Syntheses and opioid receptor binding affinities of 8-amino-2,6-methano-3-benzazocines

Article abstract of DOI:10.1021/jm020429w

8-Amino-2,6-methano-3-benzazocine derivatives have been made using Pd-catalyzed amination procedures, and their affinities for opioid receptors were assessed. The 8-amino group was hypothesized to be a replacement for the prototypic 8-OH substituent for 2

Full text of DOI:10.1021/jm020429w

838
J . Med. Chem. 2003, 46, 838-849
Syn th eses a n d Op ioid Recep tor Bin d in g Affin ities of
8-Am in o-2,6-m eth a n o-3-ben za zocin es
Mark P. Wentland,*,‡ Yingchun Ye,‡ Christopher L. Cioffi,‡ Rongliang Lou,‡ Qun Zhou,‡ Guoyou Xu,‡
Wenhu Duan,‡ Christoph M. Dehnhardt,‡ Xufeng Sun,‡ Dana J . Cohen,§ and J ean M. Bidlack§
Department of Chemistry, Rensselaer Polytechnic Institute, Troy, New York 12180, and Department of Pharmacology and
Physiology, School of Medicine and Dentistry, University of Rochester, Rochester, New York 14642
Received September 24, 2002
8-Amino-2,6-methano-3-benzazocine derivatives have been made using Pd-catalyzed amination
procedures, and their affinities for opioid receptors were assessed. The 8-amino group was
hypothesized to be a replacement for the prototypic 8-OH substituent for 2,6-methano-3-
benzazocines and related opiates. This OH group is generally required for binding yet is
implicated in unfavorable pharmacokinetic characteristics such as low oral bioavailability and
rapid clearance via O-glucuronidation. The core structures in which the 8-OH group was
replaced were cyclazocine and its enantiomers, ethylketocyclazocine and its enantiomers,
ketocyclazocine, and Mr2034. Many new analogues had high affinity for opioid receptors with
several in the subnanomolar range. Highest affinity was seen in analogues with secondary
8-(hetero)arylamino appendages. Binding to opioid receptors was enantioselective with the
(2R,6R,11R)-configuration preferred and high selectivity for µ and κ over δ opioid receptors
was observed within the series. Several derivatives were shown to have intrinsic opioid-receptor-
mediated activity in [35S]GTPγS assays.
In tr od u ction
drug-seeking behavior.10 Owing, in part, to dysphoric
side effects and a short duration of analgesic action,
further clinical development of cyclazocine ceased.1,10,11
The short duration of action observed for cyclazocine in
animals and humans may be due to O-glucuronidation;
this metabolite has been observed in humans11 and
animals.12,13
Cyclazocine [(()-1] is a member of the 2,6-methano-
3-benzazocine (aka benzomorphan) class of opioid recep-
tor-interactive agents.1-8 Numerous breakthroughs in
understanding opioid receptor pharmacology (e.g., de-
duction of multiple opioid receptor subtypes9) have been
made, in part through the study of cyclazocine and its
enantiomers (-)-2 and (+)-3.1In the 1960s and early
Opioid receptor binding studies for cyclazocine per-
formed in the late 1970s showed the compound to have
high affinity for κ and µ opioid receptors, and antinoci-
ceptive studies in rodents revealed that cyclazocine was
a κ agonist and µ antagonist.1 The opioid receptor
binding properties of cyclazocine, a racemate, reside in
the (2R,6R,11R)-isomer (-)-2, whereas the (2S,6S,11S)-
isomer (+)-3 displays high affinity for the σ receptor.14,15
The observations that cyclazocine produced opioid-
induced analgesia in humans with a lower risk of
dependency and to precipitate a somewhat milder
withdrawal from morphine suggested that the agent
might serve as a useful therapeutic for the treatment
of heroin addiction.16-18 Currently, cyclazocine is un-
dergoing NIDA-sponsored evaluation for potential clini-
cal utility as a treatment for cocaine abuse.19 Results
from this study did not support the use of cyclazocine
for this indication; however, there was some suggestion
that an analogue having a somewhat different opioid
receptor profile might be useful.
1970s, cyclazocine was evaluated in humans for anal-
gesia and as a possible treatment in postaddicts of
heroin for preventing relapse.1,10 Potent analgesia was
observed in humans dosed with cyclazocine, and after
abrupt cyclazocine withdrawal, patients did not display
In an attempt some years ago to enhance the phar-
macokinetic properties of cyclazocine (i.e., retard O-
glucuronidation and increase duration of action), we
prepared (()-4 where the 8-OH group of cyclazocine was
replaced with NH2.11 Relative to cyclazocine, we found
that (()-4 had somewhat diminished antinociception
potency in mice when delivered by the subcutaneous
route but had comparable potency when delivered
orally.11 For example, in the mouse acetylcholine writh-
* To whom correspondence should be addressed. Phone: (518) 276-
2234. Fax: (518) 276-4887. E-mail: wentmp@rpi.edu.
‡ Rensselaer Polytechnic Institute.
§ University of Rochester.
10.1021/jm020429w CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/01/2003

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 839
ing test, the po/sc ratio of ED50 values for cyclazocine
was 35 and for (()-4 this ratio was 4, indicating the
8-OH f NH2 change accounted for higher oral efficacy
than would be predicted by the sc data. Whether higher
gut wall permeability, lower first-pass or extrahepatic
metabolism, reduced clearance, or some other factor
accounts for the high oral efficacy of (()-4 is not known.
More recently, we assessed the affinities of a small
series of 8-amino-2,6-methano-3-benzazocines for opioid
receptors.20 Results from this preliminary study indi-
cated that replacing the 8-OH of cyclazocine with an
8-NH2 group conferred significantly lower affinity (>20-
fold) for µ and κ opioid receptors. However, by substitut-
ing one H of the 8-NH2 group with phenyl, affinity for
µ and κ was dramatically improved where Ki values
were now close (4-fold higher) to cyclazocine. This led
us to conclude that a new SAR for the 8-position
substituent of 2,6-methano-3-benzazocines was emerg-
ing. From this small series, we chose two optically active
analogues, the primary amine (-)-4 and the N-phenyl
analogue (-)-11 for in vivo studies to assess their
antinociception potency icv and then to determine which
opioid receptors mediate antinociception and if they
acted as antagonists of antinociception mediated by any
of the opioid receptors.21 Compounds (-)-4 and (-)-11
had similar (to each other and to (-)-cyclazocine [(-)-
2]) antinociception potency (mouse acetic acid writhing)
that was mediated by the κ opioid receptor. Coupled
with the fact that these compounds antagonize morphine-
induced antinociception leads to the conclusion that, at
the least, two members of this series are µ antagonists
and κ agonists.21
besides cyclazocine, was also investigated including
ethylketocyclazocine (EKC), ketocyclazocine, and the
optically active 3-tetrahydrofuranylmethyl-2,6-methano-
3-benzazocine, Mr2034.
Besides our early studies11 and the preliminary ac-
count of this work,20 few reports of 8-amino-2,6-methano-
3-benzazocines or 3-aminomorphinans have appeared.
A small series of 8-amino-3-benzyl-2,6-methano-3-ben-
zazocines were evaluated by Carroll and co-workers for
binding to δ receptors.23 For morphinan-related core
structures, we recently reported the synthesis and
opioid binding properties of the 3-amino derivative of
morphine.24 In addition, the 3-sulfonamido analogues
(i.e., 3-NHSO2CH3) of naltrexone and oxymorphone were
recently reported25 and a 3-aminodextromorphan ana-
logue was studied for other (than opioid receptor bind-
ing) pharmacological properties.26 This latter compound,
however, has stereochemistry that is the opposite of that
of natural opiates.
Ch em istr y
Our original method11 (Birch reduction of cyclazocine
methyl ether followed by Semmler-Wolf methodology)
to make primary amino analogue (()-4 was used to
prepare its enantiomers27 (-)-5 and (+)-6 from the
known cyclazocine enantiomers28,29 (-)-2 and (+)-3,
respectively, in nearly identical yields.11 The Birch
reduction/Semmler-Wolf method, while very useful in
making the primary and certain secondary amino
derivatives, could not be used for the rapid introduction
of a variety of amine substituents at position 8. There-
fore, we applied the methodology of Buchwald,31
Hartwig,32 and co-workers using the Pd-catalyzed ami-
nation22 of aryl triflates to our system with excellent
results, which we recently reported in a preliminary
communication.20 By use of this methodology, our first
task was to remake primary amino analogue (()-4 from
cyclazocine (()-1. This three-step process that we note
as method A is shown in Scheme 1 and detailed in the
Experimental Section, and except for a few minor
modifications, it is identical to a literature method.33
Cyclazocine (()-1 was first converted to its correspond-
ing triflate ester (()-53 in 96% yield using (CF3CO)2O/
pyridine. Conversion of (()-53 to the corresponding
benzophenone imine adduct (()-38 was accomplished
in 47% yield using Ph2CdNH, Pd2(dba)3,34 DPPF, NaO-
t-Bu in toluene at 80 °C. Adduct (()-38 was converted
The objective of the study we are now reporting,
therefore, was to prepare an expanded series of 8-amino-
2,6-methano-3-benzazocines and to assess their affini-
ties for µ, δ, and κ opioid receptors in hopes of refining
a new SAR around the 8-position. We also tested
selected analogues for functional activity in the [35S]-
GTPγS assay. These objectives are part of our overall
goal to identify orally active and long-acting analogues
of cyclazocine having potent κ agonist and µ antagonist
binding properties useful as anti-cocaine and anti-heroin
medications. The efficient synthesis of new analogues
in the series, including many in enantiomerically pure
form, was made possible by the outbreak of new
methods for the Pd-catalyzed amination of aryl tri-
flates.22 New target compounds made for this study were
benzomorphans having secondary and tertiary amine
groups at the 8-position. Secondary amines (e.g., 8-NHR)
predominated the target list because of the long-
standing doctrine that benzomorphans as well as other
µ opioid receptor interactive agents (e.g., morphine)
require H-bond donation at that site provided by the
prototypic phenolic OH.2-5 Having the 8-NHR substitu-
tion not only provided this important H-bond donor but
also allowed us to probe previously unexplored receptor
space by systematic variation of the R group. This is
made possible by the increased valence of nitrogen
compared to the valence of the oxygen of prototypic
opiates. Besides using R ) alkyl or aryl groups to
identify potential hydrophobic sites within this space,
we also used heteroaryl groups to probe potential
H-bond donor/acceptor sites. Substitution of the 8-OH
in other 2,6-methano-3-benzazocine core structures,
to (()-4 in 84% yield by treatment with NH
2OH‚HCl/
NaOAc in methanol. In similar fashion (Schemes 1 and
2), ethylketocyclazocine (EKC, (()-41),35-37 EKC enan-
tiomers (-)-4236,37 and (+)-43,36,37 ketocyclazocine (()-
48,35 and Mr2034 [(-)-50]38 were converted to their
corresponding novel 8-NH2 derivatives (()-44, (-)-45,
(+)-46, (()-49, and (-)-51 in combined (for the three
steps) yields ranging from 35% to 73% (see Table 1).
As shown in Schemes 3 and 4, new secondary and
tertiary 8-amino derivatives of numerous core benzo-
morphan structures were made using Pd-catalyzed
aminations of aryl triflates. Two known Pd source/ligand
combinations were utilized. Method B (RR′NH, Pd2-
(dba)3, DPPF, NaO-t-Bu, toluene, 80 °C)32 was es-
sentially the same as that used to convert (()-53 to (()-
38, and method C (RR′NH, Pd(OAc)2, BINAP, NaO-t-
Bu, toluene, 80 °C)22 used a different catalyst source/

