Probes for narcotic receptor mediated phenomena. 34. Synthesis and structure - Activity relationships of a potent μ-agonist δ-antagonist and an exceedingly potent antinociceptive in the enantiomeric C9-substituted 5-(3-hydroxyphenyl)-N-phenylethylmorphan

Article abstract of DOI:10.1021/jm061325e

Both of the enantiomers of 5-(3-hydroxyphenyl)-N-phenylethylmorphan with C9α-methyl, C9-methylene, C9-keto, and C9α- and C9β-hydroxy substituents were synthesized and pharmacologically evaluated. Three of the 10 compounds, (1R,5R,9S)-(-)-9-hydroxy-5-(3-hy

Full text of DOI:10.1021/jm061325e

J. Med. Chem. 2007, 50, 3765-3776
3765
Probes for Narcotic Receptor Mediated Phenomena. 34. Synthesis and Structure-Activity
Relationships of a Potent µ-Agonist δ-Antagonist and an Exceedingly Potent Antinociceptive in
the Enantiomeric C9-Substituted 5-(3-Hydroxyphenyl)-N-phenylethylmorphan Series
Anne-Ce´cile Hiebel,†,‡ Yong Sok Lee, Edward Bilsky,∞ Denise Giuvelis,∞ Jeffrey R. Deschamps,§ Damon A. Parrish,§
Mario D. Aceto,¢ Everette L. May,¢ Louis S. Harris,¢ Andrew Coop,| Christina M. Dersch,¶ John S. Partilla,¶
Richard B. Rothman,¶ Kejun Cheng,‡ Arthur E. Jacobson,‡ and Kenner C. Rice*,‡
Drug Design and Synthesis Section, Chemical Biology Research Branch, National Institute on Drug Abuse, and Center for Molecular
Modeling, DiVision of Computational Bioscience, Center for Information Technology, National Institutes of Health, DHHS, Bethesda, MD
20892, UniVersity of New England, College of Osteopathic Medicine, Biddeford, ME 04005, Laboratory for the Structure of Matter, NaVal
Research Laboratory, Washington, DC 20375, Department of Pharmacology and Toxicology, School of Medicine, Virginia Commonwealth
UniVersity, Richmond VA 23298, Department of Pharmaceutical Sciences, UniVersity of Maryland School of Pharmacy, Baltimore, MD 21201,
and Clinical Psychopharmacology Section, National Institute on Drug Abuse, Addiction Research Center, National Institutes of Health, DHHS,
Baltimore, MD 21224
ReceiVed NoVember 15, 2006
Both of the enantiomers of 5-(3-hydroxyphenyl)-N-phenylethylmorphan with C9R-methyl, C9-methylene,
C9-keto, and C9R- and C9â-hydroxy substituents were synthesized and pharmacologically evaluated. Three
of the 10 compounds, (1R,5R,9S)-(-)-9-hydroxy-5-(3-hydroxyphenyl-2-phenylethyl-2-azabicyclo[3.3.1]-
nonane ((1R,5R,9S)-(-)-10), (1R,5S)-(+)-5-(3-hydroxyphenyl)-9-methylene-2-phenethyl-2-azabicyclo[3.3.1]-
nonane ((1R,5S)-(+)-14), and (1R,5S,9R)-(-)-5-(3-hydroxyphenyl)-9-methyl-2-phenethyl-2-azabicyclo-
[3.3.1]nonane ((1R,5S,9R)-(+)-15) had subnanomolar affinity at µ-opioid receptors (Ki ) 0.19, 0.19, and
0.63 nM, respectively). The (1R,5S)-(+)-14 was found to be a µ-opioid agonist and a µ-, δ-, and κ-antagonist
in [35S]GTP-γ-S assays and was approximately 50 times more potent than morphine in a number of acute
and subchronic pain assays, including thermal and visceral models of nociception. The (1R,5R,9S)-(-)-10
compound with a C9-hydroxy substituent axially oriented to the piperidine ring (C9â-hydroxy) was a µ-agonist
about 500 times more potent than morphine. In the single-dose suppression assay, it was greater than 1000
times more potent than morphine. It is the most potent known phenylmorphan antinociceptive. The molecular
structures of these compounds were energy minimized with density functional theory at the B3LYP/6-31G*
level and then overlaid onto (1R,5R,9S)-(-)-10 using the heavy atoms in the morphan moiety as a common
docking point. Based on modeling, the spatial arrangement of the protonated nitrogen atom and the 9â-OH
substituent in (1R,5R,9S)-(-)-10 may facilitate the alignment of a putative water chain enabling proton
transfer to a nearby proton acceptor group in the µ-opioid receptor.
The 5-(3-hydroxy)phenylmorphan class of opioids (3-(2-
azabicyclo[3.3.1]nonan-5-yl)phenol, 1a, Figure 1) was originally
synthesized over 50 years ago by May and Murphy,1 and this
type of molecular structure is still of considerable current
interest.2-8 Our interest in the phenylmorphan template has been
enhanced by our finding4,9 that N-phenylethyl substitution (1b,
Figure 1) apparently transforms both enantiomers of 1a from
agonists into antagonists, as determined in [35S]GTP-γ-S assays,
Figure 1. Structures of racemic phenylmorphans.
and that suitable modification of a phenylmorphan can provide
a compound with both µ-agonist and δ-antagonist activity.10
enantiomer that was found to be a µ- and κ-antagonist.4,9 We
This combination of receptor interactions in the peptide family
noted that these compounds appeared to be less potent than a
of opioid-like agonists and antagonists has been noted to produce
racemic mixture that was formerly reported12 with a C9â-methyl
analgesia with less tolerance and dependence.11
substituent, (()-5-(3-hydroxyphenyl)-9â-methyl-2-(2′-phenyl-
The 1S,5R enantiomer of 1b (Figure 1) was a fairly potent
ethyl)-2-azabicyclo[3.3.1]nonane (1c), and we stated that the
µ-, δ-, and κ-opioid antagonist, more potent than the 1R,5S
potency difference, rather than the antagonist activity, was more
* To whom correspondence should be addressed. Telephone: 301-496-
reasonably ascribed to the formerly proposed hindered rotation
1856. Fax: 301-402-0589. E-mail: kr21f@nih.gov.
of the 5-hydroxyphenyl ring.
† Current address: Provid Pharmaceuticals, Inc., 671 US Highway 1,
The molecular conformation of the 5-(3-hydroxy)phenylmor-
phan is different from that of the morphinans, 4,5-epoxymor-
phinans and 6,7-benzomorphans, because its phenolic ring is
equatorially oriented on the piperidine ring, not constrained
axially as in these other classes of opioids. Yet many 5-(3-
hydroxy)phenylmorphan derivatives show opioid-like in vitro
and in vivo activities, and others are opioid antagonists. The
different spatial positions of the aromatic rings of the phenyl-
North Brunswick, NJ 08902.
‡ This work was initiated while these authors were members of the Drug
Design and Synthesis Section, Laboratory of Medicinal Chemistry, National
Institute of Diabetes, Digestive and Kidney Diseases, NIH.
Center for Molecular Modeling.
∞ University of New England.
§ Laboratory for the Structure of Matter.
¢ Department of Pharmacology and Toxicology.
| Department of Pharmaceutical Sciences.
¶ Clinical Psychopharmacology Section.
10.1021/jm061325e CCC: $37.00 © 2007 American Chemical Society
Published on Web 07/11/2007

3766 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Hiebel et al.
Scheme 1
morphans and the morphinans and the wide-range of molecular
structures that interact with the µ-opioid receptor have made it
difficult to find a reasonable hypothesis that could relate their
molecular structures to pharmacological activity either quanti-
tatively or qualitatively.
Because racemic 1c12 showed good affinity for opioid
receptors (Ki ) 3 nM) and displayed naltrexone-like nonselective
antagonist activity in ex vivo experiments, we thought that it
might be of interest to examine other substituents at the C9
position to see what effect they would have on the affinity and
selectivity of the 5-phenylmorphan. Thus, we have synthesized
both enantiomers of five derivatives of 5-(3-hydroxyphenyl)-
N-phenylethylmorphan with various substituents at C9. These
were the 9R-methyl, the 9-methylene, the 9-keto, and the 9R-
and 9â-hydroxy compounds. Two of them, the 9â-hydroxy
((1R,5R,9S)-(-)-10) and the 9-methylene ((1R,5S)-(+)-14)
derivatives, were found to be especially interesting. The former
was found to be extraordinarily potent as an antinociceptive,
and the latter was found to be a µ-agonist and a µ-, δ- and
κ-antagonist.
Chemistry
Phenylmorphans (1R,5R)-(+)-6, (1R,5R,9R)-(+)-8, (1R,5R,9S)-
(-)-10, (1R,5S)-(+)-14, and (1R,5S,9R)-(+)-15 and their enan-
tiomers (1S,5S)-(-)-6, (1S,5S,9S)-(-)-8, (1S,5S,9R)-(+)-10,
(1S,5R)-(-)-14, and (1S,5R,9S)-(-)-15, stemmed from common
intermediates (1R,5R)-(+)-5-(3-methoxyphenyl)-9-oxo-2-phe-
nylethyl-2-azabicyclo[3.3.1]nonane ((1R,5R)-(+)-5) and its enan-
tiomer (1S,5S)-(-)-5, respectively. The desired enantiomers of
5 were prepared from (()-2. The racemic compound 2 was
obtained in nine steps (23.8% overall yield) from commercially
available reagents using literature procedures.1,2 The enantiomers
of 2, (1R,5R)-(+)-2 and (1S,5S)-(-)-2, were obtained by
selective salt formation with D-(-)- and L-(+)-tartaric acid
(Scheme 1). The absolute configuration of the tartrate salt of
(1S,5S)-(-)-2 was determined by single-crystal X-ray analysis,
based on the configuration of the known tartaric acid used for
the preparation of the salt (Figure 2).
Figure 2. X-ray crystallographic structures of (1S,5S)-5-(3-methox-
yphenyl)-2-methyl-9-oxo-2-azabicyclo[3.3.1]nonane ((1S,5S)-(-)-2),
(1S,5S,9S)-(-)-9-hydroxy-5-(3-hydroxyphenyl-2-phenylethyl-2-azabicyclo-
[3.3.1]nonane ((1S,5S,9S)-(-)-8), and (1R,5S,9R)-(+)-5-(3-hydrox-
yphenyl)-9-methyl-2-phenethyl-2-azabicyclo[3.3.1]nonane ((1R,5S,9R)-
(+)-15).
Scheme 2a
Initial crystallization of the L-(+)-tartaric acid salt gave an
enantiomeric excess of 85% for (-)-2, as determined by NMR,
with the use of a chiral shift reagent, R-(-)-1-phenyl-2,2,2-
trifluoroethanol.13 Recrystallization of the tartrate salt gave
(1S,5S)-5-(3-methoxyphenyl)-2-methyl-9-oxo-2-azabicyclo[3.3.1]-
nonane ((1S,5S)-(-)-2) in an enantiomeric excess greater than
99% by NMR. The (1R,5R)-(+)-2 enantiomer was obtained in
the same manner by crystallization with the other antipode of
tartaric acid. Having both enantiomers in hand, the N-methyl
substituent was replaced by N-phenylethyl (Scheme 2). This
was achieved via the carbamates ethyl-5-(3-methoxyphenyl)-
9-oxo-2-azabicyclo[3.3.1]nonane-2-carboxylate, (1R,5R)-(+)-3
and its enantiomer (1S,5S)-(-)-3. For example, the secondary
amine (1R,5R)-5-(3-methoxyphenyl)-9-oxo-2-azabicyclo[3.3.1]-
nonane ((1R,5R)-(-)-4) was obtained in good yield after careful
treatment with trimethylsilyliodide.14,15 Alkylation gave the
desired N-phenylethyl substituent in nearly quantitative yield
((1R,5R)-5-(3-methoxyphenyl)-9-oxo-2-phenylethyl-2-azabicyclo-
a Reagents and conditions: (a) ethylchloroformate, K2CO3, 1,2-dichlo-
roethane, 87%; (b) TMSI, 1,2-dichloroethane, 50 °C, 72%; (c) phenyleth-
ylbromide, K2CO3, KI, acetonitrile, 95%.
[3.3.1]nonane ((1R,5R)-(+)-5)). The enantiomers (1S,5S)-(+)-4
and (1S,5S)-(-)-5 were prepared similarly.
The intermediate (1R,5R)-5-(3-hydroxyphenyl)-9-oxo-2-phe-
nylethyl-2-azabicyclo[3.3.1]nonane ((1R,5R)-(+)-6; Scheme 3)

