A novel and facile preparation of bremazocine enantiomers through optically pure N-norbremazocines

Article abstract of DOI:10.1016/j.bmc.2003.10.029

In order to provide ready access to multigram quantities of the optically pure bremazocines [(-)- and (+)-9,9-dimethyl-5-ethyl-2-hydroxy-2-(1-hydroxy- cyclopropylmethyl)-6,7-benzomorphan)], we have developed an improved non-chromatographic synthesis, and

Full text of DOI:10.1016/j.bmc.2003.10.029

Bioorganic & Medicinal Chemistry 12 (2004) 233–238
A novel and facile preparation of bremazocine enantiomers
through optically pure N-norbremazocines
Elisabeth Greiner, John E. Folk, Arthur E. Jacobson and Kenner C. Rice*
Laboratory of Medicinal Chemistry, Building 8, Room B1-23, National Institutes of Diabetes and Digestive and Kidney Diseases, National
Institutes of Health, Department of Health and Human Services, 8 Center Drive, MSC 0815, Bethesda, MD 20892-0815, USA
Received 21 July 2003; accepted 4 October 2003
Abstract—In order to provide ready access to multigram quantities of the optically pure bremazocines [(ꢀ)- and (+)-9,9-dimethyl-
5
-ethyl-2-hydroxy-2-(1-hydroxy-cyclopropylmethyl)-6,7-benzomorphan)], we have developed an improved non-chromatographic
1
synthesis, and determined the optical purity of their N-nor precursors using a rapid and relatively simple H NMR method based
on diastereomeric derivatization with optically pure 1-phenylethylisocyanate. This method of determining optical purity should be
readily amenable to similar systems containing phenolic amino functionalities. Finally, a greatly simplified methodology for intro-
duction of the N-(1-hydroxycyclopropylmethyl) substituent in bremazocine is described. The improved synthetic method—the
overall yield was increased about 3-fold—combined with the practical methodology to determine optical purity will considerably
facilitate the employment of these enantiomers as pharmacological tools for examination of the k-opioid receptor system, as well as
their evaluation as drug abuse treatment agents. This synthesis will also enable the study of these enantiomers for other, non-clas-
sical applications (e.g., treatment agents for HIV).
Published by Elsevier Ltd.
1
. Introduction
enantiomer of bremazocine is an importamt tool for
opioid receptor studies; the (+)-enantiomer, which does
not interact well with opioid receptors, can be used to
examine non-opioid pharmacological effects. We
required substantial quantities of both enantiomers for
future pharmacological studies in various primate mod-
els of behavior and disease. Since the enantiomers of
bremazocine are not commercially available and their
only published synthesis is a long, laborious multistep
process, described in the patent literature, the devel-
opment of a simplified synthesis of the pure enantiomers
has become extremely important. We now report an
improved methodology for the preparation of both
enantiomers, optically pure, and in multigram quantities.
It is now widely recognized that the enantiomers of a
drug can act on diverse biological systems and thus may
have vastly different pharmacological properties (e.g.,
1
the morphine enantiomers and the tragic case of
2
thalidomide ). Thus any studies with racemates involve
the difficult evaluation of two different drugs simulta-
neously. The enantiomers of bremazocine (Fig. 1) have
been found to have very different affinity at opioid
receptors, thus possessing very different pharmacological
activities and profiles. The opioid effects of racemic
bremazocine are due in large measure to the agonist activ-
ity of its (ꢀ)-enantiomer at the k-opioid receptor.
k-Agonists have become important pharmacological
tools and are being examined as treatment agents for
cocaine abuse due to their ability to lower excessively
elevated dopamine levels in the nucleus accumbens.
Recently, k-agonists have been evaluated for their
potential usefulness in the treatment for HIV-1 and as
cardiovascular agents. k-Agonists such as the proto-
typic ketocyclazocine, and bremazocine, are members of
1
4
3
3ꢀ5
6ꢀ8
2. Results and discussion
9
,10
Our work focused on shortening the original route,14
optimizing various steps in the published synthesis that
were found to be the most difficult and time-consuming,
and eliminating the formerly used 1-oxaspir-
o[2.2]pentane for the introduction of the N-(1-hydroxy-
cyclopropylmethyl) substituent in bremazocine.
1
1
1
2
1
3
the 6,7-benzomorphan class of opioids. The (ꢀ)-
Starting from commercially available methyl p-
methoxyphenylacetate (1) the keto compound 2 was
*
Corresponding author. Tel.: +1-301-496-1856; fax: +1-301-402-
589; e-mail: kr21f@nih.gov
0
0968-0896/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmc.2003.10.029