840 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5
Wentland et al.
Sch em e 1. Preparation of Primary Amino Targets Using Method Aa
a
(i) (CF3SO2)2O, pyr, CH2Cl2; (ii) (Ph)2CdNH, Pd2(dba)3, DPPF, NaO-t-Bu, tol, 80 °C; (iii) NH2OH, NaOAc, CH3OH or 2 N HCl, THF.
Sch em e 2. Preparation of 8-Amino Analogues of Mr2034a
a
(i) (CF3SO2)2O, pyr, CH2Cl2; (ii) RR′NH, Pd2(dba)3, DPPF, NaO-t-Bu, tol, 80 °C; (iii) NH2OH, NaOAc, CH3OH.
Sch em e 3. Preparation of Aminobenzomorphan Targets Using Methods B and Ca
a
(i) RR′NH, Pd2(dba)3, DPPF, NaO-t-Bu, tol, 80 °C (method B); (ii) RR′′NH, Pd(OAc)2, BINAP, NaO-t-Bu, tol 80 °C (method C); (iii) H2,
Pd(OH)2/C, CH3OH.
Sch em e 4. Preparation of Aminobenzomorphan Targets
52, respectively. By use of method C, (()-53 was
converted to (()-25, (()-27-30, and (()-33-37 in yields
of 25-84%. Also, by use of method C, the triflate of (-)-2
was converted to (-)-26 and (-)-31 in 53% and 57%
yield, respectively, and the triflate of (+)-3 was con-
verted to (+)-32 in 52% yield. Amination conditions were
generally not optimized, and in several instances where
lower yields of amination products were seen, the major
byproduct observed was the corresponding phenol,
which we assume was formed by cleavage of the triflate
by some as yet unidentified nucleophilic species.
Using Methods Da
By use of the same Pd source/ligand as method B but
reversing the role of the aromatic amine and aromatic
halide, an alternative strategy (method D, Scheme 4)
was used to make the 8-N-(hetero)aryl targets (()-20-
24. Compounds (()-20 and (()-21 were made by treat-
ing (()-4 with 4-CF3C6H4Br (57%) and 4-(CH3)2NC6H4-
Br (68%), respectively. Pd-catalyzed pyridinylations of
amines using 2-chloropyridine or 2, 3-, or 4-bromopyri-
dine have been reported.39 For our purposes, we used
2-chloropyridine and 3- and 4-bromopyridine under
conditions of method D to convert (()-4 to targets (()-
22-24, respectively, in yields of 25-89%.
a
(i) ArBr or 2-Cl-pyr, Pd2(dba)3, DPPF, NaO-t-Bu, tol, 80 °C.
ligand. In general, the two methods could be interchanged
without significant compromise in yield and we often
used/tried both to make a particular analogue.
As summarized in Scheme 3 and Table 1, triflate ester
(()-53 of cyclazocine was converted to (()-10 and (()-
13-19 using method B in yields of 37-64%. In yields of
29-59% and using method B, the triflates of (-)-2 and
(+)-3 were converted to (-)-11 and (+)-12, respectively,
and (()-54 and (-)-62 were converted to (()-47 and (-)-
Rather than use our original Birch reduction method
to remake the known 8-methylamino analogue (()-7,11

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 841
Ta ble 1. Opioid Receptor Binding Data for Aminobenzomorphans
Ki ( SE,a nM
yield,
mp,
°C
[3H]DAMGO [3H]naltrindole [3H]U69,593
b
compd
(()-1d
X
methodb
%
formula
(µ)
(δ)
(κ)
κ:µ:δc
OH (cyclazocine)
OH
0.32 ( 0.02 1.1 ( 0.04
0.10 ( 0.03 0.58 ( 0.06
0.18 ( 0.02
6:3:1
(-)-2d,e
(+)-3d,e
(()-4f,g
(-)-5a,i,j
(+)-6a,i,j
(()-7f,g
(()-8f
0.052 ( 0.009 11:6:1
OH
NH2
NH2
NH2
NHCH3
NHCH2CH3
NHCH(CH3)2
NHC6H5
360 ( 16
9.5 ( 0.14
1.8 ( 0.12
780 ( 78
13 ( 0.77
5.1 ( 0.74
240 ( 3.7
1100 ( 63
110 ( 15
12 ( 2.3
920 ( 27
470 ( 62
19 ( 3.1
460 ( 49
76 ( 8.2
14:3:1
27:12:1
10:7:1
2:1:1
55:36:1
6:4:1
A
70h
oil
oil
4.1 ( 0.11
1.2 ( 0.13
506 ( 37
8.5 ( 0.44
3.3 ( 0.10
78 ( 16
C18H26N2‚0.25H2O
C18H26N2
(()-9a,j
(()-10j
C21H32N2
6:2:1
B
B
B
B
B
B
B
B
B
B
D
D
D
D
D
C
C
C
C
C
C
C
C
C
C
C
C
C
59
57
59
49
39
54
64
63
62
41
57
68
44
89
25
66
53
84
51
55
45
57
52
54
48
49
25
55
C24H30N2‚0.3H2O
C24H30N2‚0.3H2O
C24H30N2‚1.5H2On
1.3 ( 0.072 9.4 ( 0.74
0.83 ( 0.036 11:7:1
(-)-11j-l NHC6H5
foam
oil
1.1 ( 0.08
46 ( 4.4
3.0 ( 0.12
8.4 ( 0.86
0.91 ( 0.12 6.2 ( 0.79
0.47 ( 0.077 8.3 ( 1.0
5.2 ( 0.08
270 ( 49
26 ( 2.0
29 ( 3.5
0.54 ( 0.01
27 ( 1.5
2.5 ( 0.14
5.7 ( 0.46
0.57 ( 0.14
10:5:1
10:6:1
10:9:1
5:3:1
(+)-12j,l,m NHC6H5
(()-13
(()-14
(()-15
(()-16a
(()-17
(()-18
(()-19
(()-20
(()-21a
(()-22
(()-23
(()-24
(()-25j
NH(4-ClC6H4)
175-178 C24H29N2Clo
NH(3,4-Cl2C6H3)
NH(4-CH3C6H4)
NH(4-CH3OC6H4)
NH(3-CH3OC6H4)
NH(2-CH3OC6H4)
NH(4-t-BuC6H4)
NH(4-CF3C6H4)
NH(4-NMe2C6H4)
NH(2-pyridinyl)
NH(3-pyridinyl)
NH(4-pyridinyl)
NHCH2C6H5
oil
C24H28N2Cl2
170-172 C25H32N2‚0.5H2O
11:7:1
180-181 C25H32N2O‚0.5H2O
0.29 ( 0.012 29:18:1
122-124 C25H32N2O‚0.25H2O 1.8 ( 0.21
13 ( 1.6
98 ( 17
61 ( 3.2
180 ( 40
75 ( 14
11 ( 4.7
1.0 ( 0.07
2.2 ( 0.10
14 ( 2.5
13:7:1
45:12:1
4:3:1
16:14:1
9:1:1
oil
oil
oil
oil
C25H32N2O‚0.75H2O 8.5 ( 1.7
C28H38N2‚H2O 18 ( 0.90
C25H29N2F3‚0.25H2O 13 ( 0.96
C26H35N3‚0.5H2O 70 ( 3.5
23H29N3‚0.125H2O 4.2 ( 0.70
11 ( 2.3
8.8 ( 0.36
9.2 ( 1.2
0.50 ( 0.061 42:33:1
2.9 ( 0.073
4.7 ( 0.15
2.1( 0.10
71 ( 13
foam
oil
C
C
C
1:3:1
23H29N3‚0.125H2O 0.64 ( 0.054 21 ( 1.8
oil
oil
oil
23H29N3‚0.5H2O
18 ( 1.4
1.5 ( 0.16
240 ( 20
57 ( 2.5
83:13:1
12:38:1
26:81:1
1:10:1
14:11:1
15:18:1
46:18:1
45:18:1
3:2:1
4:2:1
12:4:1
5:4:1
28:12:1
13:10:1
1:1:1
36:14:1
18:8:1
5:4:1
(-)-26a,j NHCH2C6H5
C25H32N2
0.67 ( 0.043 54 ( 3.6
(()-27
(()-28
(()-29j
(()-30j
NHCH2(4-CH3OC6H4)
136-138 C26H34N2O
7.8 ( 1.6
C25H31N2Cl‚0.25H2O 8.8 ( 1.3
9.0 ( 2.4
82 ( 1.9
C20H30N2‚0.125H2O 46 ( 9.2
80 ( 11
NHCH2(4-ClC6H4)
NHCH2CH2C6H5
N(CH3)2
oil
oil
oil
oil
oil
oil
100 ( 16
160 ( 8.0
1500 ( 96
810 ( 90
7.2 ( 1.7
11 ( 0.084
33 ( 2.5
C26H34N2‚0.25H2O
(-)-31p,q N(CH3)2
18 ( 1.4
(+)-32q,r N(CH3)2
C20H30N2
C22H32N2
5000 ( 310 8400 ( 380
3000 ( 27
410 ( 30
300 ( 29
560 ( 82
8.8 ( 0.40
44 ( 1.8
(()-33j
(()-34j
(()-35j
(()-36j
(()-37j
(()-38a
(()-39a,s
(()-40a
(()-41t
1-pyrrolidinyl
1-piperidinyl
770 ( 55
950 ( 84
650 ( 38
21 ( 2.8
56 ( 4.1
360 ( 29
18 ( 1.9
22 ( 0.057
1800 ( 73
3600 ( 390
2800 ( 200
250 ( 51
570 ( 33
220 ( 46
250 ( 14
170 ( 7.7
264 (dec) C23H34N2‚2HCl
140-142 C22H32N2O
1-morpholinyl
1-(4-methypiperazinyl)
N(CH3)CH2C6H5
NdC(Ph)2
oil
oil
C23H35N3
C26H34N2‚0.25H2O
C31H34N2
170 ( 5.5
7.0 ( 0.22
9.6 ( 0.45
0.62 ( 0.11
222-224 C24H28N2‚0.5H2O
X ) OH, R6 ) CH2CH3 (EKC)
0.78 ( 0.10 3.4 ( 0.41
0.73 ( 0.12 1.1 ( 0.24
(-)-42t,u X ) OH, R6 ) CH2CH3
0.17 ( 0.016 6:2:1
(+)-43t,u X ) OH, R6 ) CH2CH3
320 ( 67
7.3 ( 1.6
2.7 ( 0.40
17 ( 0.70
10 ( 2.0
3.3 ( 0.66
3.6 ( 0.51
1300 ( 150
51 ( 4.9
7.6 ( 1.4
81 ( 7.6
37 ( 3.0
20 ( 2.7
29 ( 5.3
73 ( 13 18:4:1
(()-44
X ) NH2, R6 ) CH2CH3
A
A
A
B
37h
35h
37h
51
oil
C19H26N2O‚0.8H2O
0.34 ( 0.029 150:7:1
(-)-45v,w X ) NH2, R6 ) CH2CH3
200-203 C19H26N2O‚H2O
198-200 C19H26N2O‚0.9H2O
1.1 ( 0.19
9.1 ( 0.54
2.2 ( 0.12
1.0 ( 0.24
0.80 ( 0.04
7:3:1
9:5:1
17:4:1
20:6:1
36:8:1
(+)-46w,x X ) NH2, R6 ) CH2CH3
(()-47
(()-48t
(()-49
(-)-50y
(-)-51z
X ) NHC6H5, R6 ) CH2CH3
X ) OH, R6 ) CH3
X ) NH2, R6 ) CH3
OH (Mr2034)
oil
C25H30N2O‚0.5H2O
A
73h
190-192 C18H24N2O
0.17 ( 0.02 2.5 ( 0.41
0.073 ( 0.004 34:15:1
7.8 ( 1.7
2.0 ( 0.12
NH2
A
B
65h
29
oil
C19H28N2O
14 ( 0.97
370 ( 48
35 ( 3.4
47:26:1
18:12:1
(-)-52aa NHC6H5
177-178 C25H32N2O‚0.25H2O 2.8 ( 0.35
a
b
d
See Experimental Section. Refers to the Pd-catalyzed amination step except as noted. c Ratio of the corresponding Ki values. Known
compound. See ref 28. e Eudismic ratios for µ, δ, and κ are 3600, 1897 and 1462, respectively. f Known compound. See ref 11. See
g
h
i
Experimental Section for an alternative route. Combined yield for the three-step sequence. Eudismic ratios for µ, δ, and κ are 433, 77
and 422, respectively. Reported in the preliminary account of this work (ref 20). [R]25 -79.3° (c 1.0, CHCl3). Eudismic ratios for µ, δ,
j
k
l
D
and κ are 42, 52, and 50, respectively. [R]25 +78.3° (c 1.0, CHCl3). H: calcd, 8.92; found, 8.25. N: calcd, 7.49; found, 6.66. o N: calcd,
m
n
D
7.36; found, 6.64. [R]25 -113.6° (c 0.88, EtOH). Eudismic ratios for µ, δ, and κ are 107, 10, and 167, respectively. [R]25 +125.4° (c
p
q
r
D
D
s
t
u
1.18, EtOH). Known compound. See ref 40. Known compound. See refs 35-37. Eudismic ratios for µ, δ, and κ are 438, 1182, and 429,
respectively. v [R]25 -90.4° (c 1.0, 2% aqueous HOAc). Eudismic ratios for µ, δ, and κ are 6, 11, and 8, respectively. x [R]25 +88.3° (c
w
D
D
1.0, 2% aqueous HOAc). y Known compound. See ref 38. z [R]25 -90.4° (c 1.10, MeOH). [R]25 -97.6° (c 0.84, CHCl3).
aa
D
D
we attempted the two Pd-catalyzed procedures, methods
B and C, without success. It is likely that the volatility
of methylamine reduced its effective concentration to
levels too low for success under these conditions; using

