazquez, P.; Garcia-Espana, A.; Garzon, J . Antisense
Oligodeoxynucleotides to Opioid Mu and Delta Receptors Re-
duced Morphine Dependence in Mice: Role of Delta-2 Opioid
Receptors. J . Pharmacol. Exp. Ther. 1997, 280, 1423-1431.
(18) Zhu, Y.; King, M. A.; Schuller, A. G. P.; Nitsche, J . F.; Reidl,
M.; Elde, R. P.; Unterwald, E.; Pasternak, G. W.; Pintar, J . E.
Retention of Supraspinal Delta-like Analgesia and Loss of
Morphine Tolerance in δ Opioid Receptor Knockout Mice. Neuron
1999, 24, 243-252.
(19) Ananthan, S.; Kezar, H. S., III; Carter, R. L.; Saini, S. K.; Rice,
K. C.; Wells, J . L.; Davis, P.; Xu, H.; Dersch, C. M.; Bilsky, E.
J .; Porreca, F.; Rothman, R. B. Synthesis, Opioid Receptor
Binding, and Biological Activities of Naltrexone-Derived Pyrido-
and Pyrimidomorphinans. J . Med. Chem. 1999, 42, 3527-3538.
(20) Ananthan, S.; Kezar, H. S., III; Saini, S. K.; Khare, N. K.; Davis,
P.; Dersch, C. M.; Porreca, F.; Rothman, R. B. Synthesis, Opioid
Receptor Binding, and Functional Activity of 5′-Substituted 17-
Cyclopropylmethylpyrido[2′,3′:6,7]morphinans. Bioorg. Med. Chem.
Lett. 2003, 13, 529-532.
(21) Wells, J . L.; Bartlett, J . L.; Ananthan, S.; Bilsky, E. J . In Vivo
Pharmacological Characterization of SoRI 9409, a Nonpeptidic
Opioid µ-Agonist/δ-Antagonist That Produces Limited Antino-
ciceptive Tolerance and Attenuates Morphine Physical Depend-
ence. J . Pharmacol. Exp. Ther. 2001, 297, 597-605.
(22) Xu, H.; Lu, Y.-F.; Rice, K. C.; Ananthan, S.; Rothman, R. B. SoRI
9409, a Non-Peptide Opioid µ Receptor Agonist/δ Receptor
Antagonist, Fails To Stimulate [35S]-GTP-γ-S Binding at Cloned
Opioid Receptors. Brain Res. Bull. 2001, 55, 507-511.
(23) Ananthan, S.; Khare, N. K.; Saini, S. K.; Davis, P.; Dersch, C.
M.; Porreca, F.; Rothman, R. B. Novel Ligands for the Opioid
Receptors: Synthesis and Structure-Activity Relationships
among 5′-Aryl and 5′-Heteroaryl 17-Cyclopropylmethyl-4,5R-
epoxypyrido[2′,3′:6,7]morphinans. Bioorg. Med. Chem. 2003, 11,
4143-4154.
(24) Casy, A. F.; Partfitt, R. T. Opioid Analgesics: Chemistry and
Receptors; Plenum Press: New York, 1986.
(25) Lenz, G. R.; Evans, S. M.; Walters, D. E.; Hopfinger, A. J .
Opiates; Academic Press: Orlando, FL, 1986.
(26) Maguire, P. A.; Perez, J . J .; Tsai, N. F.; Rodriguez, L.; Beatty,
M. F.; Villar, H. O.; Kamal, J . J .; Upton, C.; Casy, A. F.; Loew,
G. H. Molecular Mechanism of δ-Selectivity of Indole Analogs
of Nonpeptide Opioids. Mol. Pharmacol. 1993, 44, 1246-1251.
(36) Partilla, J . S.; Carroll, F. I.; Thomas, J . B.; Rice, K. C.;
Zimmerman, D. M.; Rothman, R. B. Opioid Peptide Receptor
Studies. 13. Characterization of Opioid Antagonists with the
[
35S]GTP-γ-S Binding Assay. Analgesia 1999, 4, 27-32.
(37) Kramer, T. H.; Davis, P.; Hruby, V. J .; Burks, T. F.; Porreca, F.
In vitro Potency, Affinity and Agonist Efficacy of Highly Selective
Delta Opioid Receptor Ligands. J . Pharmacol. Exp. Ther. 1993,
266, 577-584.
(38) Porreca, F.; LoPresti, D.; Ward, S. J . Opioid Agonist Affinity in
the Guinea-Pig Ileum and Mouse Vas Deferens. Eur. J . Phar-
macol. 1990, 179, 129-139.
(39) Xu, H.; Hashimoto, A.; Rice, K. C.; J acobson, A. E.; Thomas, J .
B.; Carroll, F. I.; Lai, J .; Rothman, R. B. Opioid Peptide Receptor
Studies. 14. Stereochemistry Determines Agonist Efficacy and
Intrinsic Efficacy in the [35S]GTP-γ-S Functional Binding Assay.
Synapse 2001, 39, 64-69.
(40) Coppola, G. M.; Hardtmann, G. E.; Huegi, B. S. Synthesis and
Reactions of 2-Aryl-3-(dimethylamino)acroleins. J . Heterocycl.
Chem. 1974, 11, 51-56.
(41) Stacey, G. J .; Hawkins, A. F.; Pearson, D. P. J .; Sunley, R. L.
Pyridine Derivatives Inducing Tillering and Agricultural Com-
positions Containing Them. Eur. Pat. Appl. 67511, 1982; Chem.
Abstr. 1983, 98, 198028.
(42) Biziere, K.; Chambon, J . P.; Hallot, A. Heterocyclic Nitrogen
Compounds. Eur. Pat. Appl. 169139, 1986; Chem. Abstr. 1986,
105, 97319.
(43) Rodbard, D.; Lenox, R. H.; Wray, H. L.; Ramseth, D. Statistical
Characterization of the Random Errors in the Radioimmuno-
assay Dose-Response Variable. Clin. Chem. 1976, 22, 350-358.
(44) MINSQ Least Squares Parameter Estimation, version 3.05;
MicroMath Inc.: St. Louis, MO, 1999.
(45) Bilsky, E. J .; Inturrisi, C. E.; Sadee, W.; Hruby, V. J .; Porreca,
F. Competitive and Non-Competitive NMDA Antagonists Block
the Development of Antinociceptive Tolerance to Morphine, but
Not to Selective µ or δ Opioid Agonists in Mice. Pain 1996, 68,
229-237.
J M030311V