Enantiomeric C9-Substituted Phenylmorphans
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16 3767
Table 1. [125I]Ioxy Binding Data for C9-Substituted
Scheme 3a
5-Hydroxyphenylmorphansa
a Reagents and conditions: (a) HBr, AcOH, reflux, 75%; (b) NaBH4,
MeOH, 90%; (c) BBr3, CH2Cl2, 72-96%; (d) superhydride, THF, 97%.
Scheme 4a
a
[
125I]IOXY binding was performed using CHO cells that were stably
transfected with either human µ-, δ-, or κ-DNA and express the human µ-,
δ-, or κ-opioid receptor, respectively. All results are (SD, n ) 3. b See ref
3.
12 was successful using either Tebbe’s reagent18,19 or Peterson
olefination20 conditions, and the latter gave higher yields. The
phenylmorphan enantiomers (1S,5S)-(-)-6, (1S,5S,9S)-(-)-8,
(1S,5S,9R)-(+)-10, (1S,5R)-(-)-14, and (1S,5R,9S)-(-)-15 were
synthesized in the same manner as their antipodes, starting from
(1S,5S)-(-)-2.
All of the final 10 enantiomers were purified by radial PLC
and recrystallization of their salts (HCl or oxalate). The relative
stereochemistries of compounds (1S,5S,9S)-(-)-8 and (1R,5S,9R)-
(+)-15 were confirmed by single-crystal X-ray crystallography
(Figure 2).
Biological Results and Discussion
Initial binding assays in human cells (Table 1) using 125I-
IOXY indicated that (1R,5R,9S)-(-)-10, (1R,5S)-(+)-14, and
(1R,5S,9R)-(+)-15 had extremely high affinity (subnanomolar)
at the µ-receptor. Both (1R,5R,9S)-(-)-10 and (1R,5S)-(+)-14
had good affinity at the δ-receptor, and the latter had subna-
nomolar affinity at the κ-receptor as well; (1R,5S,9R)-(+)-15
also had good affinity at the κ-receptor. The enantiomers of
these three compounds showed less affinity at opioid receptors,
although (1S,5R)-(-)-14 and (1S,5R,9S)-(-)-15 had good af-
finity (<10 nM) at µ- and a little less affinity at κ-receptors. It
is interesting to note that all of the C9-oxygen-containing
compounds had at least 10-fold less affinity at κ-receptors than
the C9-alkyl or alkylidene compounds. The three most interest-
ingcompounds,(1R,5R,9S)-(-)-10,(1R,5S)-(+)-14,and(1R,5S,9R)-
(+)-15 were further evaluated in [35S]GTP-γ-S assays, using
CHO cells. The (1R,5S)-(+)-14 appeared to be a very interesting
µ-agonist and a µ-, δ-, and κ-antagonist in the [35S]GTP-γ-S
assay, while (1R,5R,9S)-(-)-10, the C9â-hydroxy compound
a Reagents and conditions: (a) TMS-CH2MgCl, THF, 92%; (b) NaH,
THF, reflux, 98%; (c) Tebbe’s reagent,18,19 THF, 0 °C to room temperature,
75%; (d) BBr3, CH2Cl2, 72-96%; (e) H2, methanol, 93%.
was obtained by demethylation of the aryl methoxy group of
(1R,5R)-(+)-5. Alcohols (1R,5R,9R)-(+)-8 and (1R,5R,9S)-
(-)-10 were obtained selectively via either sodium borohy-
dride16 or lithium triethyl borohydride (superhydride)16 reduc-
tion, respectively, followed by demethylation.17
Two less polar compounds than (1R,5R,9S)-(-)-10, the
(1R,5S)-(+)-14 and (1R,5S,9R)-(+)-15 (Scheme 4), were syn-
thesized from the exocyclic methylene compound (1R,5S)-(+)-
12. The transformation leading to phenylmorphan (1R,5S)-(+)-

3768 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Table 2. Functional Data ([35S]GTP-γ-S)a
Hiebel et al.
µ-agonism
ED50
(nM)
δ-agonism
Emax
ED50
(nM)
κ-agonism
Emax
ED50
(nM)
b
Emax
%
(1R,5R,9S)-(-)-10
(1R,5S)-(+)-14
(1R,5S,9R)-(+)-15
morphine
DAMGO
SNC-80
117 ( 2
38 ( 1
14 ( 1
86 ( 3
100 ( 3
5.0 ( 0.5
2.7 ( 0.3
8.7 ( 2.2
37 ( 6
73 ( 3
n.d.
n.d.
341 ( 79
n. d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
42 ( 4
100 ( 4
26 ( 4
µ-antagonism
Kec (nM)
δ-antagonism
Ked (nM)
κ-antagonism
Ked (nM)
(1R,5R,9S)-(-)-10
n.d.
n.d.
12 ( 2
31 ( 5
610 ( 120
2.0 ( 0.2
6.0 ( 0.3
(1R,5S)-(+)-14
(1R,5S,9R)-(+)-15
naloxone
2.2 ( 0.6
4.4 ( 0.7
2.3 ( 0.3
naltrindole
norBNI
0.18 ( 0.01
0.11 ( 0.02
a [35S]GTP-γ-S binding was performed using CHO cells stably transfected with human µ-, δ-, or κ-cDNA and that stably express µ-, δ-, or κ-opiate
receptors; n.d. ) not determined. b The Emax is the extrapolated maximal stimulation where 100% is defined as the stimulation produced by 1 µM DAMGO
(465 ( 13 for µ-receptors) and 500 nM SNC80 (425 ( 15 for δ-receptors). In initial screening experiments, no compound had κ-agonist activity. c For
µ-receptors, Ke values were calculated according to the equation: [test drug]/(EC50-2/EC50-1 - 1), where EC50-2 is the EC50 value of DAMGO in the
presence of a fixed concentration of the test drug and EC50-1 is the value in the absence of the test drug. d For δ- and κ-receptors, Ke values were determined
as described in the section Data Analysis and Statistics. Each parameter value is (SD.
(the hydroxyl moiety was axially oriented in the piperidine ring),
displayed agonist activity at µ- and δ-receptors and was a weak
κ-antagonist (Table 2). The (1R,5S,9R)-(+)-15, bearing an
R-methyl substituent at C9 (equatorially oriented on the pip-
eridine ring), showed δ- and κ-antagonist activity and was a
weaker µ-agonist than the other two compounds in the [35S]-
GTP-γ-S assay. The two more interesting compounds from the
functional studies were further evaluated in standard antinoci-
ceptive assays in mice, and (1R,5R,9S)-(-)-10 was found to be
over 470 times more potent than morphine in the hot plate and
tail flick assays and 200-fold more potent in the PPQ assay,
the most potent known phenylmorphan in these assays. In the
single-dose suppression assay in the rhesus monkey, this
compound completely suppressed morphine abstinence at doses
of 0.005 and 0.03 mg/kg and was more than 1000 times more
potent than morphine. The (1R,5S)-(+)-14 analogue was over
50 times more potent than morphine in the hot plate and tail
flick and 17-fold more potent in the PPQ assay in mice. The
potency of (1R,5S)-(+)-14 as an antinociceptive was cor-
roborated by a 55 °C warm water tail-flick test in mice. The
compound produced dose- and time-related antinociceptive
effects in this assay following subcutaneous administration. Peak
antinociception occurred between 30 and 45 min postinjection,
and effects were present at 180 min following a 1 mg/kg dose
Figure 3. Dose- and time-response curve for subcutaneous (1R,5S)-
(+)-14 in the 55 °C tail-flick test.
(Figure 3). The calculated A50 value (and 95% confidence limits)
for (1R,5S)-(+)-14 was 0.11 (0.09-0.15) mg/kg. For compari-
the C5-C10 bond (see Figure 1 for numbering) due to the
â-oriented methyl substituent at C9. However, compound 1b,
son purposes, s.c. morphine produced an A50 value of 5.95
without a C9 substituent, is also a pure antagonist in the
(5.08-6.95) mg/kg. Thus, the C9â-hydroxy analogue was also
[35S]GTP-γ-S assay, and (1R,5R,9S)-(-)-10, (1R,5S)-(+)-14,
about 50-fold more potent than morphine in this assay. The
and (1R,5S,9R)-(+)-15 with a â-OH, methylene, and R-methyl,
respectively, at C9 are agonists. The latter two compounds are
effects of (1R,5S)-(+)-14 were blocked by pretreatment with
the opioid antagonist naloxone (1 mg/kg, s.c., -10 min, data
not as effective agonists as (1R,5R,9S)-(-)-10 and show potent
not shown). Both (1R,5R,9S)-(-)-10 and (1R,5S)-(+)-14 appear
antagonist activity in the [35S]GTP-γ-S assay (Table 2).
to act as potent µ-agonists in vivo, and neither showed
However, (1R,5S)-(+)-14 does not show any antagonist activity
µ-antagonist activity in the tail flick versus morphine assay. The
in vivo, presumably due to its high intrinsic efficacy (Table 3).
Thus, the transition from agonist to antagonist is very dependent
on the nature and stereochemistry of the substituent at C9 in
potential δ-opioid antagonist profile could not be ascertained
in the mouse antinociceptive assays.
the N-phenethyl-5-hydroxyphenylmorphan series. To probe the
previously assumed determinant for imparting the agonist/
Quantum Chemistry
Agonist versus Antagonist Activity. Rotation of the 5-Hy-
droxyphenyl Ring of (1R,5R,9S)-(-)-10, (1R,5S)-(+)-14, and
(1R,5S,9R)-(+)-15. The antagonist property of 1c was rational-
ized by assuming hindered rotation of the phenolic group around
antagonist property (steric hindrance), the geometries of all the
compounds in Table 1 as well as one of the enantiomers of 1b
(1R,5S) were optimized with density functional theory at the
B3LYP/6-31G* level.