842 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5
Sch em e 5. Preparation of Pyrrolo-Fused Analogue (()-39a
Wentland et al.
a
(i) BrCH2CH(OCH2CH3)2, KHCO3, DMF; (ii) (CF3CO)2O, CF3CO2H; (iii) KOH, MeOH.
Sch em e 6. Preparation of Indolo-Fused Analogue (()-40a
a
(i) C6H5NHNH2‚HCl, HOAc.
excess gaseous methylamine did not help. Since we had
a supply of the N-methyl-N-phenylmethyl analogue (()-
37, we removed the benzyl group with H2/Pd(OH)2/C
(Scheme 3) to give (()-7 in 44% yield. The isopropy-
lamino analogue (()-9 was made in 71% yield through
reductive amination of (()-4 with (CH3)2CO, NaBH3CN,
and ZnCl2.
oxygen present in the reaction mixture. Rather than go
through elaborate experiments to isolate and character-
ize (()-68, we simply carried out the reaction as before
and obtained (()-40 in 21% yield. Cyclization of the
enamine intermediate from (()-66 with phenylhydra-
zine could have occurred at either the 7- or 9-position,
giving two regioisomeric (dihydro)carbazoles. The struc-
ture of the product was assigned (()-40 by proton NMR
where the two benzenoid H atoms that are part of the
benzomorphan ring system are observed as two doublets
coupled to each other at δ 7.15 ppm (d, J ) 8.3 Hz, 1H)
and 7.21 ppm (d, J ) 8.3 Hz, 1H). Thus, we conclude
that the regiochemistry of the Fischer indole cyclization
occurred in the predicted orientation where the putative
enamine intermediate (()-67 giving rise to the observed
product would be stabilized through conjugation relative
to the other possible enamine structure.
In an attempt to gain insight into the nature of the
bioactive conformation of these ligands, two cyclic
probes, indole derivative (()-39 and carbazole analogue
(()-40, were made. The synthesis (and in vivo antinoci-
ception properties) of indole (()-39 has been reported
from an 8-amino-3-carboethoxy-2,6-methano-3-benza-
zocine intermediate.40 We felt the harshly acidic condi-
tions of this procedure could not be used to convert (()-4
directly to (()-39 because the cyclopropyl group might
not survive. We utilized an alternative procedure41
shown in Scheme 5 and described in the Experimental
Section. Compound (()-4 was treated with bromo-
acetaldehyde diethyl acetal, KHCO3, and DMF to pro-
vide intermediate (()-64. Cyclization of this compound
with (CF3CO)2O and CF3CO2H occurred at the 9-posi-
tion of the benzene ring to give trifluoroacetamide (()-
65, which was hydrolyzed in KOH/CH3OH to provide
the desired target (()-39. Assignment of structure, and
in particular where the point of cyclization occurred (9-
versus 7-position of (()-64), was made using proton
NMR where the two benzenoid H atoms appeared as
sharp singlets at δ 7.30 and 7.33 ppm.
Resu lts a n d Discu ssion
Affinities of 8-amino-2,6-methano-3-benzazocine tar-
gets and their 8-OH counterparts for µ, δ, and κ opioid
receptors are reported in Table 1 and were assessed
using published procedures.43 The first part of this
discussion involves the bioisosteric replacement of NH2
for the 8-OH group on the following core structures:
cyclazocine and both enantiomers, EKC and both enan-
tiomers, ketocyclazocine, and Mr2034 [(-)-50], an opti-
cally active 3-tetrahydrofuranylmethyl-2,6-methano-3-
benzazocine. Compared to cyclazocine (()-1, the primary
amino analogue (()-4 displayed 30-fold and 23-fold
lower affinity for µ and κ receptors, respectively. This
significantly lower affinity is not consistent with results
from our earlier in vivo studies6 where (()-4 was only
5-fold less potent than (()-1 in rodent models of anti-
nociception dosed sc and equipotent po. We assume,
therefore, that (()-4 has substantially enhanced phar-
macokinetic properties relative to cyclazocine especially
when administered orally. Against δ, (()-4 had very low
By use of a known procedure,42 carbazole target (()-
40 was made from an enol ether intermediate [(()-66]
we previously reported.11 Treatment of (()-66 with
phenylhydrazine hydrochloride and acetic acid gave only
compound (()-40 (Scheme 6) and no expected dihydro-
carbazole intermediate (()-68 was observed in our first
attempts. It is likely that intermediate (()-68 was
formed, but despite the fact that it was performed under
a blanket of N2, it was oxidized by small amounts of

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 843
affinity compared to µ and κ receptors. For both cycla-
zocine and the primary amino analogue, the active
enantiomers at opioid receptors are the (-)-(2R,6R,11R)-
isomers (-)-2 and (-)-5, respectively; eudismic ratios
versus µ were 3600 and 433, respectively, and ratios
versus κ were 1462 and 423, respectively.
respectively, than the primary amine comparator (()-
4. In fact, the affinity of (()-10 approached within 4-fold
of the subnanomolar affinity of cyclazocine [(()-1]. As
we observed in several other examples, the enantiomer
of (()-10 that displays the higher affinity is the (-)-
(2R,6R,11R)-isomer (-)-11; however, large eudismic
ratios were not seen (42, 52, and 50 versus µ, δ, and κ,
respectively). We have no explanation at present to
account for this.
Ethylketocyclazocine [EKC, (()-41] is a well-docu-
mented benzomorphan that has high affinity for opioid
receptors. EKC acts as a κ agonist and, at higher doses,
a µ agonist in monkeys.44,45 Additionally, EKC has been
shown to block cocaine self-administration in nonhuman
primates.46,47 Upon a comparison of the novel 8-NH2
derivative (()-44 to (()-41, higher affinity than would
be predicted by the cyclazocine case was seen against
all receptor types. The amino analogue (()-44 had 9-fold
and 15-fold lower affinity than EKC (()-41 for µ and δ,
respectively; against κ, comparable affinity was ob-
served. As expected, the active enantiomers at opioid
receptors for both EKC and the amino derivative are
(-)-(2R,6R,11R)-isomers (-)-42 and (-)-45, respectively;
however, a rather large divergence in eudismic ratios
was observed between the two. For the EKC case,
eudismic ratios for the two enantiomers, (-)-42, and (+)-
43 versus µ, δ, and κ were 438, 1182, and 429, respec-
tively. For the 8-aminoEKC example, the eudismic
ratios for the two enantiomers, (-)-45, and (+)-46 versus
µ, δ, and κ were only 6, 11, and 8, respectively. When
an unexpectedly low eudismic ratio is observed, one
suspects that the “inactive” enantiomer, in this case (+)-
46, is contaminated with the “active” isomer [i.e., (-)-
45]. In our case, however, the amino enantiomers were
made from the very same samples of the 8-OH enanti-
omers [(-)-42 and (+)-43] that were put into the
receptor binding assays and had large ratios. At present,
we have no explanation for the low eudismic ratios of
(-)-45 and (+)-46.
From the data just described, it appears that the
presence of a 1-keto group is beneficial for the 8-OH f
8-NH2 bioisosteric replacement strategy. Further sup-
port for this conclusion comes from study of another
1-keto core system, ketocyclazocine (()-48, where nearly
identical affinities against all receptor types were seen
for the 8-OH [(()-48] and 8-NH2 [(()-49] analogues.
We studied replacement of the 8-OH of Mr2034 [(-)-
50].38 This compound, and related derivatives, was
pioneered by Merz and co-workers and is characterized
by having very high potency in mouse antinociception
models.38 When we replaced the 8-OH of (-)-50 with
NH2 to provide (-)-51, we found that (-)-51 displayed
considerably less affinity than (-)-50 for all receptor
types. Specifically, Ki values were 82-fold, 148-fold, and
107-fold greater for (-)-51 than (-)-50 against µ, δ, and
κ, respectively.
To determine if the substantial benefit of adding a
phenyl group to the 8-N of (()-4 was transferable to
other core structures, we evaluated the corresponding
N-phenyl analogues in the EKC and Mr2034 core
structures. Compared to 8-aminoEKC [(()-44], the
N-phenyl analogue (()-47 had nearly the same affinity
for µ and δ and displayed 6-fold less affinity for κ
receptors. For the Mr2034 core structure, addition of a
phenyl group to the primary amine of (-)-51 resulted
in a compound [(-)-52] having 5-fold, 11-fold, and 4-fold
higher affinity for µ, δ, and κ, respectively. It appears
that the benefits of the N-phenyl group vary depending
on the core benzomorphan structure.
We made a considerable effort to understand what
role the phenyl group of (()-10 has in enhancing the
binding affinity of (()-4. We implemented a Topliss
method approach48 to probe what effect, if any, various
physicochemical parameters of substituents on the
phenyl ring have on binding affinity. The standard five
compounds that are part of the first phase of the Topliss
method were made. Review of the binding data for (()-
10 (8-NHC6H5), (()-13 [NH(4-ClC6H4)], (()-14 [NH(3,4-
Cl2C6H3)], (()-15 [NH(4-CH3C6H4)], and (()-16 [NH(4-
CH3OC6H4)] revealed a rank order (from highest to
lowest) of affinity for µ and κ of (()-16 > (()-15 g (()-
10 > (()-13 > (()-14. Topliss scheme analyses of this
rank order suggested that -σ (i.e., electron donation)
was an important determinant of activity. Of the several
analogues suggested for additional synthesis by the
Topliss algorithm, we choose the NH(4-NMe2C6H4)
derivative (()-21. This compound, however, had lower
affinity (8- to 149-fold less affinity for µ and 2- to 30-
fold less affinity for κ receptors) than the five probes,
leading to the likely conclusion that the larger (than
methyl or methoxy) 4-dimethylamino group has a nega-
tive steric interaction with the two receptors. In an
attempt to support or refute this conclusion, we made
(()-19, which has the bulky electron-donating group
tert-butyl at the 4-position of the N-phenyl ring. Since
(()-19 had affinity for µ and κ that is between that of
(()-15 (4-CH3) and (()-21 (4-NMe2), our understanding
of the electronic and/or steric effects of the 4-substituent
was not clarified.
We also made the following three additional ana-
logues in this subseries: 3-CH3O [(()-17], 2-CH3O [(()-
18], and 4-CF3 [(()-20]. For the two methoxy regioiso-
mers (()-17 and (()-18, binding affinity, while sub-
stantial, was somewhat lower (4-fold and 18-fold for µ,
respectively; 2-fold and 8-fold for κ, respectively) than
that of (()-16; the 4-CH3O derivative has the highest
binding affinity within the entire series. These data
suggest that not only the presence of an electron-
donating group is important but also affinity is depend-
ent on that group’s position on the phenyl ring. Last,
We also studied the effect on receptor binding affinity
upon substitution of the 8-N of cyclazocine analogue (()-
4. When the primary amine of (()-4 was substituted
with monoalkyl groups, affinity against all receptor
types was not appreciably affected by methyl [(()-7] or
ethyl [(()-8] substituents; however, it was substantially
reduced (e.g., 25-fold for µ) by isopropyl substitution
[(()-9]. For secondary amine substituents containing a
phenyl ring, however, substantial affinity was observed.
Against µ and κ, for example, the phenylamino analogue
(()-10 displayed 7-fold and 5-fold greater affinity,

844 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5
Wentland et al.
the 4-CF3 phenyl derivative (()-20 displayed somewhat
less affinity than the 4-Cl [(()-13] and 3,4-Cl2 [(()-14]
analogues.
subseries paralleled that of the N-phenyl series. Com-
pound (()-27, the 4-methoxybenzyl derivative, had 5-
and 15-fold less affinity for µ and κ, respectively, than
(()-25. This result contrasts with the N-phenyl case
where 4-methoxy [(()-16] resulted in a 3-fold increase
in affinity relative to H [(()-10]. When a 4-Cl was
introduced into the benzyl of (()-27, affinity decreased
6-fold for µ and 3-fold for κ receptors. This was similar
to the effect the 4-Cl had on the N-phenyl analogue (()-
10.
The 2-, 3-, and 4-pyridinyl analogues of (()-10 were
made to probe the effects of aza substitution on the
phenyl ring. Subnanomolar affinity for µ and κ was
observed for the 3-pyridinyl compound (()-23 that was
comparable to (()-10. The 2-pyridinyl analogue (()-22
displayed 6-fold and 18-fold lower affinity than (()-23
for µ and κ receptors, respectively. The 4-pyridinyl
isomer (()-24 had 28-fold and 6-fold lower affinity than
(()-23 for µ and κ receptors, respectively. In an attempt
to rationalize the substantial differences in affinity
within this small subseries of structurally similar
analogues, one argument would be that the best affinity
would be seen in those analogues where the H-bond
donor ability of the anilino NH is the greatest. Since
the compounds with the highest affinity, the phenyl
([(()-10]) and 3-pyridinyl ([(()-23]) derivatives, are the
ones predicted to have the weakest H-bond donor
characteristics based on electronic consideration, other
factors must be considered. These limited data suggest
that the enhanced affinity of the phenyl ([(()-10]) and
3-pyridinyl ([(()-23]) compounds relative to the affinity
of others [e.g., 8-(2-pyridinylamino), 8-(4-pyridinylami-
no), 8-amino) is due to optimized interactions of the
(hetero)aromatic group itself with the target protein
rather than to the role the (hetero)aromatic group plays
in modulating the physicochemical properties of other
groups in the pharmacophore. One such interaction
might be hydrophobic in nature; another possibility is
that the 3-N of (()-23 acts as an H-bond acceptor for a
putative complimentary donor on the receptor.
Extension of the phenyl group by one more methylene
gave the phenethyl analogue (()-29, which had some-
what lower affinity (approximately 7-fold) for µ than
either the corresponding benzyl or phenyl derivative (()-
25 or (()-10, respectively. Against κ, (()-29 had only
2-fold lower affinity than (()-25 and had 13-fold lower
affinity than (()-10.
Several 8-tertiary amine derivatives of cyclazocine
were also made. Relative to primary amine (()-4, the
8-dimethylamino analogue (()-30 had 8- to 9-fold lower
binding affinity for µ and κ. Consistent with SAR studies
for traditional opiates where the prototypic phenolic OH
is required for binding to opioid receptors,2-5 these data
indicate that at least one H on the 8-position N is
required for reasonable binding affinity. While it is
possible that (()-30 can donate an H-bond to a compli-
mentary acceptor site on the opioid receptor when in
the protonated state, the low pKa (<5) for such aromatic
amines decreases the likelihood of any significant pro-
tonation at the pH (7.5) of the assay. These binding
data, however, appear inconsistent with our previous
in vivo antinociception studies where we observed that
(()-4 and (()-30 had the same high potency in the
acetylcholine writhing test (mouse) when dosed sc.11 It
is likely that the potent antinociception properties of
(()-30 is a result of its metabolism to (()-4 and/or (()-
7, two compounds that have good affinity for opioid
receptors. We have no direct evidence, however, to
support or refute this assumption. We made the two
enantiomers of (()-30 and found that the modest
binding affinity resides in the (-)-(2R,6R,11R)-isomer
(-)-31 [versus (+)-32].
The N-(1-pyrrolidinyl), N-(1-piperidinyl), and N-(1-
morpholinyl) analogues (()-33, -34, and -35, respec-
tively, all had very low affinity for the receptors;
however, there were two tertiary amine derivatives that
bucked this trend and displayed reasonable affinity.
These were the N-[1-(4-methylpiperazinyl)] analogue
(()-36 (Ki ) 21 nM versus µ; Ki ) 8.8 nM versus κ) and
the N(CH3)CH2C6H5 derivative (()-37 (Ki ) 56 nM
versus µ; Ki ) 44 nM versus κ). As discussed in an
earlier section, the N-benzyl group is responsible for a
6-fold increase in affinity for µ relative to NH (compare
(()-25 and (()-4) and had little effect on κ affinity. We
can only speculate that the enhanced affinity of the
tertiary amine derivative (()-37 for µ relative to other
tertiary amine analogues is to due to the benefits of the
N-benzyl appendage and may compensate somewhat for
the lack of the NH H-bond donor. For (()-36, the distal
N of the piperazine is undoubtedly protonated at pH
7.5 and thus may sustain an electrostatic or H-bond
interaction with a complementary site of the receptor.
Last, we tested imine intermediate (()-38, which dis-
played very low affinity for all receptors. Factors
In an attempt to identify the bioactive conformation
of the N-alkyl and/or aryl derivatives, we made two
conformationally restricted analogues as probes. These
are the indole derivative (()-39 and carbazole (()-40,
and we observed good to moderate binding affinity for
both. In addition to (()-39 being a potential mimic of
the N-ethyl derivative (()-8, the indole NH fits nicely
into the classic mold of a bioisosteric replacement for a
phenolic OH.49 Relative to (()-8, indole (()-39 had 4-fold
less affinity for µ and had nearly comparable affinity
for κ. Carbazole derivative (()-40 could be considered a
close analogue of the N-phenyl derivative (()-10 in that
it incorporates an ortho carbon of the phenyl appendage
of (()-10 into an additional ring with the 7-position of
the benzomorphan ring system. Modest binding affinity
was seen for (()-40 that was 17- and 12-fold less than
(()-10 for µ and κ, respectively. Few conclusions can be
drawn from these data other than the following: (a) we
have yet another example of the utility of an indole NH
to be a surrogate for a phenolic OH; (b) the bioactive
conformation of (()-10 is probably not that displayed
by (()-40.
To further probe the receptor space that the phenyl
of (()-10 occupies, we extended the distance between it
and the benzomorphan core by one and two methylenes.
Compared to (()-10, the benzylamino analogue (()-25
had comparable affinity for µ receptors but had 6-fold
lower affinity for κ. Consistent with our other observa-
tions, the activity of racemic (()-25 resides in the (-)-
(2R,6R,11R)-isomer (-)-26. We made two substituted
(on phenyl) derivatives to see if the SAR in the N-benzyl

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 845
Ta ble 2. EMax and EC50 Values for the Stimulation of
[
data. For (-)-cyclazocine [(-)-2] and the 8-NH2 analogue
(-)-5, similar maximal stimulation of [35S]GTPγS bind-
ing to κ receptors was seen. (-)-EKC [(-)-42] produced
the greatest stimulation (Emax ) 170%) of [35S]GTPγS
binding to κ receptors among all compounds tested and
was substantially more efficacious than its 8-NH2
analogue (-)-45. In each of the other two pairs of 8-OH
versus 8-NH2 analogues evaluated in the κ assay, (()-
48 versus (()-49 (ketocyclazocine core) and (-)-50
versus (-)-51 (Merz core), similar efficacy was observed.
The two pairs of 1-keto analogues [(-)-42 versus (-)-
45 and (()-48 versus (()-49] were studied for stimula-
tion of [35S]GTPγS binding to µ receptors (Table 3)
where a different activity profile was seen. Both 8-NH2
compounds showed higher efficacy (1.4- to 2-fold) than
their 8-OH counterparts. We also compared the effects
of phenyl substitution on the 8-N on [35S]GTPγS binding
by using the (-)-cyclazocine-related pair (-)-5 versus
(-)-11 for κ and using the EKC-related pair (()-44
versus (()-47 for κ and µ receptors. For both κ and µ,
the N-phenyl group was responsible for a modest
increase in efficacy (1.1- to 1.5-fold).
35S]GTPγS Binding to the Human κ Opioid Receptora
compd
Emax (% maximal stimulation)
EC50 (nM)
(()-1
90 ( 7.5
100 ( 2.6
80 ( 6.5
120 ( 7.7
90 ( 8.8
170 ( 11
100 ( 6.0
110 ( 3.2
110 ( 8.6
80 ( 1.2
70 ( 6.0
70 ( 9.2
70 ( 5.7
140 ( 12
120 ( 14
1.6 ( 0.60
1.2 ( 0.30
20 ( 0.12
3.8 ( 0.21
10 ( 0.94
1.1 ( 0.30
4.4 ( 1.6
2.8 ( 0.80
5.9 ( 1.0
3.3 ( 0.17
9.7 ( 2.1
4.0 ( 0.38
330 ( 46
7.0 ( 4.3
18 ( 4.3
(-)-2
(-)-5
(-)-11
(()-23
(-)-42
(()-44
(-)-45
(()-47
(()-48
(()-49
(-)-50
(-)-51
U 50,488
spiradoline
a
See Experimental Section. Data are the mean Emax and EC50
values ( SE from at least three separate experiments, performed
in triplicate. For calculation of the Emax values, the control
[
35S]GTPγS binding was set at 0%.
Ta ble 3. Emax and EC50 Values for the Stimulation of
35S]GTPγS Binding to the Human µ Opioid Receptora
[
compd
Emax (% maximal stimulation)
EC50 (nM)
Con clu sion s
(-)-2
50 ( 6.0
50 ( 5.6
70 ( 2.8
60 ( 5.3
100 ( 1.7
70 ( 1.3
40 ( 0.21
80 ( 5.5
120 ( 12
4.5 ( 0.50
15 ( 4.1
5.5 ( 1.1
62 ( 3.6
92 ( 6.3
22 ( 3.1
21 ( 3.2
180 ( 12
110 ( 9.0
Analysis of opioid receptor binding data for this
expanded series of 8-amino-2,6-methano-3-benzazocines
has led us to an improved understanding of the rela-
tionship between affinity for the receptors and the
nature of the 8-substituent. With respect to the isosteric
replacement of the 8-OH of several benzomorphans with
8-NH2, the effect of the amino group on binding affinity
is dependent on the nature of the ligand core and that
the 8-NH2 group most closely mimics the binding
properties of 8-OH when the core structure contains a
1-keto group (i.e, EKC and ketocyclazocine) and the
receptor type is κ. This conclusion about the interde-
pendence of functional groups on binding affinity is
further supported by data from our morphine study
where we found that 3-amino-3-deshydroxymorphine
had 60-fold lower affinity for µ than morphine.24
The significant enhancement in affinity upon mono-
substitution of the 8-N with (hetero)aryl-containing
groups suggests that the receptor has a specific comple-
mentary site(s) of interaction with these appendages.
On the basis of the SAR that has emerged, we conclude
that hydrophobics play an important role in binding and
that the hydrophobic pocket may contain an H-bond
donor group complementary to a heteroaryl (e.g., 3-py-
ridinyl) acceptor. This binding pocket has not been
previously explored to any great extent because most
opiates have no opportunity to probe this receptor space
because of the lower (than nitrogen) valence of the
oxygen of the prototypic 8-OH group. Disubstitution of
the 8-amine provides ligands having relatively low
affinity that we conclude is due to the absence of an NH
donor as would be predicted from historical SAR data
of 8-OH opiates.
(-)-11
(-)-42
(()-44
(-)-45
(()-47
(()-48
(()-49
DAMGO
a
See Experimental Section. Data are the mean Emax and EC50
values ( SE from at least three separate experiments, performed
in triplicate. For calculation of the Emax values, the control
[
35S]GTPγS binding was set at 0%.
contributing to this observation could be the lack of an
NH as H-bond donor and/or that the steric bulk of the
benzophenone imine moiety might exceed the size of the
complimentary pocket(s) in the receptors.
Intrinsic opioid-receptor-mediated activity for selected
members of the series was determined using [35S]GTPγS
binding assays at κ and µ receptors. All compounds
tested in these assays had high affinities (Ki < 4 nM)
for κ and receptors and several pairs were chosen to
compare the effect of the following structural changes
on intrinsic activity: (a) 8-OH versus 8-NH2 [(-)-2 and
(-)-5, (-)-42 and (-)-45, (()-48 and (()-49, and (-)-50
and (-)-51]; (b) 8-NH2 versus 8-NHphenyl [(-)-5 and
(-)-11, and (()-44 and (()-47]. Results for κ and µ opioid
receptor stimulation of [35S]GTPγS binding are shown
in Tables 2 and 3, respectively. Efficacy in these assays
is quantified by the Emax value (the percent of maximal
stimulation of control), and affinity for the receptor is
quantified by an EC50 value that represents the con-
centration of compound that produced 50% of the
maximal stimulation obtained by the compound. Intrin-
sic activity was observed in all compounds tested. For
example, the 8-NH(3-pyridinyl) analogue (()-23, that
has high affinity (Ki ) 0.50 nM) for κ in the binding
assay, showed an Emax of 90% (Table 2) comparable to
the value displayed by cyclazocine (()-1. Within this
small subset of compounds, when tested for both κ and
µ receptor stimulation, these benzomorphans were
generally more efficacious in stimulating κ than µ, as
would be predicted from the binding data and historical
Besides the high affinity for opioid receptors, other
attributes of 8-OH-containing benzomorphans that were
sustained by many 8-amino substitutions are (a) high
enantiospecificity of binding, with the (2R,6R,11R)-
isomers [(2S,6R,11R)- for the 1-keto compounds] being
the active enantiomer; (b) high selectivity for µ and κ
over δ opioid receptors, and (c) intrinsic activity as
demonstrated in [35S]GTPγS assays.