Enantiomeric C9-Substituted Phenylmorphans
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16 3769
Table 3. In Vivo Assaysa
TF vs M
(mg/kg)
SDS
(mg/kg)
cmpd
HP
PPQ
TF
(1R,5R,9S)-(-)-10
0.002
0.002
0.004
inactivec
(1)
CS
(0.001-0.003)b
0.02
(0.001-0.004)b
0.02
(0.003-0.006)c
0.03
(0.006-0.03)d
CS
(1R,5S)-(+)-14
morphine‚SO4
inactivef
(1, 10, 30)
inactive
(1, 10, 30)
(0.01-0.04)e
0.85
(0.39-1.9)
(0.016-0.03)e
0.4
(0.2-0.8)
(0.02-0.04)f,g
1.92
(0.89-4.14)
(0.75-1.0)h
CS
(3.0)
a ED50, mg/kg, sc (95% confidence limits); HP ) hot plate assay, PPQ ) phenylquinone assay, and TF ) tail flick assay, in mice. The TF vs M ) tail
flick vs morphine (µ-antagonist assay) in mice; SDS ) single dose suppression in monkeys, and CS ) complete suppression. Vehicle 5% hydroxypropyl-
â-cyclodextrin in H2O for injection. b Highest dose tested was 0.01 mg/kg. c Straub tail, ataxia, quick onset, increased locomotor activity, and convulsions
were noted at 1 mg/kg. d Potency estimate > 1000× morphine. e Highest dose tested was 0.1 mg/kg. f Straub tail, increased locomotor activity. g Highest
dose tested was 1 mg/kg. h Potency estimate was about 3× morphine.
Figure 5. Superposed ab inito energy minimized structures of (1R,5S)-
1b and (1R,5R,9S)-(-)-10. Atoms represented by colors as follows:
white, hydrogen; green, carbon; blue, nitrogen; red, oxygen.
are separated by an energy barrier of 6.6 kcal/mol and 8.0 kcal/
mol, depending on the direction of the phenol ring rotation. In
terms of thermochemical stability, the rotamer at the dihedral
angle of -19.5° is 0.20 kcal/mol more stable than the rotamer
at 129.5°. Apparently, the presence of the â-oriented (axial to
the piperidine ring) hydroxyl moiety at C9 raises the first
rotational energy barrier of (1R,5R,9S)-(-)-10 by 3.4 kcal/mol,
as compared to that of (1R,5S)-1b, resulting from the steric
repulsion between the C9-OH and the C11-H (or C15-H) of
the phenol (see Figure 1 for numbering). Nonetheless, because
of the small energy barrier and comparable energy, the two
rotamers of (1R,5R,9S)-(-)-10 exist at room temperature and
undergo a rapid interconversion on a sub-microsecond scale.
Figure 4. (a) Potential energy of (1R,5S)-1b vs the dihedral angle of
The rigid fitting of (1R,5S)-1b onto (1R,5R,9S)-(-)-10 using
the nine atoms of the morphan moiety, that is, C1-C9, as a
common docking site gives a 0.04 Å root-mean-square deviation
(rmsd), suggesting essentially no change in the backbone of the
two compounds. Moreover, the spatial orientation of the phenol
of (1R,5R,9S)-(-)-10 at -19.5° is quite comparable to that of
(1R,5S)-1b at -2.7°, as depicted in Figure 5. Even the rotation
of the dihedral angle, C15-C10-C5-C9, of (1R,5R,9S)-(-)-
10 to match that of (1R,5S)-1b (i.e., from -19.5° to -2.7° and
129.5° to 177.4°) requires only a small energy penalty of ∼0.6
kcal/mol and ∼1.0 kcal/mol, respectively. These small energy
differences clearly indicate that all phenolic ring conformations
in (1R,5S)-1b are readily accessible by the phenolic ring of
(1R,5R,9S)-(-)-10, and thus, the difference in the spatial position
of the phenolic ring between (1R,5S)-1b and (1R,5R,9S)-(-)-
C15-C10-C5-C9; (b) potential energy of (1R,5R,9S)-(-)-10 vs the
dihedral angle of C15-C10-C5-C9.
We first varied the dihedral angle of C15-C10-C5-C9 in
increments of 10° while relaxing the rest of the structure of 1b,
(1R,5R,9S)-(-)-10, (1R,5S)-(+)-14, and (1R,5S,9R)-(+)-15 to
assess the effect of the C9 substituent on the rotational energy
barrier of the 5-hydroxyphenyl moiety. The rotation of the
5-hydroxyphenyl of (1R,5S)-1b gives two stable conformers at
the dihedral angle of -2.7° and 177.4°, with the former more
stable than the latter by 0.25 kcal/mol (Figure 4a); these two
rotamers are separated by an energy barrier of 3.2 kcal/mol and
3.5 kcal/mol. The variation of the dihedral angle of C15-C10-
C5-C9 of (1R,5R,9S)-(-)-10 (Figure 4b) from -19.5° to 339.5°
also results in two stable rotamers at -19.5° and 129.5° that

3770 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Hiebel et al.
Table 4. Energetics of the Rotamers Calculated with Density Functional Theory at the B3LYP/6-31G* Level
rotational energy
dihedral angle of C15-C10-C5-C9
barriers of the phenolb
(kcal/mol)
compound
(1R,5S)-1b
(deg) of rotamers
-2.7
-19.5
-58.1
53.0
177.4 (0.25)a
129.5 (0.20)
61.9 (-0.26)
233.0 (0.6)
3.2
3.5
8.0
8.3
7.5
(1R,5R,9S)-(-)-10
6.6
8.7
7.5
(1R,5S)-(+)-14
121.9 (-0.22)
(1R,5S,9R)-(+)-15
a Values in parenthesis represent the relative stability (kcal/mol) with respect to the first rotamer. b With respect to the first rotamer.
10 is not responsible for the antagonist or agonist activity.
Similar energy barriers and energetics were obtained with the
9-methylene and 9R-methyl substitution as reported in Table
4. If we assume that the µ-opioid receptor did not stabilize a
particular rotamer, the quantum chemical studies that we report
here suggest that it is unlikely that ex vivo potency or agonist
versus antagonist behavior of the 9â-methyl compound12 could
be ascribed to a rotamer; these results do not support the
hypothesis that a C9â-methyl moiety in the phenylmorphan
could produce sufficient steric hindrance to effect pharmacologi-
cal activity.
As we previously noted, it is surprising to find C9-substituted
5-phenylmorphans containing an N-phenylethyl moiety that are
µ-agonists. Although compounds with an N-phenylethyl moiety
in the morphinan, 6,7-benzomorphan, and 4,5-epoxymorphinan
series of analgesics are well-known to display increased
antinociceptive potency over the related N-methyl compound,
all of the N-phenylethyl substituted 5-phenylmorphans that have
formerly been reported were pure µ-antagonists. There is no
rationale, that we are aware of, that satisfactorily relates agonist
and antagonist effects to molecular structure in this opioid series.
Agonist Potency. Overlay of (1R,5R,9S)-(-)-10, (1R,5S)-
(+)-14, and (1R,5S,9R)-(+)-15. It is possible to rationalize the
high agonist potency of (1R,5R,9S)-(-)-10 if we hypothesize
that the agonist activity of this series of compounds might be
correlated with a proton transfer of the charged nitrogen atom
via a channeled water molecule(s) to an appropriate amino acid
in the µ-opioid receptor. If that occurred, a subsequent confor-
mational change in the receptor induced by the proton transfer
might initiate the signaling process that culminates in agonist
activity.
The morphan moiety of (1R,5R,9S)-(-)-10 fits well to the
corresponding atoms of the geometry optimized structure of
morphine (rmsd <0.14 Å), whereas its enantiomer, (1S,5S,9R)-
(+)-10, which was found to have 60-fold less affinity for the
µ-receptor, does not readily fit. A similar trend is also observed
for (1R,5S)-(+)-14 and (1R,5S,9R)-(+)-15, as shown in Table
1. Among (1R,5R,9S)-(-)-10, (1R,5S)-(+)-14, and (1R,5S,9R)-
(+)-15, the stereochemistry of the morphan is identical except
at C9, suggesting that the differences in their binding affinities
and pharmacological responses arise from the chemical nature
of the hydroxyl, methylene, and methyl substituent as well as
from their spatial orientation at C9. To probe this possibility,
the geometry optimized structures of (1R,5S)-(+)-14 and
(1R,5S,9R)-(+)-15 were overlaid onto (1R,5R,9S)-(-)-10 using
the morphan atoms as a common docking site as shown in
Figure 6a. The rmsd of the fitting of (+)-14 and (+)-15 onto
(-)-10 is 0.04 Å and 0.05 Å, respectively. The C9-hydroxyl
oxygen, methylene carbon, and methyl carbon are positioned
2.6 Å, 3.4 Å and 3.9 Å, respectively, away from their respective
protonated nitrogen. We observed that the C9â-hydroxyl in
(1R,5R,9S)-(-)-10 is especially well-positioned to form a water-
mediated H-bond with the protonated nitrogen. Insertion of a
water molecule between the C9â-hydroxyl and the nitrogen atom
and subsequent geometry optimization (Figure 6b) suggests that
Figure 6. (a) Superposed structures of (1R,5R,9S)-(-)-10, (1R,5S)-
(+)-14, and (1R,5S,9R)-(+)-15. Dotted lines indicate the distance
between the heavy atoms of the C9-substituents and the protonated
tertiary nitrogen atom. The C9-hydroxyl oxygen in (1R,5R,9S)-(-)-
10, methylene carbon in (1R,5S)-(+)-14, and methyl carbon in
(1R,5S,9R)-(+)-15 are positioned 2.6 Å, 3.4 Å, and 3.9 Å, respectively,
away from the protonated nitrogen. (b) H-bonding interaction of a
putative water molecule with the 9â-OH and the protonated tertiary
nitrogen of (1R,5R,9S)-(-)-10.
a water molecule can be part of a seven-membered ring involved
in H-bonding with the OH and the tertiary nitrogen. With the
addition of a second water molecule, a short H-bonded water