846 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5
Wentland et al.
organic phase was washed twice with water and once with
brine. After the organic extracts (Na2SO4) were dried and
concentrated in vacuo, (()-53 (1.42 g, 96%) was obtained as a
red-orange oil: 1H NMR (500 MHz, CDCl3) δ 7.15 (d, 1H, J )
8.5 Hz), 7.12 (d, 1H, J ) 2.4 Hz), 7.00 (dd, 1H, J ) 8.5, 2.6
Hz), 3.19 (m, 1H), 2.98-2.94 (d, 1H, J ) 18.6 Hz), 2.67-2.77
(m, 2H), 2.52 (m, 1H), 2.34 (m, 1H), 1.99 (m, 3H), 1.38 (s, 3H),
1.33 (m, 1H), 0.89 (m, 1H), 0.85 (d, 3H, J ) 7.1 Hz), 0.51 (d,
2H, J ) 8 Hz), 0.12 (m, 2H); IR (film) νmax 2965, 1716, 1418,
1215 cm-1; MS (CI) m/z 404 (M + H)+. Anal. (C19H24F3NO3S‚
0.5H2O) C, H, N.
We believe the knowledge gained from this study will
facilitate the design of new high-affinity opioid receptor-
interactive ligands modified at the 8-position. In fact,
using the knowledge we learned that the receptor can
accommodate large hydrophobic 8-substituents capable
of sustaining H-bond contacts with the protein, we
recently made and evaluated several opiates containing
a carboxamide (CONH2) group at the phenolic OH
position.50,51 Many of these novel ligands display the
high (subnanomolar) affinity for µ, κ, and/or δ as their
phenolic OH counterparts, and in one instance (cycla-
zocine core), this modification resulted in a substantially
longer (2 versus 15 h) duration of antinociception in
mice.52 Encouraged by these results, we continue to
focus our efforts in further exploration of 8- and 3-posi-
tion modification of benzomorphans and morphinans,
respectively.
(()-3-(Cyclop r op ylm et h yl)-N-(d ip h en ylm et h ylen e)-
1,2,3,4,5,6-h exa h yd r o-cis-6,11-d im et h yl-2,6-m et h a n o-3-
ben za zocin -8-a m in e [(()-38]. Meth od A. An oven-dried 25
mL two-neck flask equipped with reflux condenser was placed
into a N2-filled glovebox where it was charged with Pd2(dba)3
(0.052 g, 0.057 mmol), DPPF (0.094 g, 0.171 mmol), and NaO-
t-Bu (0.109 g, 1.14 mmol). The system was capped with a
rubber septum and removed from the glovebox. Dry toluene
(4 mL) was then added to the mixture via syringe, and the
resulting dark suspension was stirred at 25 °C for 10 min. A
solution containing triflate (()-53 (0.460 g, 1.14 mmol) and
benzophenone imine (0.248 g, 1.368 mmol) in 3 mL of toluene
was added, and the mixture was heated to 80°C with vigorous
stirring. At approximately 6 h, the reaction had gone to
completion as evidenced by TLC. It was allowed to cool, diluted
with 20 mL of CH2Cl2, filtered over Celite, and then directly
preadsorbed onto silica gel. The crude reaction mixture was
purified by flash column chromatography using hexanes/ethyl
acetate as eluent (20:1 f 10:1 f 3:1 gradient) to yield (()-38
(0.245 g, 47%) as a yellow-orange oil: 1H NMR (500 MHz,
CDCl3) δ 7.75 (d, 2H, J ) 7.5 Hz), 7.46 (m, 1H), 7.40 (m, 2H),
7.23 (m, 3H), 7.08 (m, 2H), 6.87 (d, 1H, J ) 8.1 Hz), 6.63 (m,
1H), 6.44 (m, 1H), 3.08 (m, 1H), 2.81 (d, 1H, J ) 18.2 Hz),
2.53-2.63 (m, 2H), 2.44 (m, 1H), 2.30 (m, 1H), 1.67-1.90 (m,
3H), 1.07 (s, 3H), 0.97 (m, 1H), 0.85 (m, 1H), 0.75 (d, 3H, J )
6.8 Hz), 0.50 (m, 2H), 0.10 (m, 2H); IR (film) νmax 2961, 2916,
1594, 1571, 1491 cm-1; MS (CI) m/z 435 (M + H)+. Anal.
(C31H34N2 ) C, H, N.
Exp er im en ta l Section
Proton NMR and in certain cases 13C NMR [Varian Unity-
500 NMR] data, direct insertion probe (DIP) chemical ioniza-
tion mass spectra (Shimadzu GC-17A GC-MS mass spectrom-
eter), and infrared spectra (Perkin-Elmer Paragon 1000 FT-
IR spectrophotometer) were consistent with the assigned
1
structures of all test compounds and intermediates. H NMR
multiplicity data are denoted by s (singlet), d (doublet), t
(triplet), q (quartet), m (multiplet), dd (doublet of doublets),
and br (broad). Coupling constants are in hertz. Carbon,
hydrogen, and nitrogen elemental analyses for all novel targets
were performed by Quantitative Technologies Inc., White-
house, NJ , and were within ( 0.4% of theoretical values except
as noted; the presence of water was confirmed by proton NMR.
Melting points were determined on
a Meltemp capillary
melting point apparatus and are uncorrected. Optical rotation
data were obtained from a Perkin-Elmer 241 polarimeter.
Reactions were generally performed under an argon or N2
atmosphere. Amines used in the Pd-catalyzed amination
reactions and racemic 2,2′-bis(diphenylphosphino)-1,1′-binaph-
thyl (BINAP) were purchased from Aldrich Chemical Co. and
used as received unless otherwise indicated. Tris(dibenzylide-
neacetone)dipalladium(0) [Pd2(dba)3], Pd(OAc)2, and 1,1′-bis-
(diphenylphosphino)ferrocene (DPPF) were purchased from
Strem Chemicals, Incorporated. THF was distilled from sodium/
benzophenone ketyl. Toluene and Et2O were distilled from
sodium metal. Pyridine was distilled from KOH. Methylene
chloride was distilled from CaH2. DMF and DMSO were
distilled from CaH2 under reduced pressure. Methanol was
dried over 3 Å molecular sieves prior to use. Silica gel (Bodman
Industries, ICN SiliTech 2-63 D 60A, 230-400 mesh) was used
for flash column chromatography.
(()-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-2,6-m eth a n o-3-ben za zocin -8-a m in e [(()-4].
Meth od A. Ketimine adduct (()-38 (1.459 g, 3.36 mmol) was
dissolved in 100 mL of methanol, and NH2OH‚HCl (0.467 g,
6.72 mmol) and NaOAc (0.716 g, 8.73 mmol) were added. The
reaction mixture was stirred at 25 °C for 30 min, and it was
concentrated in vacuo. The residue was purified by flash
column chromatography (CH3OH/CH2Cl2/NH4OH,1:9:0.1) to
give (()-411
(0.765 g, 84%) as an oil: 1H NMR (500 MHz,
CDCl3) δ 6.83 (d, 1H, J ) 8.0 Hz), 6.57 (d, 1H, J ) 2.4 Hz),
6.47 (dd, 1H, J ) 8.1, 2.4 Hz), 3.50 (b, 2H), 3.10 (m, 1H), 2.82
(d, 1H, J ) 18.1 Hz), 2.68 (m, 1H), 2.58 (m, 1H), 2.47 (m, 1H),
2.31 (m, 1H), 2.04 (m, 1H), 1.89 (m, 1H), 1.83 (m, 1H), 1.34
(m, 1H), 1.32 (s, 3H), 0.85 (m, 1H), 0.85 (d, 3H, J ) 7.1 Hz),
0.49 (m, 2H), 0.10 (m, 2H); MS (CI) m/z 271 (M + H)+.
(-)-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-2,6-m eth a n o-3-ben za zocin -8-a m in e [(-)-5]
a n d (+)-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-
cis-6,11-dim eth yl-2,6-m eth an o-3-ben zazocin -8-am in e [(+)-
6]. By use of procedures11 identical to those reported for
preparing (()-4, cyclazocine enantiomers28,29 (-)-2 and (+)-3
were converted to (-)-5 and (+)-6, respectively, in yields nearly
identical to those reported (48% overall) for the racemate. For
(()-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-N-(4-m eth oxyp h en yl)-2,6-m eth a n o-3-ben -
za zocin -8-a m in e [((-16)]. Meth od B. Pd2(dba)3 (0.060 g,
0.066 mmol), DPPF (0.062 g, 0.197 mmol), and NaO-t-Bu
(0.126 g, 1.313 mmol) were added to an oven-dried 25 mL two-
neck flask equipped with reflux condenser contained in an N2-
filled glovebox. The flask was removed from the glovebox after
being capped with a rubber septum, and 4 mL of dry toluene
was added to the mixture via syringe. The resulting dark
suspension was stirred at ambient temperature for 10 min
when a solution of triflate (()-53 (0.300 g, 0.744 mmol) and
p-anisidine (0.110 g, 0.892 mmol) in 3 mL of toluene was
added. The resulting mixture was heated to 80 °C with
vigorous stirring. As judged by TLC, the reaction was complete
after 3 h. After cooling to ambient temperature, the mixture
was diluted with 20 mL of CH2Cl2, filtered over Celite, and
then preadsorbed onto silica gel. The crude reaction mixture
was purified by flash chromatography using hexanes/ethyl
acetate as eluent (20:1 f 10:1 f 3:1 gradient) to provide (()-
16 as a light-yellow oil that crystallized upon treatment with
(-)-5: mp 87-89 °C; [R]25 -118° (c 1.0, MeOH)]. Anal.
D
(C18H26N2‚0.25H2O) C, H, N. For (+)-6: mp 88-90 °C; [R]25
+118° (c 1.0, MeOH)]. Anal. (C18H26N2) C, H, N.
D
(()-3-(Cyclop r op ylm eth yl)-1,2,3,4,5,6-h exa h yd r o-cis-6,-
11-d im e t h yl-2,6-m e t h a n o-3-b e n za zocin -8-ylt r iflu or o-
m eth a n esu lfon ic Acid Ester [(()-53]. Meth od A. To a
stirred suspension of cyclazocine [(()-1, 1.00 g, 3.68 mmol] and
pyridine (0.76 g, 9.58 mmol) in 20 mL of dry CH2Cl2 under an
N2 atmosphere was added trifluoromethanesulfonic anhydride
(1.25 g, 4.42 mmol) at 0 °C. The resulting bright yellow-orange
suspension gradually became a solution after 15 min and was
stirred at 25 °C for an additional 2 h. The reaction was then
quenched with 20 mL of saturated NaHCO3 solution. The

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 847
1
ether (0.179 g, 0.475 mmol, 64%): mp 180-181 °C; H NMR
(500 MHz, CDCl3) δ 7.00 (m, 2H), 6.90 (m, 1H), 6.84 (m, 3H),
6.74 (m, 1H), 5.43 (bs, 1H), 3.78 (s, 3H), 3.15 (m, 1H), 2.87-
2.84 (d, J ) 18.0 Hz, 1H), 2.74 (m, 1H), 2.72 (m, 1H), 2.49-
2.34 (m, 2H), 2.17 (m, 1H), 2.07 (m, 1H), 1.93 (m, 1H), 1.36
(m, 1H), 1.31 (s, 3H), 1.26 (m, 1H), 0.86 (d, J ) 6.9 Hz, 3H),
0.51 (m, 2H), 0.11 (m, 2H); 13C NMR (125 MHz, CDCl3) δ
154.54, 143.01, 142.72, 136.71, 128.36, 127.80, 120.68, 114.61,
114.10, 113.78, 69.99, 59.84, 57.07, 55.54, 49.89, 45.97, 42.01,
41.67, 36.42, 25.47, 23.07, 14.22, 9.24, 4.03, 3.59; IR (CH2Cl2)
νmax 3356, 2094, 1652, 1506 cm-1; MS (CI) m/z 377 (M + H)+.
Anal. (C25H32N2O‚0.5H2O) C, H, N.
CDCl3) δ 6.89 (d, 1H, J ) 8.0 Hz), 6.50 (m, 1H), 6.43 (m, 1H),
3.53 (b, 1H), 3.10 (m, 1H), 2.83 (m, 1H), 2.82 (s, 3H), 2.69 (m,
1H), 2.60 (m, 1H), 2.48 (m, 1H), 2.32 (m, 1H), 2.05 (m, 1H),
1.90 (m, 1H), 1.84 (m, 1H), 1.36 (m, 1H), 1.35 (s, 3H), 0.86 (m,
1H), 0.86 (d, 3H, J ) 6.8 Hz), 0.50 (m, 2H), 0.11 (m, 2H); MS
(CI) m/z 285 (M + H)+.
(()-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-N-(2-m eth yleth yl)-2,6-m eth a n o-3-ben za zo-
cin -8-a m in e [(()-9]. Compound (()-4 (0.104 g, 0.385 mmol)
was dissolved in 20 mL of MeOH, and acetone (0.027 g, 0.462
mmol), NaBH3CN (0.024 g, 0.385 mmol), and ZnCl2 (0.027 g,
0.193 mmol) were added. The reaction mixture was stirred at
25 °C for 12 h and quenched with 10 mL of 1 N NaOH solution.
After evaporation to remove most of the MeOH, the resulting
solution was extracted three times with EtOAc. The combined
EtOAc extracts were washed with brine, dried over anhydrous
Na2SO4, filtered, and concentrated in vacuo to dryness. Flash
column chromatography of the resulting mixture provided the
desired product (()-9 (0.085 g, 70%) as a yellow semisolid: 1H
NMR (500 MHz, CDCl3) δ 6.84 (d, 1H, J ) 8.0 Hz), 6.46 (d,
1H, J ) 2.4 Hz), 6.40 (dd, 1H, J ) 8.1, 2.7 Hz), 3.58 (m, 1H),
3.09 (m, 1H), 2.81 (d, 1H, J ) 18.0 Hz), 2.68 (m, 1H), 2.57 (m,
1H), 2.47 (m, 1H), 2.32 (m, 1H), 2.05 (m, 1H), 1.88 (m, 1H),
1.83 (m, 1H), 1.35 (m, 1H), 1.32 (s, 3H), 1.20 (d, 3H, J ) 2.0
Hz), 1.18 (d, 3H, J ) 2.2 Hz), 0.85 (m, 1H), 0.85 (d, 3H, J )
7.1 Hz), 0.49 (m, 2H), 0.10 (m, 2H); IR (film) νmax 2962, 2915,
2834, 1613, 1506, 1462, 1429, 1379, 1320, 1256, 1177, 1099,
801 cm-1; MS (CI) m/z 313 (M + H)+. Anal. (C21H32N2) C, H,
N.
(-)-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-N-(p h en ylm eth yl)-2,6-m eth a n o-3-ben za zo-
cin -8-a m in e [(-)-26]. Meth od C. A flask was put into a
nitrogen-filled glovebox and charged with the triflate of (-)-2
(0.430 g, 1.067 mmol), Pd(OAc)2 (0.0048 g, 0.0213 mmol),
BINAP (0.0146 g, 0.0235 mmol), and NaO-t-Bu (0.144 g, 1.49
mmol). The flask was then capped with a rubber septum and
removed from the glovebox. A solution of benzylamine (0.137
g, 1.28 mmol) in 10 mL of dry toluene was then added to the
mixture via syringe. The reaction mixture was then heated to
80 °C in an oil bath with vigorous stirring. After 8 h, the
reaction mixture was cooled to 25 °C, diluted with ether,
filtered, and concentrated in vacuo. The resulting crude oil was
purified by flash column chromatography to give (-)-26 (0.205
g, 53%) as a colorless oil: 1H NMR (500 MHz, CDCl3) δ 7.27-
7.41 (m, 5H), 6.87 (d, 1H, J ) 8.1 Hz), 6.55 (d, 1H, J ) 2.5
Hz), 6.47 (dd, 1H, J ) 8.1, 2.4 Hz), 4.28 (d, 2H, J ) 4.9 Hz),
3.84 (b, 1H), 3.11 (m, 1H), 2.83 (d, 1H, J ) 18.1 Hz), 2.70 (m,
1H), 2.60 (m, 1H), 2.48 (m, 1H), 2.32 (m, 1H), 2.06 (m, 1H),
1.89 (m, 1H), 1.84 (m, 1H), 1.34 (m, 1H), 1.32 (s, 3H), 0.86 (m,
1H), 0.86 (d, 3H, J ) 7.0 Hz), 0.50 (m, 2H), 0.11 (m, 2H). IR
(film) νmax 2913, 1613, 1505, 1321, 456, 451 cm-1; MS (CI) m/z
7-(Cyclop r op ylm eth yl)-5,6,7,8,9,10-h exa h yd r o-cis-10,-
12-d im eth yl-6,10-m eth a n o-1H-p yr r olo[2,3-i][3]ben za zo-
cin e [(()-39]. A modification of a known procedure41 for the
conversion of anilines to indoles was used as follows. A mixture
of (()-4 (0.252 g, 0.93 mmol), bromoacetaldehyde diethyl acetal
(0.203 g, 1.03 mmol), KHCO3 (0.103 g, 1.03 mmol), and DMF
(5 mL) was stirred at 125 °C for 20 h. The solvent was removed
in vacuo, and the residue was purified by flash column
chromatography (CHCl3/MeOH/H2O, 15 mL:1 mL:2 drops) to
give (()-64 (0.194 g, 54%) having proton NMR and MS spectra
consistent with the desired structure. Compound (()-64 (0.070
g, 0.18 mmol) was stirred at 90 °C in a sealed tube for 72 h
with (CF3CO)2O (0.35 mL) and CF3CO2H (0.70 mL). After
concentration of the mixture in vacuo, the crude product was
purified by flash column chromatography (CH2Cl2/MeOH, 15:
1) to give 0.36 g of intermediate (()-65 having proton NMR
and MS spectra consistent with the desired structure. Without
further purification or characterization, (()-65 was treated
with 2 mL of 5% KOH in MeOH for 2 h at 25 °C. After
concentration in vacuo, the residue was partitioned between
H2O and CH2Cl2 and the organic layer was dried (Na2SO4) and
concentrated to give a crude product that was purified by
preparative silica gel TLC (CHCl3/MeOH/NH4OH, 33:7:1) to
give 0.019 g (53% from (()-64) of (()-3940 as a foam: 1H NMR
(500 MHz, CDCl3) δ 8.04 (br s, 1H), 7.33 (s, 1H), 7.30 (s, 1H),
7.14 (d, 1H, J ) 2.5 Hz), 6.43 (d, 1H, J ) 2.2 Hz), 3.27 (s, 1H),
3.10 (d, 1H, J ) 8.1 Hz), 2.93 (m, 1H), 2.77 (m, 1H), 2.59 (m,
1H), 2.44 (m, 1H), 2.13 (m, 1H), 2.04 (m, 2H), 1.46 (s, 3H),
1.37 (d, 1H, J ) 13.5 Hz), 1.25 (s, 3H), 0.93 (m, 1H), 0.90 (d,
J ) 7.1 Hz, 2H), 0.54 (d, J ) 7.8 Hz, 1H), 0.17 (m, 1H); MS
(CI) m/z 295 (M + H)+.
361 (M + H)+; [R]25 -101.9 ° (c 1.06, EtOH). Anal. (C25H32N2
D
) C, H, N.
(()-3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-cis-
6,11-d im eth yl-N-(4-d im eth yla m in op h en yl)-2,6-m eth a n o-
3-ben za zocin -8-a m in e [(()-21]. Meth od D. An oven-dried
25 mL two-neck flask equipped with reflux condenser was
placed into an N2-filled glovebox where it was charged with
Pd2(dba)3 (0.023 g, 0.025 mmol), DPPF (0.042 g, 0.076 mmol),
and NaO-t-Bu (0.005 g, 0.508 mmol). The system was capped
with a rubber septum and removed from the glovebox. Dry
toluene (4 mL) was then added to the mixture via syringe, and
the resulting dark suspension was stirred at 25 °C for 10 min.
A solution containing 4-bromo-N,N-dimethylaniline (0.101 g,
0.508 mmol) and amine (()-4 (0.164 g, 0.610 mmol) in 3 mL
of toluene was added, and the mixture was heated to 80°C
with vigorous stirring. After 6 h, the reaction had gone to
completion as evidenced by TLC. The mixture was allowed to
cool, diluted with 20 mL of CH2Cl2, filtered over Celite, and
then directly preadsorbed onto silica gel. The crude reaction
mixture was purified by flash column chromatography using
hexanes/ethyl acetate as eluent (20:1 f 10:1 f 3:1 gradient)
yielding (()-21 (0.134 g, 68%) as a brown oil: 1H NMR (500
MHz, CDCl3) δ 7.03 (d, 2H, J ) 1.7 Hz), 6.90 (d, 1H, J ) 1.9
Hz), 6.80 (s, 1H), 6.75 (d, 2H, J ) 1.7 Hz), 6.70 (d, 1H, J ) 1.8
Hz), 5.3 (m, 1H), 3.39 (m, 1H), 2.92 (s, 6H), 2.89-2.83 (m, 3H),
2.78 (m, 1H), 2.57 (m, 1H), 2.56 (m, 1H), 2.17 (m, 1H), 2.07
(m, 1H), 1.41 (m, 1H), 1.34, (s, 3H), 1.08 (m, 1H), 0.91 (d, 3H,
J ) 7.1 Hz), 0.61 (d, 2H, J ) 8.1 Hz), 0.23 (m, 2H); IR (CH2-
Cl2) νmax 3356, 2094, 1652, 1506 cm-1; MS (CI) m/z 390 (M +
H)+. Anal. (C26H35N3‚0.5H2O) C, H, N.
(()-7-(Cyclop r op ylm eth yl)-5,6,7,8,9,10-h exa h yd r o-cis-
10,12-d im eth yl-6,10-m eth a n o-1H-in d olo[2,3-i][3]ben za zo-
cin e [(()-40]. Phenylhydrazine hydrochloride (0.053 g, 0.364
mmol) was added to a solution of methyl enol ether (()-6611
(0.100 g, 0.348 mmol) in acetic acid (1 mL) under nitrogen.
The mixture was stirred at room temperature for 1 h and then
at reflux for 2 h. The reaction mixture was taken up in 10 mL
of ethyl acetate and washed with saturated sodium bicarbonate
solution and water. The organic phase was dried (Na2SO4) and
concentrated in vacuo to give a crude product as a brown oil.
Purification by flash column chromatography (CH2Cl2/MeOH/
NH4OH, 25:1:0.05) provided (()-40 (0.025 g, 21%) as white
3-(Cyclop r op ylm et h yl)-1,2,3,4,5,6,-h exa h yd r o-N,6,11-
tr im eth yl-cis-2,6-m eth a n o-3-ben za zocin -8-a m in e [(()-7].
Compound (()-37 (0.080 g, 0.21 mmol) was dissolved in 5 mL
of anhydrous MeOH, and 50% Pd(OH)2/C (0.040 g) was added
slowly. The suspension was placed in a Parr hydrogenation
apparatus and was shaken for 24 h at 25 °C at a pressure of
72 psi. The reaction mixture was then filtered over Celite, and
the filtrate was concentrated in vacuo. Preparative silica gel
TLC (CH3OH/CH2Cl2/NH4OH, 1:9:0.1) gave the desired prod-
uct (()-711 (0.027 g, 44%) as white foam: 1H NMR (500 MHz,
1
needles: mp 222-224 °C; H NMR (500 MHz, CDCl3) δ 8.34
(d, 1H, J ) 8.3 Hz), 8.15 (s, 1H), 7.40 (d, 1H, J ) 8.6 Hz), 7.35