Enantiomeric C9-Substituted Phenylmorphans
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16 3771
chain can form, which in turn effectively polarizes the hydrogen
of the tertiary nitrogen. This observation has led us to
hypothesize that, in µ-opioid receptors, a putative H-bonded
water chain pulls the proton from the protonated tertiary amine
and then facilitates a proton hopping along the chain, a process
known as a Grotthuss mechanism,21 to a nearby proton acceptor,
possibly Asp, a well-conserved residue among the opioid
G-protein coupled receptors (GPCR). Proton transfer via a
H-bonded water chain inside proteins is a well-known process.22
A change in the local protonation state is also known to cause
substantial conformational changes in proteins such as cyto-
chrome oxidase,23 bacteriorhodopsin,24 and green fluorescent
protein.25 Accordingly, analogous conformational changes in the
µ-opioid receptor caused by the proton transfer may provide
the signal that induces µ-agonist activity through the opioid
receptor.
The possibility that proton transfer influences the agonist
activity of an opioid was first raised by Belleau et al.26 and
Kolb and Scheiner.27 They envisioned a direct proton transfer
from the nitrogen atom. While this kind of direct transfer is
feasible, structural evidence from recent X-ray diffraction studies
support proton transfer via an aligned H-bonded water chain.
Examples of proteins with an interior water chain includes
bacteriorhodopsin,28 a GPCR, cytochrome-c oxidase,23 green
fluorescent protein,25 and gramicidin.29 In particular, the X-ray
structure of bacteriorhodopsin of the N′ state28 has shown a
H-bonded water chain connecting the proton donor Asp-85 to
the Schiff base nitrogen of retinal, and quantum chemical
dynamics simulations have indicated that this water chain
facilitates the proton hopping during the transition from the late
M to the N state.30 In addition, the stability induced by
H-bonding through a water molecule(s), as shown in Figure 6b,
lends further support to the premise that the agonist activity of
(1R,5R,9S)-(-)-10 is enhanced because of the presence of the
9-hydroxyl moiety in (-)-10, which may promote the alignment
of water molecule(s).
The hydrophobic methylene and methyl substituents at C9
may also aid the alignment of water molecule(s) by restricting
the movement of putative water molecules at the opioid binding
site. This proposition is supported by recent computer simula-
tions that have shown that the confined hydrophobic environ-
ment of the interior of both bacteriorhodopsin30 and carbon
nanotubes31,32 favors the alignment of a H-bonded water chain.
The putative water chain induced by the C9-methylene of
(1R,5S)-(+)-14 and the C9-methyl of (1R,5S,9R)-(+)-15, how-
ever, may not be quite as well aligned as in (1R,5R,9S)-(-)-
10, making the proton transfer somewhat less favorable because
these C9 moieties are spatially further from the N-atom. This
may give rise to the more modest agonist behavior of (1R,5S)-
(+)-14. The 9R-methyl of (1R,5S,9R)-(+)-15 is even further
from the tertiary nitrogen (3.9 Å vs 2.6 Å and 3.4 Å for the
related C9â-hydroxy and the methylene compounds, respec-
tively) and, thus, a weaker µ-agonist than the other two
compounds in the [35S]GTP-γ-S assay and in vivo (data not
shown). Nevertheless, the 9R-methyl still seems to be fairly
effective in aligning water molecules and the compound has
some activity as an agonist, unlike 1b that was a pure antagonist
ex vivo.4,9 The distance of these moieties from the nitrogen atom
appear, at least in this limited sample of three compounds, to
be compatible with the alignment of the water chain and,
perhaps, with their in vivo activity.
the µ-opioid receptor, and two of these produced extremely
potent antinociceptive effects in standard assays of nociception.
One of them, (1R,5S)-(+)-14, exhibited a profile consistent with
a mixed-acting µ-agonist and a µ-,δ- and κ-antagonist ex vivo.
That compound, however, was found to have morphine-like
actions in the single dose suppression assay in monkeys and,
in an antinociceptive assay, showed typical signs of morphine-
like agonist activity (Straub tail and stereotypic circling
behavior). It did not show any antagonist activity in vivo.
Quantum chemical calculations did not provide any assurance
that the change in types of activities (agonist vs antagonist) could
be ascribed to the presence or absence of steric hindrance caused
by a C9-substituent. The question of what distinguishes opioid
agonists and antagonists at the molecular level remains intrigu-
ing. Although agonist activity for an N-phenylethyl-substituted
phenylmorphan was unexpected, our hypothesis of the facilita-
tion of aligned water molecules by a properly positioned
9-hydroxy substituent and a protonated nitrogen atom enabling
water-assisted proton transfer from the ligand to an amino acid
in the µ-receptors’ binding pocket, provides a possible explana-
tion of the remarkable µ-opioid agonist potency found for the
9â-hydroxy compound. This hypothesis is being further exam-
ined, and the results that are obtained will be reported in the
future.
Experimental Section
The experimental procedures are written for the (1S,5S)-
enantiomers. The (1R,5R)-enantiomers were prepared following the
same conditions. All reactions were performed in oven-dried
glassware under an argon atmosphere and magnetically stirred,
unless otherwise noted. All melting points were determined on a
Thomas-Hoover melting point apparatus and are uncorrected. Radial
preparative layer chromatography (radial PLC) was performed on
a Chromatotron (Harrison Associates, Palo Alto, CA) using glass
plates coated with 1, 2, or 4 mm layers of Kieselgel 60 PF254
containing gypsum. Proton (1H NMR in CDCl3, unless otherwise
stated, with TMS at δ 0.0 ppm) and carbon (13C NMR with CDCl3
at δ 77.23 ppm) spectra were recorded on a Varian XL-300
instrument at 300 MHz for 1H NMR and 75 MHz for 13C NMR. A
temperature of 55 °C was used when needed for rotamer coales-
cence. Mass spectra (ES) and high-resolution mass spectra (HRMS)
were obtained using a JEOL SX102. Thin layer chromatography
(TLC) was performed on 250 mm Analtech GHLF silica gel plates
using various gradients of CHCl3/MeOH containing 1% NH4OH
or gradients of ethyl acetate/n-hexanes. Visualization was ac-
complished under UV or by staining in an iodine chamber.
Elemental analyses were performed by Atlantic Microlabs, Inc.,
Norcross, GA.
Quantum Chemical Methods. Geometry optimization and
energetic calculations for all the compounds in Table 1 have been
conducted in the gaseous phase with the density functional theory
at the level of B3LYP/6-31G*.33 No thermal energy and solvents
effects were taken into account in calculating rotational energy
barriers. Superposition of these geometry-optimized structures was
carried out using the rigid-fit of Quanta 2000 (Accelrys). This fit
method minimizes the root-mean-square deviation between refer-
ence and working compounds as a function of both translational
and rotational (or orientational) degrees of freedom.
Antinociceptive (Hot Plate, Tail-Flick, Phenylquinone) Assays
in Mice and Single-Dose Suppression Assay in Monkeys. These
assays were run as previously noted.34
Warm Water Tail-Flick Test. Antinociception was assessed
at 55 °C. The latency to the first sign of a rapid tail-flick was taken
as the behavioral endpoint. Each mouse was first tested for baseline
latency by immersing its tail in the water and recording the time to
response. Mice not responding within 5 s were excluded from
further testing. Mice were then administered the test compound
and tested for antinociception at 10, 20, 30, 45, 60, 90, 120, and
Conclusion
The synthesis of the enantiomers of the various C9-substituted
phenylmorphans yielded compounds with very high affinity for

3772 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Hiebel et al.
180 min postinjection. A cutoff of 10 s was used to avoid tissue
damage. Antinociception was calculated by the following for-
mula: % antinociception ) 100 × (test latency-control latency)/
(10-control latency). Dose-response lines were constructed at times
of peak agonist effect and analyzed by linear regression using
FlashCalc software (Dr. Michael Ossipov, University of Arizona,
Tucson, AZ). All A50 values (95% confidence limits) shown are
calculated from the linear portion of the dose-response curve. A
minimum of three doses/curve and 10 mice/dose were used.
Binding and Efficacy Assays. Cell Culture and Membrane
Preparation. The recombinant CHO cells (hMOR-CHO, hDOR-
CHO, and hKOR-CHO) were produced by stable transfection with
the respective human opioid receptor cDNA and were provided by
Dr. Larry Toll (SRI International, CA). The cells were grown on
plastic flasks in DMEM (100%; hDOR-CHO and hKOR-CHO) or
DMEM/F-12 (50%/50%) medium (hMOR-CHO) containing 10%
FBS and G-418 (0.10-0.2 mg/mL) under 95% air/5% CO2 at
37 °C. Cell monolayers were harvested and frozen at -80 °C.
[35S]GTP-γ-S Binding Assays. On the day of the assay, cells
were thawed on ice for 15 min and homogenized using a polytron
in 50 mM Tris-HCl, pH 7.4, containing 4 µg/mL leupeptin, 2 µg/
mL chymostatin, 10 µg/mL bestatin, and 100 µg/mL bacitracin.
The homogenate was centrifuged at 30 000 × g for 10 min at
4 °C, and the supernatant was discarded. The membrane pellets
were resuspended in binding buffer and were used for [35S]GTP-
γ-S binding assays. [35S]GTP-γ-S binding was determined as
described previously.35 Briefly, test tubes received the following
additions: 50 µL buffer A (50 mM Tris-HCl, pH 7.4, containing
100 mM NaCl, 10 mM MgCl2, 1 mM EDTA), 50 µL GDP in buffer
A/0.1% BSA (final concentration ) 10 µM), 50 µL drug in buffer
A/0.1% BSA, 50 µL [35S]GTP-γ-S in buffer A/0.1% BSA (final
concentration ) 50 pM), and 300 µL of cell membranes (50 µg of
protein) in buffer B. The final concentrations of reagents in the
[35S]GTP-γ-S binding assays were: 50 mM Tris-HCl, pH 7.4,
containing 100 mM NaCl, 10 mM MgCl2, 1 mM EDTA, 1 mM
DTT, 10 µM GDP, and 0.1% BSA. Incubations proceeded for 3 h
at 25 °C. Nonspecific binding was determined using GTP-γ-S (40
µM). Bound and free [35S]-GTP-γ-S were separated by vacuum
filtration through GF/B filters. The filters were punched into 24-
well plates to which was added 0.6 mL LSC-cocktail (Cytoscint).
Samples were counted, after an overnight extraction, in a Trilux
liquid scintillation counter at 27% efficiency.
Opioid Binding Assays. [125I]IOXY (6â-iodo-3,14-dihydroxy-
17-cyclopropylmethyl-4,5R-epoxymorphinan; SA ) 2200 Ci/mmol)
was used to label µ-, δ-, and κ-binding sites. On the day of the
assay, cell pellets were thawed on ice for 15 min then homogenized
with a polytron in 10 mL/pellet of ice-cold 10 mM Tris-HCl, pH
7.4. Membranes were then centrifuged at 30 000 × g for 10 min,
resuspended in 10 mL/pellet ice-cold 10 mM Tris-HCl, pH 7.4,
and again centrifuged 30 000 × g for 10 min. Membranes were
then resuspended in 25 °C 50 mM Tris-HCl, pH 7.4 (∼100 mL/
pellet hMOR-CHO, 50 mL/pellet hDOR-CHO and 120 mL/pellet/
hKOR-CHO). All assays took place in 50 mM Tris-HCl, pH 7.4,
with a protease inhibitor cocktail [bacitracin (100 µg/mL), bestatin
(10 µg/mL), leupeptin (4 µg/mL), and chymostatin (2 µg/mL)], in
a final assay volume of 1.0 mL. Nonspecific binding was
determined using 10 µM naloxone. Triplicate samples were filtered
with Brandell Cell Harvesters (Biomedical Research & Develop-
ment, Inc., Gaithersburg, MD) over Whatman GF/B filters after a
2 h incubation at 25 °C. The filters were punched into 12 × 75
mm glass test tubes and counted in a Micromedic gamma counter
at 80% efficiency. Opioid binding assays had ∼30 µg protein per
assay tube. Inhibition curves were generated by displacing a single
concentration of radioligand by 10 concentrations of drug.
[35S]-GTP-γ-S binding) were determined using either the program
MLAB-PC (Civilized Software, Bethesda, MD) or KaleidaGraph
(version 3.6.4, Synergy Software, Reading, PA). For functional Ki
determinations, except where otherwise indicated, we determined
the ability of test agents to inhibit DAMGO (1 µM)-, SNC80 (500
nM)- or (-)-U50,488 (500 nM)-stimulated [35S]-GTP-γ-S binding,
using the appropriate cell line. The data of at least three experiments
(10 points each) were pooled, and the data were fit to the two-
parameter logistic equation for the best-fit estimates of the IC50
and N values: Y ) 100/(1 + ([INHIBITOR]/IC50)N), where Y is
the percent of control value. The functional Ki values were then
calculated from the IC50 values according to the equation: Ki )
IC50/(1 + [D]/ED50(D), where D is the concentration of stimulation
agent and ED50(D) is the ED50 of that agent for stimulating [35S]-
GTP-γ-S binding. For opioid binding experiments, the data were
fit to the two-parameter logistic equation for the best-fit estimates
of the IC50 and N values.
Sources. [35S]GTP-γ-S (SA ) 1250 Ci/mmol) was obtained from
DuPont NEN (Boston, MA). Various opioid peptides were provided
by Multiple Peptide System via the Research Technology Branch,
NIDA. [125I]IOXY was prepared as described37,38 using pre-
cursor (3-acetoxy-6R-[[(trifluoromethyl)sulfonyl]oxy]-l4-hydroxy-
17-(cyclopropylmethyl)-4,5R-epoxymorphinan) synthesized in the
laboratory of Dr. Thomas Prisinzano (University of Iowa, Iowa City,
IA). The sources of other agents are published.36
Optical Resolution of (()-5-(3-Methoxyphenyl)-2-methyl-
9-oxo-2-azabicyclo[3.3.1]nonane ((()-2). The racemate (15.85 g,
(()-2) was diluted in 100 mL of 60% acetone/40% methanol and
l-tartaric acid (9.15 g) was dissolved in 75 mL of 60% acetone/
40% methanol. The optimal target concentration of the salt was 1
g/7 mL. Both solutions were heated to reflux prior to mixing. The
l-tartaric acid solution was quickly added to the base, and the
resulting mixture was vigorously mixed. Snow flake-like crystals
appeared almost instantaneously. After cooling, the solid was
filtered (10.43 g, 41.7%), and a small portion was free-based and
analyzed by NMR with (R)-(-)-1-phenyl-2,2,2-trifluoroethanol as
chiral shift reagent (ee > 85%).13 The salt was recrystallized from
84 mL of 90% methanol/water to yield 5.26 g (33%) of enantio-
merically pure (1S,5S)-(-)-2 after free-basing (>99% by NMR):
[R]20D -34.5° (c 1.34, CHCl3). The initial filtrate and mother liquors
were recovered, evaporated, and free-based (9.23 g, 58%). The
absolute stereochemistry of (1S,5S)-(-)-2 was established by X-ray
analysis of the L-tartrate salt. The residual material from the
preparation of the levo enantiomer, which was enriched with
(1R,5R)-(+)-2, was diluted in 61 mL of 60% acetone/40% methanol.
D-Tartaric acid (5.33 g) was dissolved in 40 mL of 60% acetone/
40% methanol. Both solutions were heated to reflux prior to mixing.
The D-tartaric acid solution was quickly added to the base, and the
resulting mixture was vigorously mixed. Snow flake-like crystals
appeared almost instantaneously. After cooling, the solid was
filtered (12.2 g, 49%, ee > 70%). After evaporation, the filtrate
was worked up to obtain an additional 2.19 g (9%). The salt was
recrystallized from 90% methanol/10% water (1 g/8 mL) to obtain
enantiomerically pure (1R,5R)-(+)-2: [R]20D +35.5° (c 1.2, CHCl3)
after conversion to the free-base.
(1S,5S)-(-)-Ethyl-5-(3-methoxyphenyl)-9-oxo-2-azabicyclo-
[3.3.1]nonane-2-carboxylate, (1S,5S)-(-)-3). The free-base oil,
(1S,5S)-(-)-2 (8.72 g, 33.6 mmol), was diluted in 67 mL of
anhydrous dichloroethane. To the reaction mixture was added K2-
CO3 (2.21 g, 168 mmol) followed by ethylchloroformate (16.05
mL, 168 mmol). The resulting mixture was refluxed overnight. After
that time, some remaining starting material was present and more
K2CO3 (882 mg, 67.2 mmol) and ethylchloroformate (3.21 mL,
33.6 mmol) were added. The reaction was refluxed for an additional
7 h. Water was then added, and the basicity was adjusted to pH 9
with 28% NH4OH. The aqueous layer was extracted with CHCl3,
and the combined organic extracts were washed with brine and
dried over Na2SO4. The crude mixture was evaporated in vacuo
and purified by flash chromatography (silica gel, 2% MeOH/CH2-
Cl2) to afford (1S,5S)-(-)-3 as a white solid (9.32 g, 87%) and
some recovered starting material (820 mg, 9%). Mp 92-93 °C; 1H
Data Analysis and Statistics. For [35S]GTP-γ-S binding experi-
ments, the percent stimulation of [35S]GTP-γ-S binding was
calculated according to the following formula: (S - B)/B × 100,
where B is the basal level of [35S]GTP-γ-S binding and S is the
stimulated level of [35S]GTP-γ-S binding.36 EC50 values (the
concentration that produces fifty percent maximal stimulation of
[35S]GTP-γ-S binding) and Emax (% of maximal stimulation in the