848 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5
Wentland et al.
(3) Fu¨rst, S.; Hosztafi, S.; Friedmann, T. Structure-Activity Rela-
tionships of Synthetic and Semisynthetic Opioid Agonists and
Antagonists. Curr. Med. Chem. 1995, 1, 423-440.
(4) Mascarella, S. W.; Bai, X.; Williams, W.; Sine, B.; Bowen, W.
D.; Carroll, F. I. (+)-cis-N-(Para-, Meta-, and Ortho-substituted
benzyl)-N-normetazocines: Synthesis and Binding Affinity at the
[3H]-(+)-Pentazocine-Labeled (σ1) Site and Quantitative Struc-
ture-Affinity Relationship Studies. J . Med. Chem. 1995, 38,
565-569.
(5) Reden, J .; Reich, M. F.; Rice, K. C.; J acobson, A. E.; Brossi, A.
Deoxymorphines: Role of the Phenolic Hydroxyl in Antinocice-
ption and Opiate Receptor Interactions. J . Med. Chem. 1979,
22, 256-259.
(6) Hedberg, M. H.; J ohansson, A. M.; Fowler, C. J .; Terenius, L.;
Hacksell, U. Palladium-Catalyzed Synthesis of 3C-Substituted
3-Deoxymorphines. Bioorg. Med. Chem. Lett. 1994, 4, 2527-
2532.
(7) Kubota, H.; Rothman, R. B.; Dersch, C.; McCullough, K.; Pinto,
J .; Rice, K. C. Synthesis and Biological Activity of 3-Substituted
3-Desoxynaltrindole Derivatives. Bioorg. Med. Chem. Lett. 1998,
8, 799-804.
(8) Palmer, D. C.; Strauss, M. J . Benzomorphans: Synthesis,
Stereochemistry, Reactions and Spectroscopic Characterizations.
Chem. Rev. 1977, 77, 1-36.
(9) Martin, W. R.; Eades, C. G.; Thompson, J . A.; Huppler, R. E.;
Gilbert, P. E. The effects of morphine- and nalorphine-like drugs
in the nondependent and morphine-dependent chronic spinal
dog. J . Pharmacol. Exp. Ther. 1976, 197, 517-532.
(10) Fink, M.; Freedman, A. M.; Resnick, R.; Zaks, A. In Agonist and
Antagonist Actions of Narcotic Analgesic Drugs; Kosterlitz, H.
W., Collier, H. O. J ., Villareal, J . G., Eds.; Macmillan: New York,
1971; pp 266-276.
(11) Wentland, M. P.; Albertson, N. F.; Pierson, A. K. Synthesis and
Pharmacology of 8-Amino-3-(cyclopropylmethyl)-1,2,3,4,5,6-
hexahydro-cis-6,11-dimethyl-2,6-methano-3-benzazocine and Re-
lated Compounds. J . Med. Chem. 1980, 23, 71-74.
(12) Mule, S. J .; Clements, T. H.; Gorodetzky, C. W. Metabolic fate
of cyclazocine-3H in dogs. J . Pharmacol. Exp. Ther. 1968, 160,
387-396.
(t, 1H, J ) 7.2 Hz), 7.21 (d, 1H, J ) 8.3 Hz), 7.18 (t, 1H, J )
7.7 Hz), 7.15 (d, 1H, J ) 8.3 Hz), 3.29 (s, 1H), 3.02 (d, 1H d, J
) 5.9, 8.1 Hz), 2.93 (d, 1H, J ) 8.1 Hz), 2.84 (d, 1H, J ) 11.0
Hz), 2.59 (d, 1H, J ) 11.7 Hz), 2.50 (dd, 1H, J ) 6.4, 12.7 Hz),
2.30 (dd, 1H, J ) 6.8, 12.7 Hz), 2.07 (m, 3H), 1.77 (s, 3H), 1.67
(m, 1H), 0.93 (d, 3H, J ) 7.1 Hz), 0.91 (m, 1H), 0.50 (m, 2H),
0.13 (m, 1H), 0.08 (m, 1H); IR (CHCl3) νmax 3357, 2974, 1413,
1049 cm-1; MS (CI) m/z 345 (M + H)+. Anal. (C24H28N2‚0.5H2O)
C, H, N.
Ra d iola beled Liga n d Bin d in g Assa ys. Binding assays
used to screen compounds are similar to those previously
reported.43 Guinea pig brain membranes, 500 µg of membrane
protein, were incubated with 12 different concentrations of the
compound in the presence of 1 nM [3H]U69,59353 (κ), 0.25 nM
[3H]DAMGO54 (µ), or 0.2 nM [3H]naltrindole55 (δ) in a final
volume of 1 mL of 50 mM Tris-HCl, pH 7.5 at 25 °C. Incubation
times of 60 min were used for [3H]U69,593 and [3H]DAMGO.
Because of a slower association of [3H]naltrindole with the
receptor, a 3 h incubation was used with this radioligand.
Samples incubated with [3H]naltrindole also contained 10 µM
MgCl2 and 0.5 mM phenylmethylsulfonyl fluoride. Nonspecific
binding was measured by inclusion of 10 µM naloxone. The
binding was terminated by filtering the samples through
Schleicher & Schuell no. 32 glass fiber filters using a Brandel
48-well cell harvester. The filters were subsequently washed
three times with 3 mL of cold 50 mM Tris-HCl, pH 7.5, and
were counted in 2 mL of Ecoscint A scintillation fluid. For [3H]-
naltrindole and [3H]U69,593 binding, the filters were soaked
in 0.1% polyethylenimine for at least 60 min before use. IC50
values were calculated by a least-squares fit to a logarithm-
probit analysis. Ki values of unlabeled compounds were
calculated from the equation
IC50
(13) Lynn, R. K.; Smith, R. G.; Leger, R. M.; Deinzer, M. L.; Griffin,
D.; Gerber, N. Identification of glucuronide metabolites of
benzomorphan narcotic analgesic drugs in bile from the isolated
perfused rat liver by gas chromatography and mass spectrom-
etry. Drug Metab. Dispos. 1977, 5, 47-55.
(14) Walker, J . M.; Bowen, W. D.; Walker, F. O.; Matsumoto, R. R.;
de Costa, B.; Rice, K. C. Sigma receptors: Biology and function.
Pharmacol. Rev. 1990, 42, 355-402.
Ki )
1 + S
where S ) (concentration of radioligand)/(Kd of radioligand).56
Data are the mean ( SE from at least three experiments
performed in triplicate.
[
35S]GTP γS Bin d in g Assa ys. In a final volume of 0.5 mL,
12 different concentrations of each test compound were
incubated with 15 µg (κ) or 7.5 µg (µ) of CHO cell membranes
that stably expressed either the human κ or µ opioid receptor.
The assay buffer consisted of 50 mM Tris-HCl, pH 7.4, 3 mM
MgCl2, 0.2 mM EGTA, 3 µM GDP, and 100 mM NaCl. The
final concentration of [35S]GTPγS was 0.080 nM. Nonspecific
binding was measured by inclusion of 10 µM GTPγS. Binding
was initiated by the addition of the membranes. After an
incubation of 60 min at 30°C, the samples were filtered
through Schleicher & Schuell no. 32 glass fiber filters. The
filters were washed three times with cold 50 mM Tris-HCl,
pH 7.5, and were counted in 2 mL of Ecoscint scintillation
fluid. Data are the mean Emax and EC50 values ( SE from at
least three separate experiments, performed in triplicate. For
calculation of the Emax values, the control [35S]GTPγS binding
was set at 0%.
(15) de Costa, B. R.; Bowen, W. D.; Hellewell, S. B.; Walker, J . M.;
Thurkauf, A.; J acobson, A. E.; Rice, K. C. Synthesis and
evaluation of optically pure [3H]-(+)-pentazocine, a highly potent
and selective radioligand for sigma receptors. FEBS Lett. 1989,
251, 53-58.
(16) Martin, W. R.; Fraser, H. F.; Gorodetsky, C. W.; Rosenberg, D.
E. Studies of the dependence-producing potential of the narcotic
antagonist 2-cyclopropylmethyl-2′-hydroxy-5,9-dimethyl-6,7-ben-
zomorphan (cyclazocine), WIN 20740, ARC II-3. J . Pharmacol.
Exp. Ther. 1965, 150, 426-436.
(17) Resnick, R.; Fink, M.; Freedman, A. M. Cyclazocine treatment
of opiate dependence: a progress report. Comp. Psychiatry 1971,
12, 496-501.
(18) Harris, L. S.; Pierson, A. K. Some narcotic antagonists in the
benzomorphan series. J . Pharmacol. Exp. Ther. 1964, 143, 141-
148.
(19) Preston, K. L.; Umbricht, A.; Schroeder, J . R.; Abreu, M.;
Pickworth, W. B. Cyclazocine: Comparison to Hydromorphone
and Interaction with Cocaine National Institute on Drug Abuse.
Drug Alcohol Depend. 2001, 63 (Suppl. 1), S126.
Ack n ow led gm en t. We gratefully acknowledge the
discussions with the late Professors Sydney Archer and
Arthur G. Schultz of Rensselaer and also those with
Professor Frank Hauser (University at Albany). Also
acknowledged is the contribution of Ms. Hongmei Zhang
of Rensselaer. We thank J ames Hilbert, Brian Knapp,
and Emily DeVita of the University of Rochester for
excellent technical assistance. Funding of this research
was from NIDA (Grants DA01674, DA03742, DA12180,
and KO5-DA00360).
(20) Wentland, M. P.; Xu, G.; Cioffi, C. L.; Ye, Y.; Duan, W.; Cohen,
D. J .; Colasurdo, A. M.; Bidlack, J . M. 8-Aminocyclazocine
Analogues: Synthesis and Structure-Activity Relationships.
Bioorg. Med. Chem. Lett. 2000, 10, 183-187.
(21) Bidlack, J . M.; McLaughlin, J . P.; Wentland, M. P. Partial
Opioids: Medications for the Treatment of Pain and Drug Abuse.
Ann. N. Y. Acad. Sci. 2000, 909, 1-11.
(22) Hartwig, J . F. Transition Metal Catalyzed Synthesis of Ary-
lamines and Aryl Ethers from Aryl Halides and Triflates: Scope
and Mechanism. Angew. Chem., Int. Ed. 1998, 37, 2046-2067.
(23) Danso-Danquah, R.; Bai, X.; Zhang, X.; Mascarella, S. W.;
Williams, W.; Sine, B.; Bowen, W.; Carroll, F. I. Synthesis and
σ Binding Properties of 2′-Substituted 5,9-Dimethyl-6,7-benzo-
morphans. J . Med. Chem. 1995, 38, 2978-2985.
(24) Wentland, M. P.; Duan, W.; Cohen, D. J .; Bidlack, J . M. Selective
Protection and Functionalization of Morphine: Synthesis and
Opioid Receptor Binding Properties of 3-Amino-3-desoxymor-
phine Derivatives. J . Med. Chem. 2000, 43, 3558-3565.
(25) McCurdy, C. R.; J ones, R. M.; Portoghese, P. S. Investigation of
Phenolic Bioisosterism in Opiates: 3-Sulfonamido Analogues of
Naltrexone and Oxymorphone. Org. Lett. 2000, 2, 819-821.
Refer en ces
(1) Archer, S.; Glick, S. D.; Bidlack, J . M. Cyclazocine Revisited.
Neurochem. Res. 1996, 21, 1369-1373.
(2) Aldrich, J . V. In Burger’s Medicinal Chemistry and Drug
Discovery; Wolff, M. E., Ed.; J ohn Wiley & Sons: New York,
1996; Vol. 3, pp 321-441.