Enantiomeric C9-Substituted Phenylmorphans
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16 3773
NMR (55 °C) δ 7.29-7.23 (m, 1H), 6.81-6.78 (m, 3H), 4.39
(broad s, 1H), 4.19 (q, 3H, J ) 7.2 Hz), 3.79 (s, 3H), 3.30 (ddd,
1H, J ) 5.4, 10.5 and 13.8 Hz), 2.6 (broad s, 1H), 2.51 (ddd,
1H, J ) 5.6, 10.5 and 14.3 Hz), 2.39-2.31 (m, 2H), 2.26-2.15
(m, 2H), 1.86-1.71 (m, 2H), 1.28 (t, 3H, J ) 7.2 Hz); 13C NMR
(55 °C) δ 211.1, 159.8, 145.7, 129.3, 119.7, 114.0, 111.9,
63.6, 61.8, 55.5, 53.2, 42.0, 41.3, 38.6, 18.2, 14.9; HRMS
(ES+) calcd for C18H23NO4 (M + H)+, 318.1705; found, 318.1692.
3.29-3.21 (m, 1H), 2.95-2.73 (m, 5H), 2.50-2.33 (m, 4H), 2.25-
2.10 (m, 2H), 1.82-1.65 (m, 2H); 13C NMR (55 °C) δ 215.2, 155.6,
146.2, 140.2, 129.3, 128.9, 128.6, 126.3, 119.3, 115.0, 114.0, 68.9,
58.9, 53.0, 49.0, 40.9, 39.4, 34.6, 34.1, 19.3; HRMS (ES+) calcd
for C22H25NO2, 336.1964 (M + H)+; found, 336.1963; [R]20
D
-31.9° (c 1.115, CHCl3). Anal. (C22H26BrNO2): C, H, N. For
(1R,5R)-(+)-6: [R]20 +32.5° (c 0.965, CHCl3). HBr salt: mp
D
199-202 °C. Anal. (C22H26BrNO2‚0.5H2O): C, H, N.
[R]20 -40.6° (c 1.25, CHCl3). Anal. (C18H23NO4): C, H, N. For
(1S,5S,9S)-(-)-9-Hydroxy-5-(3-methoxyphenyl-2-phenylethyl-
2-azabicyclo[3.3.1]nonane ((1S,5S,9S)-(-)-7). Methanol (5 mL)
was placed in a flask. To this was added thionyl chloride (0.38
mL, 1.4 mmol, 2 M solution in CH2Cl2), and the resulting mixture
was stirred for 30 min at room temperature and then cooled to
0 °C. A solution of the phenolic amine (1S,5S)-(-)-5 (400 mg,
1.14 mmol) in 2 mL of methanol was then added, and the salt was
allowed to form for 30 min at 0 °C. Sodium borohydride (180 mg,
4.8 mmol) was then slowly added, and the resulting mixture was
warmed to room temperature. The reaction was quenched with a
saturated solution of NaHCO3, and the aqueous layer was extracted
with Et2O. The combined organic extracts were washed with brine
and dried over Na2SO4. The residue was evaporated and purified
by radial PLC (silica, gradient CH2Cl2 to 5% MeOH/CH2Cl2) to
afford the R-alcohol (1S,5S,9S)-(-)-7 (361 mg, 90%) as a white
solid: mp 97-98 °C; 1H NMR δ 7.32-7.17 (m, 6H), 7.06 (d, 1H,
J ) 7.8 Hz), 7.03 (t, 1H, J ) 2.2 Hz), 6.75 (dd, 1H, J ) 2.2 and
8 Hz), 4.30 (d, 1H, J ) 3.6 Hz), 3.78 (s, 3H), 3.15 (s, 1H), 3.0 (td,
1H, J ) 5 and 11.8 Hz), 2.93-2.80 (m, 5H), 2.27-2.06 (m, 2H),
2.00-1.71 (m, 7H); 13C NMR δ 160.0, 149.7, 140.7, 129.7, 128.9,
128.5, 126.2, 118.3, 112.6, 111.2, 73.4, 57.6, 56.6, 55.3, 49.6, 40.8,
39.4, 34.7, 28.9, 22.0, 18.7; HRMS (ES+) calcd for C23H29NO2,
D
(1R,5R)-(+)-3: [R]20 +40.4° (c 1.12, CHCl3).
D
(1S,5S)-(+)-5-(3-Methoxyphenyl)-9-oxo-2-azabicyclo[3.3.1]-
nonane ((1S,5S)-(+)-4). The carbamate (1S,5S)-(-)-3 (10.17 g,
32.04 mmol) was dissolved in 66 mL of anhydrous dichloroethane.
Trimethylsilyliodide (6.81 mL, 48.06 mmol) was added dropwise
at room temperature. The resulting reaction mixture was then
warmed to 55 °C overnight. An additional 1.6 mL (11.3 mmol) of
trimethylsilyliodide was added, and the reaction was further stirred
at 55 °C for 3 h. The reaction was quenched with an aqueous
solution of sodium thiosulfate and basified with 28% NH4OH. The
aqueous layer was extracted with CH2Cl2. The combined organic
extracts were washed with brine and dried with Na2SO4. The crude
mixture was evaporated in vacuo and purified by flash chroma-
tography (silica gel, gradient 2% to 5% MeOH/CH2Cl2) to afford
the product (1S,5S)-(+)-4 as an oil (5.63 g, 72%). Some starting
material was recovered (800 mg, 8%). 1H NMR δ 7.30-7.25
(m, 1H), 6.86-6.78 (m, 3H), 3.80 (s, 3H), 3.62-3.52 (m, 2H),
3.00 (dt, 1H, J ) 6 and 13.8 Hz), 2.54-2.22 (m, 7H), 2.13-2.01
(m, 1H), 1.83-1.74 (m, 1H); 13C NMR δ 216.6, 159.4, 145.9, 129.1,
119.8, 113.9, 111.4, 61.9, 55.4, 53.8, 44.4, 42.1, 40.9, 35.9, 20.0;
HRMS (ES+)) calcd for C15H19NO2 (M + H)+, 246.1494; found,
246.1487; [R]20 +66.2° (c 1.4, CHCl3). For (1R,5R)-(-)-4:
352.2277 (M + H)+; found, 352.2277; [R]20 -37.7° (c 0.87,
D
D
[R]20 -65.4° (c 0.86, CHCl3).
CHCl3). Anal. (C23H29NO2): C, H, N. For (1R,5R,9R)-(+)-7: [R]20
+38.4° (c 1.47, CHCl3). Anal. (C23H29NO2): C, H, N.
D
D
(1S,5S)-(-)-5-(3-Methoxyphenyl)-9-oxo-2-phenylethyl-2-
azabicyclo[3.3.1]nonane ((1S,5S)-(-)-5). The secondary amine
(1S,5S)-(+)-4 (4.29 g, 17.5 mmol) was dissolved in anhydrous
acetonitrile (60 mL), and K2CO3 (7.22 g, 52.5 mmol) and KI (722
mg, 4.35 mmol) were added. 2-Phenylethylbromide (3.62 mL, 26.3
mmol) was then added in one portion, and the resulting reaction
mixture was refluxed for 20 h. An additional 1.2 mL of 2-phenyl-
ethylbromide was added, and the reaction mixture was refluxed
for 4 h to drive the reaction to completion. The solids were filtered,
and the resulting filtrate was concentrated in vacuo. The residue
was purified by flash chromatography (silica gel, hexanes then
gradient CH2Cl2 to 2% MeOH/CH2Cl2) to afford the (1S,5S)-
(1S,5S,9S)-(-)-9-Hydroxy-5-(3-hydroxyphenyl-2-phenylethyl-
2-azabicyclo[3.3.1]nonane ((1S,5S,9S)-(-)-8). The methoxyphenyl
compound (1S,5S,9S)-(-)-7 (400 mg, 1.14 mmol) was dissolved
in 6 mL anhydrous CH2Cl2 and cooled to -78 °C. A solution of
BBr3 (1 M, 5.7 mL, 5.7 mmol) was then added dropwise, and the
reaction was allowed to warm up slowly overnight. The reaction
was cooled back to -78 °C and quenched with H2O and 28% NH4-
OH. The resulting heterogeneous mixture was warmed to room
temperature and vigorously stirred for 1 h, until the CH2Cl2 layer
was clear. The aqueous layer was extracted with CH2Cl2, and the
combined organic extracts were washed with brine and dried over
Na2SO4. The crude was evaporated and purified by radial PLC
(silica, 94.5/5/0.5 CHCl3/MeOH/28% NH4OH) to afford (1S,5S,9S)-
(-)-8 as an off-white foam (367 mg, 96%). The relative stereo-
chemistry of (1S,5S,9S)-(-)-8 was confirmed by X-ray analysis.
1H NMR δ 7.30-7.25 (m, 2H), 7.22-7.11 (m, 4H), 6.97 (s, 1H),
6.90 (d, 1H, J ) 7.5 Hz), 6.68 (dd, 1H, J ) 1.6 and 8 Hz), 4.40 (d,
1H, J ) 3.6 Hz), 3.20 (s, 1H), 3.00-2.96 (m, 2H), 2.85 (d, 4H, J
) 3 Hz), 2.28-2.05 (m, 2H), 1.99-1.75 (m, 6H); 13C NMR δ
156.9, 149.4, 139.9, 129.9, 129.0, 128.7, 126.4, 117.9, 114.5, 113.6,
72.4, 57.3, 27.0, 49.8, 40.3, 38.9, 33.7, 28.6, 21.9, 18.0; HRMS
(ES+) calcd for C22H27NO2, 338.2121 (M + H)+; found, 338.2120;
1
(-)-5 (5.79 g, 95%) as clear crystals: mp 68-70 °C; H NMR δ
7.32-7.18 (m, 6H), 6.85-6.77 (m, 3H), 3.80 (s, 3H), 3.35 (t, 1H,
J ) 3 Hz), 3.28-3.20 (m, 1H), 2.93-2.73 (m, 5H), 2.51-2.35
(m, 4H), 2.22-2.11 (m, 2H), 1.82-1.62 (m, 2H); 13C NMR δ 214.6,
159.4, 146.3, 140.3, 129.1, 128.9, 128.6, 126.3, 119.9, 113.9, 111.4,
69.2, 59.1, 55.4, 53.1, 49.1, 40.8, 39.5, 34.7, 34.4, 19.2; HRMS
(ES+) calcd for C23H27NO2 (M + H)+, 350.2106; found, 350.2120;
[R]20D -29.4° (c 1.40, CHCl3,). Anal. (C23H27NO2): C, H, N. For
(1R,5R)-(+)-5: [R]20 +30° (c 1.18, CHCl3).
D
(1S,5S)-(-)-5-(3-Hydroxyphenyl)-9-oxo-2-phenylethyl-2-
azabicyclo[3.3.1]nonane ((1S,5S)-(-)-6). The methoxyamine (1S,5S)-
(-)-5 (261 mg, 0.75 mmol) was dissolved in 3.4 mL of glacial
acetic acid. A 48% aqueous HBr solution (3.4 mL) was then added,
and the reaction was allowed to reflux overnight. The cloudy
mixture became clear and yellow upon heating. The mixture was
cooled to 0 °C, diluted with H2O, and basified to pH 10 with a
40% NaOH solution. The aqueous layer was extracted with EtOAc/
CHCl3/MeOH (8:1:1). The combined organic extracts were washed
with H2O and brine, dried over Na2SO4, and concentrated in vacuo.
The crude residue was purified by radial PLC (silica, gradient CH2-
Cl2 to 2% MeOH/CH2Cl2) to afford (1S,5S)-(-)-6 (188 mg, 75%)
as a white foam. The HBr salt of (1S,5S)-(-)-6 was formed in Et2O
(1 mL) with 30% HBr in acetic acid (0.12 mL) and recrystallized
from absolute ethanol to give a white solid. The analyses were
performed on the free base, unless otherwise noted. 1H NMR
(free-base, 55 °C) δ 7.32-7.17 (m, 6H), 6.80 (d, 1H, J ) 7.5 Hz),
6.73-6.68 (m, 2H), 5.41 (broad s, 1H), 3.36 (t, 1H, J ) 3 Hz),
[R]20 -23.6° (c 1.51, CHCl3). Anal. (C22H27NO2‚0.4 H2O): C,
D
H, N. For (1R,5R,9R)-(+)-8: [R]20D +24.1° (c 1.7, CHCl3). Anal.
(C22H27NO2‚0.3 H2O): C, H, N.
(1S,5S,9R)-(+)-9-Hydroxy-5-(3-methxyphenyl-2-phenylethyl-
2-azabicyclo[3.3.1]nonane ((1S,5S,9R)-(+)-9). The 9-keto com-
pound (1S,5S)-(-)-5 (200 mg, 0.57 mmol) was dissolved in 5 mL
of anhydrous THF and cooled to -78 °C. A solution of lithium
triethyl borohydride (superhydride; 1 M, 0.86 mL, 0.86 mmol) was
added dropwise, and the resulting mixture was stirred for 45 min
at -78 °C. The reaction was quenched with H2O, and the basicity
was adjusted to pH 9 with 28% NH4OH. The aqueous layer was
extracted with Et2O. The combined organic extracts were washed
with brine and dried over Na2SO4. The residue was evaporated and
purified by radial PLC (silica, gradient CH2Cl2 to 2% MeOH/CH2-
Cl2) to afford the â-alcohol (1S,5S,9R)-(+)-9 (195 mg, 97%) as a
thick oil. 1H NMR δ 7.30-7.17 (m, 6H), 7.00 (d, 1H, J ) 8.1 Hz),

3774 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Hiebel et al.
6.96-6.94 (m, 1H), 6.72 (dd, 1H, J ) 2.1 and 8.1 Hz), 4.05
(s, 1H), 3.79 (s, 3H), 3.50 (broad s, 1H), 3.11-2.91 (m, 2H), 2.88-
2.77 (m, 4H), 2.35-2.24 (m, 2H), 2.04 (dd, 1H, J ) 5.1 and 13.8
Hz), 1.92-1.84 (m, 2H), 1.69-1.50 (m, 4H); 13C NMR δ 159.8,
151.0, 140.5, 129.3, 128.8, 128.6, 126.3, 118.1, 112.4, 110.6, 71.7,
58.9, 56.4, 55.3, 48.3, 41.1, 40.9, 34.5, 30.0, 25.0, 22.9; HRMS
(ES+) calcd for C23H29NO2, 352.2277 (M + H)+; found, 352.2261;
[R]20 +55.5° (c 1.26, CHCl3,). For (1R,5R,9S)-(-)-9: [R]20
Alternate method: The starting ketone (1S,5S)-(-)-5 (2 g, 5.7
mmol) was diluted in 30 mL of anhydrous THF and cooled to 0
°C. Tebbe’s reagent18,19 (11.4 mL, 5.7 mmol, 0.5 M solution in
toluene) was slowly added. The resulting mixture was stirred at 0
°C for 1 h and then slowly warmed up to room temperature for an
additional 2 h. The reaction was cooled to 0 °C and 50 mL of Et2O
was added. The reaction was quenched carefully with 1.8 N NaOH.
A very vigorous gas evolution took place and a thick green
precipitate formed. Magnesium sulfate was added, and the mixture
was allowed to stir an additional 5 min. The solids were
filtered and washed with EtOAc. The filtrate was concentrated and
purified by flash chromatography (silica, gradient, 1-5% MeOH/
1%NH4OH/CHCl3) to afford the product (1S,5R)-(-)-12 in 75%
yield (1.48 g) as a pale yellow oil.
D
D
-55.1° (c 1.08, CHCl3).
(1S,5S,9R)-(+)-9-Hydroxy-5-(3-hydroxyphenyl-2-phenylethyl-
2-azabicyclo[3.3.1]nonane ((1S,5S,9R)-(+)-10). The procedure was
identical to the one performed to obtain the C9R-alcohol (1S,5S,9S)-
(-)-8. The crude residue was purified by radial PLC (silica, 94.5/
5/0.5 CHCl3/MeOH/NH3) to yield 90% of the â-alcohol (1S,5S,9R)-
(+)-10 as an off-white foam. The HCl salt was formed with (SOCl2/
MeOH). 1H NMR δ 7.30-7.25 (m, 3H), 7.20-7.11 (m, 3H), 6.91
(d, 1H), 6.85 (s, 1H), 6.57 (dd, 1H, J ) 2.3 and 8.3 Hz), 4.35
(broad s, 1H), 4.06 (d, 1H, J ) 3.9 Hz), 3.10-2.75 (m, 7H), 2.33-
2.22 (m, 2H), 2.02 (dd, 1H, J ) 5 and 13.7 Hz), 1.90-1.83
(m, 3H), 1.67-1.47 (m, 3H); 13C NMR δ 156.3, 150.5, 140.3,
129.4, 128.8, 128.6, 126.3, 117.3, 113.2, 113.0, 71.7, 58.8, 56.4,
48.3, 40.8, 40.7, 34.3, 29.7, 24.8, 22.7; HRMS (ES+) calcd for
C22H27NO2, 338.2120 (M + H)+; found, 338.2112; [R]20D +50.9°
(c 1.18, CHCl3). (1S,5S,9R)-(+)-10‚HCl: mp 138-142 °C.
Anal. (C22H28ClNO2‚0.5H2O): C, H, N. For (1R,5R,9S)-(-)-10:
(1S,5R,9S)-(-)-5-(3-Methoxyphenyl)-9-methyl-2-phenethyl-2-
azabicyclo[3.3.1]nonane ((1S,5R,9S)-(-)-13). The C9-methylene
compound (1S,5R)-(-)-12 (610 mg, 1.73 mmol) was placed in a
Parr shaker vessel along with 9 mL of anhydrous MeOH and
platinum oxide (40 mg, 0.17 mmol). The mixture was shaken for
2 h under a 35 psi atmosphere of hydrogen. The mixture was filtered
over celite. Concentration of the residue and purification by radial
PLC (silica, 2% MeOH/CH2Cl2) afforded the product (1S,5R,9S)-
(-)-13 (566 mg, 1.59 mmol) in 93% yield as a pale yellow oil. 1H
NMR δ 7.33-7.21 (m, 6H), 7.00 (d, 1H, J ) 8 Hz), 6.964-6.959
(m, 1H), 6.74-6.71 (m, 1H), 3.80 (s, 3H), 3.03 (dd, 2H, J ) 3.5
and 9.8 Hz), 2.89-2.80 (m, 5H), 2.47-2.45 (m, 1H), 2.25-2.05
(m, 2H), 1.99-1.90 (m, 3H), 1.85-1.75 (m, 2H), 1.59-1.54
(m, 1H), 0.76 (d, 3H, J ) 7.2 Hz); 13C NMR δ 159.7, 152.2, 141.0,
129.2, 129.0, 128.6, 126.2, 118.3, 112.3, 110.5, 60.1, 58.3, 58.2,
55.4, 50.3, 41.8, 40.1, 38.5, 34.9, 28.5, 22.2, 18.7, 14.4; HRMS
(ES+) calcd for C24H31NO, 350.2484 (M + H)+; found, 350.2474;
[R]20 -38.4° (c 1.43, CHCl3). For (1R,5S,9R)-(+)-13: [R]20
[R]20 -51.8° (c 0.82, CHCl3). Anal. (C22H28ClNO2‚0.6H2O):
D
C, H, N.
(1S,5S)-(-)-9-Hydroxy-5-(3-hydroxyphenyl-2-phenylethyl-9-
trimethylsilylmethyl-2-azabicyclo[3.3.1]nonane ((1S,5S)-(-)-11).
The ketone (1S,5S)-(-)-5 (1.5 g 4.3 mmol) was dissolved in 25
mL of anhydrous THF and cooled to 0 °C. An ethereal solution of
(trimethylsilylmethy)magnesium chloride (17.2 mL, 17.2 mmol)
was added. After 20 min, the ice bath was removed and the mixture
was warmed to room temperature over 1 h, and refluxed overnight.
The reaction was cooled to 0 °C and quenched with dilute NH4-
OH, CHCl3 was added, and the organic phase was separated. The
aqueous phase was filtered to remove the thick magnesium salts
and extracted with CHCl3. The combined organic phases were
washed with brine, dried over Na2SO4, and evaporated under
reduced pressure. The residue was purified by radial PLC (silica,
gradient hexanes/EtOAc 100% to 90% followed by CH2Cl2) to
afford the pure compound (1S,5S)-(-)-11 as a pale yellow oil (1.73
g, 92%). 1H NMR δ 7.43-7.29 (m, 8H), 6.91-6.88 (m, 1H), 3.94
(s, 3H), 3.23-3.18 (m, 2H), 2.96-2.83 (m, 5H), 2.77-2.66
(m, 2H), 2.12-1.55 (m, 8H), 0.52 (d, 1H, J ) 14.7 Hz), 0.00 (s,
9H); 13C NMR δ 159.5, 149.0, 141.2, 129.0, 128.8, 128.5, 126.1,
120.3, 114.7, 110.8, 77.1, 62.0, 57.8, 55.4, 48.7, 45.1, 34.8, 33.4,
24.1, 22.3, 19.4, 0.8; HRMS (ES+) calcd for C27H39NO2Si, 438.2828
(M + H)+; found, 438.2824; [R]20D -17.0° (c 1.575, CHCl3). For
D
D
+39.5° (c 1.31, CHCl3).
(1S,5R)-(-)-5-(3-Hydroxyphenyl)-9-methylene-2-phenethyl-
2-azabicyclo[3.3.1]nonane ((1S,5R)-(-)-14). The methoxyphenyl
compound (1S,5R)-(-)-12 was submitted to the same reaction
conditions that were used to obtain the phenolic alcohol (1S,5S,9S)-
(-)-8. Purification of the crude residue by radial PLC (silica,
gradient 2% to 5% MeOH/CH2Cl2) gave the free phenol in 96%
yield. It was converted to the oxalate salt: light tan solid; mp 113-
115 °C. Oxalate: 1H NMR (CD3OD) δ 7.40-7.29 (m, 5H), 7.21
(t, 1H, J ) 8 Hz), 6.89-6.84 (m, 2H), 6.74-6.71 (m, 1H), 5.28
(s, 1H), 4.66 (s, 1H), 4.31 (s, 1H), 3.83 (m, 1H), 3.51-3.45 (m,
3H), 3.16-3.10 (m, 2H), 2.57 (m, 1H), 2.42-2.38 (m, 1H), 2.24-
1.95 (m, 6H); 13C NMR δ 156.1, 152.9, 149.7, 140.3, 129.1, 128.94,
128.90, 128.6, 126.3, 119.3, 115.4, 113.7, 109.6, 63.4, 58.6, 48.7,
43.9, 39.0, 37.5, 34.2, 31.0, 21.0; RD -27.6° (c 0.79, MeOH). Anal.
(C25H29NO5‚0.1H2O): C, H, N. HRMS of the free base (ES+) calcd
for C23H28NO, 334.2171 (M + H)+; found, 334.2171. For (1R,5S)-
(+)-14 oxalate: [R]20D +26.5° (c 0.79, MeOH). Anal. (C25H29NO5‚
0.1H2O): C, H, N.
(1R,5R)-(+)-11: [R]20 +17.2° (c 1.325, CHCl3).
D
(1S,5R)-(-)-5-(3-Methoxyphenyl)-9-methylene-2-phenethyl-
2-azabicyclo[3.3.1]nonane ((1S,5R)-(-)-12). The tertiary alcohol
(1S,5S)-(-)-11 (1.73 g, 3.96 mmol) was diluted in anhydrous THF.
Sodium hydride (632 mg, 15.8 mmol) was added, and the mixture
was allowed to reflux overnight. The reaction was cooled to 0 °C
and stopped by the addition of H2O and 28% NH4OH. The aqueous
phase was extracted with CH2Cl2, dried over Na2SO4, and concen-
trated in vacuo. The residue was purified by radial PLC (silica,
gradient CH2Cl2, 5% MeOH/CH2Cl2) to afford (1S,5R)-(-)-12 as
(1S,5R,9S)-(-)-5-(3-Hydroxyphenyl)-9-methyl-2-phenethyl-2-
azabicyclo[3.3.1]nonane ((1S,5R,9S)-(-)-15). The methoxyphenyl
compound (1S,5R,9S)-(-)-13 was submitted to the same reaction
conditions used to obtain the phenol (1S,5S,9S)-(-)-8. After
purification by radial PLC (silica, gradient 2% to 5% MeOH/CH2-
Cl2), the product (1S,5R,9S)-(-)-15 was isolated in 72% yield as a
1
light tan solid, mp 173-174 °C. H NMR δ 7.33-7.15 (m, 6H),
6.96 (d, 1H, J ) 7.8 Hz), 6.87 (s, 1H), 6.66 (dd, 1H, J ) 2.4 and
7.8 Hz), 3.06-3.01 (m, 2H), 2.92-2.79 (m, 5H), 2.48 (s, 1H),
2.25-2.05 (m, 2H), 1.95-1.74 (m, 5H), 1.59-1.57 (m, 2H); 13C
NMR δ 156.9, 151.3, 139.8, 129.4, 128.9, 128.7, 126.4, 117.4,
113.8, 113.7, 58.3, 58.7, 50.3, 40.6, 39.0, 38.0, 33.5, 28.0, 21.8,
18.1, 14.1; HRMS (ES+) calcd for C23H30NO, 336.2327 (M + H)+;
1
a pale yellow oil in 98% yield (1.34 g, 3.56 mmol). H NMR
(CDCl3, 300 MHz) δ 7.32-7.20 (m, 6H), 6.98 (d, 1H, J ) 8.4
Hz), 6.95-6.94 (m, 1H), 6.76 (dd, 1H, J ) 2.4 and 8.4 Hz), 4.83
(s, 1H), 4.20 (s, 1H), 3.81 (s, 3H), 3.43 (m, 1H), 3.11-3.03 (m,
1H), 2.87-2.75 (m, 5H), 2.37-2.28 (m, 1H), 2.24-2.00 (m, 5H),
1.71-1.67 (m, 1H), 1.59-1.48 (m, 1H); 13C NMR (CDCl3, 75
MHz) δ 159.3, 154.3, 150.3, 140.9, 129.0, 128.8, 128.6, 126.2,
120.3, 114.2, 110.7, 108.9, 63.9, 59.2, 55.4, 49.0, 44.2, 39.1, 38.3,
34.8, 32.4, 20.4; HRMS (ES+) calcd for C24H29NO, 348.2327
found, 336.2326; [R]20 -36.1° (c 0.79, CHCl3). Anal. (C23H29-
D
NO‚0.1H2O): C, H, N. For (1R,5S,9R)-(+)-15: [R]20 +35.9°
D
(c 0.79, CHCl3). Anal. (C23H29NO‚0.1H2O): C, H, N. The relative
stereochemistry of (1R,5S,9R)-(+)-15 was confirmed by X-ray
analysis.
(M + H)+; found, 348.2325; [R]20 -58.9 (c 1.87, CHCl3). For
X-ray Crystal Structure of (1S,5S)-(-)-2, (1S,5S,9S)-(-)-8,
and (1R,5S,9R)-(+)-15. Single-crystal X-ray diffraction data were
D
(1R,5S)-(+)-12: [R]20 +59.5 (c 1.20, CHCl3).
D