8-Amino-2,6-methano-3-benzazocines
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 5 849
(26) (a) Newman, A. H.; Bevan, K.; Bowery, N.; Tortella, F. C.
Synthesis and Evaluation of 3-Substituted 17-Methylmorphinan
Analogs as Potential Anticonvulsant Agents. J . Med. Chem.
1992, 35, 4135-4142. (b) Tortella, F. C.; Britton, P.; Williams,
A.; Lu, X. C. M.; Newman, A. H. Neuroprotection (Focal
Ischemia) and Neurotoxicity (Electroencephalographic) Studies
in Rats with AHN649, a 3-Amino Analog of Dextromorphan and
Low-Affinity N-Methyl-D-aspartate Antagonist. J . Pharmacol.
Exp. Ther. 1999, 291, 399-408.
(27) For all optically active compounds reported in this paper, the
(-)-isomers are (2R,6R,11R) or (2S,6R,11R) and the (+)-isomers
are (2S,6S,11S) or (2R,6S,11S) depending if there is a 1-keto
group or not.
(28) Tullar, B. F.; Harris, L. S.; Perry, R. L.; Pierson, A. K.; Soria, A.
E.; Wetterau, W. F.; Albertson, N. F. Benzomorphans. Optically
Active and trans Isomers. J . Med. Chem. 1967, 10, 383-386.
(29) Optical purity was assessed by reacting known intermediates
(the enantiomers of normetazocine) with (S)-(-)-methylbenzyl-
isocyanate and evaluation of the proton NMR of the resulting
ureas using a known procedure (ref 30).
(30) Brine, G. A.; Berrang, B.; Hayes, J . P.; Carroll, F. I. An Improved
Resolution of (()-cis-N-Normetazocine. J . Heterocycl. Chem.
1990, 27, 2139-2143.
(31) Wolfe, J . P.; Buchwald, S. L. Palladium-Catalyzed Amination
of Aryl Triflates. J . Org. Chem. 1997, 62, 1264-1267.
(32) Louie, J .; Driver, M. S.; Hamann, B. C.; Hartwig, J . F. Palladium-
Catalyzed Amination of Aryl Triflates and Importance of Triflate
Addition Rate. J . Org. Chem. 1997, 62, 1268-1273.
(33) Wolfe, J . P.; A° hman, J .; Sadighi, J . P.; Singer, R. A.; Buchwald,
S. L. An Ammonia Equivalent for the Palladium-Catalyzed
Amination of Aryl Halides and Triflates. Tetrahedron Lett. 1997,
38, 6367-6370.
(34) Pd2(dba)3, tris(dibenzylideneacetone)dipalladium(0); DPPF, 1,1′-
bis(diphenylphosphino)ferrocene; BINAP, 2,2′-bis(diphenylphos-
phino)-1,1′-binaphthyl.
(35) Michne, W. F.; Albertson, N. F. Analgetic 1-Oxidized-2,6-
methano-3-benzazocines. J . Med. Chem. 1972, 15, 1278-1281.
(36) Albertson, N. F. 1,2,3,4,5,6-Hexahydro-1-oxo or hydroxy-3-acyl-
2,6-methano-3-benzazocines and Corresponding 1-Esters thereof.
U.S. Patent 3936462, 1976.
(42) Kuroki, M.; Tsunashima, Y. The Chemistry of Carbazoles. VII.
Syntheses of Methylcarbazoles. J . Heterocycl. Chem. 1981, 18,
709-714.
(43) Archer, S.; Seyed-Mozaffari, A.; J iang, Q.; Bidlack, J . M. 14R,-
14′â-[Dithiobis[2-oxo-2,1-ethanediyl)imino]]bis(7,8-dihydromor-
phinone) and 14R,14′â-[Dithiobis[(2-oxo-2,1-ethanediyl)imino]]-
bis-7,8-dihydro-N-(cyclopropylmethyl)normorphinone: Chemistry
and Opioid Binding Properties. J . Med. Chem. 1994, 37, 1578-
1585.
(44) Gmerek, D. E.; Dykstra, L. A.; Woods, J . H. Kappa opioids in
rhesus monkeys. III. Dependence associated with chronic ad-
ministration. J . Pharmacol. Exp. Ther. 1987, 242, 428-436.
(45) Butelman, E. R.; France, C.; Woods, J . H. Apparent pA2 analysis
of respiratory depressant effects of alfentanil, etonitazene,
ethylketocyclazocine (EKC) and Mr2033 in rhesus monkeys. J .
Pharmacol. Exp. Ther. 1993, 264, 145-151.
(46) Negus, S. S.; Mello, N. K.; Portoghese, P. S.; Lin, C. E. Effects
of kappa opioids on cocaine self-administration by rhesus
monkeys. J . Pharmacol. Exp. Ther. 1997, 282, 44-55.
(47) Mello, N. K.; Negus, S. S. Effects of kappa opioid agonists on
cocaine- and food-maintained responding by rhesus monkeys.
J . Pharmacol. Exp. Ther. 1998, 286, 812-824.
(48) Topliss, J . G. A Manual Method for Applying the Hansch
Approach to Drug Design. J . Med. Chem. 1977, 20, 463-469.
(49) Patani, G. A.; LaVoie, E. J . Bioisosterism: A Rational Approach
in Drug Design. Chem. Rev. 1996, 96, 3147-3176.
(50) Wentland, M. P.; Lou, R.; Ye, Y.; Cohen, D. J .; Richardson, G.
P.; Bidlack, J . M. 8-Carboxamidocyclazocine Analogues: Rede-
fining the Structure-Activity Relationships of 2,6-Methano-3-
benzazocines. Bioorg. Med. Chem. Lett. 2001, 11, 623-626.
(51) Wentland, M. P.; Lou, R.; Dehnhardt, C. M.; Duan, W.; Cohen,
D. J .; Bidlack, J . M. 3-Carboxamido Analogues of Morphine and
Naltrexone: Synthesis and Opioid Receptor Binding Properties.
Bioorg. Med. Chem. Lett. 2001, 11, 1717-1721.
(52) Bidlack, J . M.; Cohen, D. J .; McLaughlin, J . P.; Lou, R.; Ye, Y.;
Wentland, M. P. 8-Carboxamidocyclazocine: A Long-Acting,
Novel Benzomorphan. J . Pharmacol. Exp. Ther. 2002, 302, 374-
380.
(53) Xia, Q.; Tai, K. K.; Wong, T. M. Chronic morphine treatment
increases the number, but decreases the affinity of [3H]-U 69593
binding sites in the rat heart. Life Sci. 1992, 50, 1143-1148.
(54) Pivovarchik, M. V.; Grinevich, V. P. The binding of the opioid
receptors of the rat brain to specific ligands under acute alcohol
intoxication. Vestsi Nats. Akad. Navuk Belarusi, Ser. Biyal.
Navuk 2000, 72-74; Chem. Abstr. 2001, 134, 337050.
(55) Dorn, C. R.; Markos, C. S.; Dappen, M. S.; Pitzele, B. S. Synthesis
of tritium-labeled naltrindole. J . Labelled Compd. Radiopharm.
1992, 31, 375-380.
(56) Cheng, Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (KI) and the concentration of inhibitor which causes
50% inhibition (I50) of an enzymatic reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
(37) Albertson, N. F. 1-Oxygenated-2,6-methano-3-benzazocines. U.S.
Patent 4161598, 1979.
(38) Merz, H.; Stockhaus, K. N-[(Tetrahydrofuryl)alkyl] and N-
(Alkoxyalkyl) Derivatives of (-)-Normetazocine, Compounds
with Differentiated Opioid Action Profiles. J . Med. Chem. 1979,
22, 1475-1483.
(39) Wagaw, S.; Buchwald, S. L. The Synthesis of Aminopyridines:
A Method Employing Palladium-Catalyzed Carbon-Nitrogen
Bond Formation. J . Org. Chem. 1996, 61, 7240-7241.
(40) Allen, R. C.; Wettlaufer, D. G. Preparation of 2,6-Methanopyr-
rolo-3-benzazocines as Analgesics. U.S. Patent 4929622, 1990.
(41) Nordlander, J . E.; Catalane, D. B.; Kotian, K. D.; Stevens, R.
M.; Haky, J . E. Synthesis of Indoles from N-(Trifluoroacetyl)-
2-anilino Acetals. J . Org. Chem. 1981, 46, 778-782.
J M020429W