Enantiomeric C9-Substituted Phenylmorphans
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16 3775
collected at 93 °K using MoK radiation and a Bruker SMART 1000
CCD area detector. Crystals were prepared for data collection by
coating with high viscosity microscope oil (Paratone-N, Hampton
Research). Corrections were applied for Lorentz, polarization, and
absorption effects. The structures were solved by direct methods
and refined by full-matrix least squares on F2 values using the
programs found in the SHELXTL suite.39 Parameters refined
included atomic coordinates and anisotropic thermal parameters
for all nonhydrogen atoms. Hydrogen atoms on carbons were
included using a riding model (coordinate shifts of C applied to H
atoms), with the C-H distance set at 0.96 Å. The absolute
configuration was set based on the known configuration of
the starting materials for the synthesis or a cocrystallized chiral
salt.
(1S,5S)-(-)-2: An oil-coated 0.52 × 0.29 × 0.19 mm3 crystal
of (1S,5S)-(-)-2 was mounted on a glass rod and transferred
immediately to the cold stream (-180 °C) on the diffractometer.
The crystal was orthorhombic in space group P21, with unit cell
dimensions a ) 6.0987(19) Å, b ) 6.907(2) Å, and c ) 22.526(7)
Å. Data were 95.6% complete to 27.98° θ (approximately 0.75 Å).
The asymmetric unit contains one molecule plus one molecule of
L-tartaric acid.
(1S,5S,9S)-(-)-8: An oil-coated 0.68 × 0.28 × 0.19 mm3 crystal
of (1S,5S,9S)-(-)-8 was mounted on a glass rod and transferred
immediately to the cold stream (-180 °C) on the diffractometer.
The crystal was orthorhombic in space group P212121 with unit
cell dimensions a ) 7.1948(9) Å, b ) 10.3050(14) Å, and c )
24.896(4) Å. Data were 97.0% complete to 28.35° θ (approximately
0.75 Å). The asymmetric unit contains a single molecule.
(1R,5S,9R)-(+)-15: An oil-coated 0.48 × 0.21 × 0.16 mm3
crystal of (1R,5S,9R)-(+)-15 was mounted on a glass rod and
transferred immediately to the cold stream (-180 °C) on the
diffractometer. The crystal was orthorhombic in space group P212121
with unit cell dimensions a ) 8.3455(13) Å, b ) 10.9441(18) Å,
and c ) 19.225(3) Å. Data were 94.3% complete to 28.33° θ
(approximately 0.75 Å). The asymmetric unit contains a single
molecule.
the data can be obtained, free of charge, on application to CCDC,
12 Union Road, Cambridge, CB2 1EZ, U.K. [fax: +44(0)-1223-
336033; e-mail: deposit@ccdc.cam.ac.uk]. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
References
(1) May, E. L.; Murphy, J. G. Structures Related to Morphine. IV.
m-Substituted Phenylcyclohexane Derivatives. J. Org. Chem. 1955,
20, 1197-1201.
(2) Linders, J. T. M.; Flippen-Anderson, J. L.; George, C. F.; Rice, K.
C. An Expedient Synthesis of 9-Keto-2-methyl-5-(dimethoxyphenyl)-
morphans. Tetrahedron Lett. 1999, 40, 3905-3908.
(3) Adah, S. A.; Jacobson, A. E.; Rice, K. C.; Dersch, C. M.; Horel, R.;
Rothman, R. B. Development of Analogs of N-Phenethylphenyl-
morphan as Potential Narcotic Antagonists. Abstracts of Papers of
the American Chemical Society 2000, 220, U578-U578.
(4) Hashimoto, A.; Coop, A.; Rothman, R. B.; Dersch, C.; Xu, H.; Horel,
R.; George, C.; Jacobson, A. E.; Rice, K. C. Synthesis and
Pharmacology of Optically Pure N-Substituted Phenylmorphans as
Opioid Receptor Antagonists. In Problems of Drug Dependence,
1999; Harris, L. S., Ed.; National Institute on Drug Abuse Research
Monograph 180; NIH Publication No. 00-4737; National Institutes
of Health: Washington DC, 2000; p 250.
(5) Thomas, J. B.; Gigstad, K. M.; Fix, S. E.; Burgess, J. P.; Cooper, J.
B.; Mascarella, S. W.; Cantrell, B. E.; Zimmerman, D. M.; Carroll,
F. I. A Stereoselective Synthetic Approach to N-Alkyl-4â-methyl-
5-phenylmorphans. Tetrahedron Lett. 1999, 40, 403-406.
(6) Linders, J. T. M.; Mirsadeghi, S.; Flippen-Anderson, J. L.; George,
C.; Jacobson, A. E.; Rice, K. C. Probes for Narcotic Receptor
Mediated Phenomena. Part 30. Synthesis of rac-(3R,6aS,11aR)-2-
Methyl-1,3,4,5,6,11a-hexahydro-2H-3,6a-methanobenzofuro[2,3-c]a-
zocin-8-ol, an Epoxy Isomer of 5-Phenylmorphan. HelV. Chim. Acta
2003, 86, 484-493.
(7) Carroll, F. I.; Zhang, L.; Mascarella, S. W.; Navarro, H. A.; Rothman,
R. B.; Cantrell, B. E.; Zimmerman, D. M.; Thomas, J. B. Discovery
of the First N-Substituted 4â-Methyl-5-(3-hydroxyphenyl)morphan
to Possess Highly Potent and Selective Opioid δ Receptor Antagonist
Activity. J. Med. Chem. 2004, 47, 281-284.
(8) Kim, I. J.; Dersch, C. M.; Rothman, R. B.; Jacobson, A. E.; Rice, K.
C. A Critical Structural Determinant of Opioid Receptor Interaction
with Phenolic 5-Phenylmorphans. Bioorg. Med. Chem. 2004, 12,
4543-4550.
(9) Hashimoto, A.; Jacobson, A. E.; Rothman, R. B.; Dersch, C. M.;
George, C.; Flippen-Anderson, J. L.; Rice, K. C. Probes for Narcotic
Receptor Mediated Phenomena. Part 28: New Opioid Antagonists
from Enantiomeric Analogues of 5-(3-Hydroxyphenyl)-N-phenyl-
ethylmorphan. Bioorg. Med. Chem. 2002, 10, 3319-3329.
(10) Cheng, K.; Kim, I.-J.; Lee, M. J.; Dersch, C. M.; Rothman, R. B.;
Jacobson, A. E.; Rice, K. C. Opioid Ligands with Mixed Properties
from Substituted Enantiomeric N-Phenethyl-5-phenylmorphans. Syn-
thesis of a µ-Agonist, δ-Antagonist, and δ-Inverse Agonists. Org.
Biomol. Chem. 2007, 5, 1177-1190.
(11) Schiller, P. W.; Weltrowska, G.; Berezowska, I.; Nguyen, T. M. D.;
Wilkes, B. C.; Lemieux, C.; Chung, N. N. TIPP Opioid Peptide
Family: Development of δ Antagonists, δ Agonists, and Mixed µ
Agonist/δ Antagonists. Biopolymers 1999, 51, 411-425.
(12) Thomas, J. B.; Zheng, X. L.; Mascarella, S. W.; Rothman, R. B.;
Dersch, C. M.; Partilla, J. S.; Flippen-Anderson, J. L.; George, C.
F.; Cantrell, B. E.; Zimmerman, D. M.; Carroll, F. I. N-Substituted
9â-Methyl-5-(3-hydroxyphenyl)morphans Are Opioid Receptor Pure
Antagonists. J. Med. Chem. 1998, 41, 4143-4149.
(13) Pirkle, W. H.; Burlingame, T. G.; Beare, S. D. Optically Active NMR
Solvents VI. The Determination of Optical Purity and Absolute
Configuration of Amines. Terahedron Lett. 1968, 9, 5849-5852.
(14) Lott, R. S.; Chauhan, V. S.; Stammer, C. H. Trimethylsilyl Iodide
as a Peptide Deblocking Agent. J. Chem. Soc., Chem. Commun. 1979,
495-496.
(15) Raucher, S.; Bray, B. L.; Lawrence, R. F. Synthesis of (()-
Catharanthine, (+)-Anhydrovinblastine, and (-)-Anhydrovincovaline.
J. Am. Chem. Soc. 1987, 109, 442-446.
(16) Kodato, S.; Linders, J. T. M.; Gu, X.-H.; Yamada, K.; Flippen-
Anderson, J. L.; Deschamps, J. R.; Jacobson, A. E.; Rice, K. C.
Synthesis of rac-(1R,4aR,9aR)-2-Methyl-1,3,4,9a-tetrahydro-2H-1,-
4a-propanobenzofuro[2.3-c]pyridin-6-ol. An Unusual Double Rear-
rangement Leading to the ortho- and para-f Oxide-Bridged Phenyl-
morphan Isomers. Org. Biomol. Chem. 2004, 2, 330-336.
(17) Rice, K. C. A Rapid, High-Yield Conversion of Codeine to Morphine.
J. Med. Chem. 1977, 20, 164-165.
(18) Schwarz, J. B.; Meyers, A. Synthesis of Vicinal Stereogenic Tertiary
and Quaternary Centers Using Chiral Bicyclic Lactams and Diaste-
reoselective Protonation. Asymmetric Synthesis of (+)-Laurene. J.
Org. Chem. 1995, 60, 6511-6514.
Acknowledgment. The research of the Drug Design and
Synthesis section, CBRB, NIDA, was supported by the NIH
Intramural Research Programs of the National Institute of
Diabetes and Digestive and Kidney Diseases and the National
Institute on Drug Abuse, and the latter supported the research
of the Clinical Psychopharmacology section. Y.S.L. thanks Dr.
Sergio Hassan for many helpful discussions. The quantum
chemical study utilized the high-performance computer capabili-
ties of the Helix Systems at the NIH (http://helix.nih.gov) and
the PC/LINUX clusters at the Center for Molecular Modeling
of the NIH (http://cit.nih.gov), and this research was supported
by the NIH Intramural Research Program through the Center
for Information Technology. The authors also acknowledge the
expert technical assistance of Mario Ayestas (NIDA). We thank
Dr. John Lloyd (NIDDK) for the mass spectral data. Some of
the pharmacological data were obtained under the auspices of
the Drug Evaluation Committee, College on Problems of Drug
Dependence. Dr. A. Coop is the recipient of an independent
scientist award from NIDA DA K02-19634.
Note Added after ASAP Publication. This manuscript was
released ASAP on July 11, 2007, with some citations of ref 12
that should be absent from the second paragraph of the first
page, from the first and second paragraphs of Quantum
Chemistry, and from the second paragraph of Conclusion. The
correct version was posted on August 2, 2007.
Supporting Information Available: Atomic coordinates for
(1S,5S)-(-)-2, (1S,5S,9S)-(-)-8, and (1R,5S,9R)-(+)-15 have been
deposited with the Cambridge Crystallographic Data Centre (depo-
sition numbers 625527, 625528, 625529, respectively). Copies of

3776 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 16
Hiebel et al.
(19) Jung, M. E.; Pontillo, J. Synthetic Approach to Potential Precursors
of Sclerophytin A. Tetrahedron 2003, 59, 2729-2736.
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J.
W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghava-
chari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-
Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople,
J. A. Gaussian 03, revision B.04; Gaussian, Inc.: Pittsburgh, PA,
2003.
(20) Clive, D. L. J.; Sun, S. Synthesis of the Racemic Tetracyclic Core
of CP-225,917sA Model Compound Lacking The Sidearms of the
Natural Product. Terahedron Lett. 2001, 42, 6267-6270.
(21) de Grotthuss, C. J. T. Sur La De´composition De L’eau Et Des Corps
Qu’elle Tient En Dissolution AÅ L’aide De L’e´lectricite´ Galvanique.
Ann. Chim. 1806, LVIII, 54-74.
(22) Decoursey, T. E. Voltage-Gated Proton Channels and Other Proton
Transfer Pathways. Physiol. ReV. 2003, 83, 475-579.
(23) Wikstrom, M.; Ribacka, C.; Molin, M.; Laakkonen, L.; Verkhovsky,
M.; Puustinen, A. Gating of Proton and Water Transfer in the
Respiratory Enzyme Cytochrome C Oxidase. Proc. Natl. Acad. Sci.
U. S. A 2005, 102, 10478-10481.
(24) Lanyi, J. K.; Schobert, B. Crystallographic Structure of the Retinal
and the Protein after Deprotonation of the Schiff Base: The Switch
in the Bacteriorhodopsin Photocycle J. Mol. Biol. 2002, 321, 727-
737.
(34) Aceto, M. D.; Bowman, E. R.; Harris, L. S.; Hughes, L. D.; Kipps,
B. R.; Lobe, S. L.; May, E. L. Dependence Studies of New
Compounds in the Rhesus Monkey, Rat and Mouse. In Problems of
Drug Dependence 2005, Proceedings of the 67th Annual Scientific
Meeting, The College on Problems of Drug Dependence, Inc.; Dewey,
W. L., Ed.; U.S. Department of Health and Human Services, National
Institutes of Health, National Institute on Drug Abuse: Washington,
D.C., 2006; NIDA Research Monograph 186, pp 188-228.
(35) Xu, H.; Hashimoto, A.; Rice, K. C.; Jacobson, 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.
(25) Leiderman, P.; Huppert, D.; Agmon, N. Transition in the Temper-
ature-Dependence of GFP Fluorescence: From Proton Wires to
Proton Exit. Biophys. J. 2006, 90, 1009-1018.
(26) Belleau, B.; Conway, T.; Ahmed, F. R.; Hardy, A. D. Importance of
the Nitrogen Lone Electron Pair Orientation in Stereospecific Opiates.
J. Med. Chem. 1974, 17, 907-908.
(27) Kolb, V. M.; Scheiner, S. New Insights in the Clastic Binding
Hypothesis for Opiate Receptor Interactions. 2. Proton-Transfer
Mechanism. J. Pharm. Sci. 1984, 73, 719-723.
(36) Xu, H.; Wang, X.; Wang, J.; Rothman, R. B. Opioid Peptide Receptor
Studies. 17. Attenuation of Chronic Morphine Effects after Antisense
Oligodeoxynucleotide Knock-Down of RGS9 Protein in Cells
Expressing the Cloned Mu Opioid Receptor. Synapse 2004, 52, 209-
217.
(28) Schobert, B.; Brown, L. S.; Lany, J. K. Crystallographic Structures
of the M and N Intermediates of Bacteriorhodopsin: Assembly of a
Hydrogen-Bonded Chain of Water Molecules between Asp-96 and
the Retinal Schiff Base. J. Molec. Biol. 2003, 330, 553-570.
(29) Arseniev, A. S.; Barsukov, I. L.; Bystrov, V. F.; Lomize, A. L.;
Ovchinnikov Yu, A. 1H-NMR Study of Gramicidin A Transmem-
brane Ion Channel. Head-to-Head Right-Handed, Single-Stranded
Helices. FEBS Lett. 1985, 186, 168-174.
(37) de Costa, B. R.; Iadarola, M. J.; Rothman, R. B.; Berman, K. F.;
George, C.; Newman, A. H.; Mahboubi, A.; Jacobson, A. E.; Rice,
K. C. Probes for Narcotic Receptor Mediated Phenomena. 18.
Epimeric 6R-Iodo-3,14-dihydroxy-17-(cyclopropylmethyl)-4,5R-ep-
oxymorphinans and 6â-Iodo-3,14-dihydroxy-17-(cyclopropylmethyl)-
4,5R-epoxymorphinans as Potential Ligands for Opioid Receptor
Single Photon Emission Computed TomographysSynthesis, Evalu-
ation, and Radiochemistry of 125I-6â-Iodo-3,14-Dihydroxy-17-(cy-
clopropylmethyl)-4,5R-epoxymorphinan. J. Med. Chem. 1992, 35,
2826-2835.
(30) Lee, Y. S.; Krauss, M. Dynamics of Proton Transfer in Bacterior-
hodopsin. J. Am. Chem. Soc. 2004, 126, 2225-2230.
(31) Berezhkovskii, A.; Hummer, G. Single-File Transport of Water
Molecules through a Carbon Nanotube. Phys. ReV. Lett. 2002, 89,
064503.
(32) Hassan, S. A.; Hummer, G.; Lee, Y.-S. Effects of Electric Fields on
Proton Transport through Water Chains. J. Chem. Phys. 2006, 124,
204510-204518.
(33) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G.
A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
(38) Ni, Q.; Xu, H.; Partilla, J. S.; de Costa, B. R.; Rice, K. C.; Rothman,
R. B. Selective Labeling of κ2 Opioid Receptors in Rat Brain by
[
125I]IOXY: Interaction of Opioid Peptides and Other Drugs with
Multiple κ2a Binding Sites. Peptides 1993, 14, 1279-1293.
(39) SHELXTL, v6.10; Bruker AXS, Inc.: Madison, WI, 2000.
JM061325E