234
E. Greiner et al. / Bioorg. Med. Chem. 12 (2004) 233–238
prepared by a Reformatzky reaction of 1 with methyl-2-
zation of 6 to 7. However, we found DMF to be an
unsatisfactory solvent for this reaction because it pro-
moted the formation of unidentified by-products.
Changing the solvent to toluene and employing catalytic
amounts of EtOH to initiate the reaction increased the
yield of 7 from 5 from less than 50% to about 60%.
Compound 8 was prepared from 7 in five steps as out-
1
4
bromoisobutyrate as previously described (Scheme 1).
However, we developed a distinctly improved workup
procedure by changing the extraction conditions, which
afforded 2 in 92% yield versus the reported 60–65%
1
4
yield. After quenching the reaction with sulfuric acid,
the solution of 2 in Et O was washed with water until
2
1
5
the pH was neutral before washing with aqueous base in
order to prevent formation of an inorganic salt
lined in the original report. Key intermediate 8, (ꢁ)-
N-norbremazocine, was resolved into its enantiomers,
(+)-8 and (ꢀ)-8, through diastereomeric salt formation
(
Na SO ) and emulsion. This salt formation was found
2 4
1
4
to occur in the original procedure when the organic
extract was immediately basified. The elimination of
the salt facilitated the work up and precluded loss of
product by its entrapment in the salt fraction.
with 3-bromocamphor-8-sulfonic acid (Scheme 3).
However, analysis of the optical purity of the enantio-
1
4
mers was not reported. Because optically pure enan-
tiomers were essential at this stage of the synthesis, we
devised a simple NMR method for the determination of
optical purity as described below.
The published route to intermediate 5 was formerly
1
4
conducted in three steps: (1) conversion of 2 into a
benzyliminoester 3 with TiCl in toluene, (2) reduction
In the original synthetic pathway17 the enantiomers of
bremazocine (+)-9 and (ꢀ)-9 were prepared using 1-
oxaspiro[2.2]pentane. This reagent is generated in two
steps from methallyl chloride in low overall yield (about
4
of 3 to benzylamine 4 by catalytic hydrogenation over
Pt/C, and (3) a second hydrogenation with Pd as cata-
lyst to afford the debenzylated primary amine 5 (Scheme
1
8
1
dure by a one-pot reductive amination of ketone 2 to
). We simplified the synthesis of 5 to a two-step proce-
15%). Its preparation is laborious and requires special
precautions due to the volatile nature of the inter-
18
mediate methylenecyclopropane.
1
5
benzylamine 4 utilizing TiCl /NaCNBH followed by
4
3
N-debenzylation of 4 to the primary amine 5. The N-
debenzylation of compound 4 as described in the origi-
nal procedure proved troublesome in our hands and did
not result in reproducible yields. This problem was
overcome by conducting the reaction under catalytic
In order to avoid the problematic preparation and use
of 1-oxaspiro[2.2]pentane, our synthesis of (+)-9 and
(ꢀ)-9 employed 1-acetoxycyclopropanecarboxylic acid,
1
9
prepared by simple acetylation of the commercially
available 1-hydroxycyclopropanecarboxylic acid. The
synthesis of bremazocine was then accomplished
through 1-[3-(dimethylamino)propyl]-3-ethyl-carbodi-
imide (EDCI) mediated coupling of 1-acetoxy-
cyclopropanecarboxylic acid with the enantiopure N-
norbremazocines (+)-8 and (ꢀ)-8, respectively, to give
(+)-8a and (ꢀ)-8a as crystalline intermediates. This was
followed by simultaneous reduction of the amide and
1
6
transfer hydrogenolysis conditions employing ammo-
nium formate as the hydrogen source to give the amine
5
. These modifications provided convenient and rela-
tively quick access to intermediate 5 and helped increase
the overall yield of 5 from 1 from 22 to 59%.
Compound 6 was prepared by treatment of 5 with
methyl acrylate (Scheme 2) followed by acetic acid and
formaldehyde in an Eschweiler–Clark reaction, accord-
ester groups with AlH to give the bremazocine enan-
3
1
4
ing to the literature. In the next step, sodium hydride
1
tiomers (+)-9 and (ꢀ)-9. Our values for optical rotation
4
17
in DMF was originally used in the Dieckman cycli-
were comparable with those reported, although the
reported values were stated without mention of solvent
or concentration. The equal and opposite optical rota-
tions of our enantiomers (+)-9 and (ꢀ)-9 suggested
complete resolution. However, we required an indepen-
dent determination of enantiopurity of the isomers. Our
Figure 1. Absolute stereochemistry of bremazocine enantiomers.
Scheme. 2. Reagents: (a) methyl acrylate, cat. HOAc, HCOOH,
HCHO; (b) NaH, toluene.
Scheme. 1. Reagents: (a) Zn, methyl 2-bromoisobutyrate, THF; (b)
TiCl , benzylamine, Ch Cl ; (c) NaCNBH ; (d) HCO NH , Pd/C,
MeOH.
4
2
2
3
2
4
Scheme. 3. Reagents: (a) 1-Acetoxycyclopropanecarboxylic acid,
EDCI, CH Cl ; (b) LiAlH , H SO , THF.
2
2
4
2
4

E. Greiner et al. / Bioorg. Med. Chem. 12 (2004) 233–238
235
methodology for evaluation of enantiopurity as descri-
bed below showed that the N-norbremazocine enantio-
mers (+)-8 and (ꢀ)-8 were more than 99% pure.
pure samples within the limits of detection. In order to
measure the experimental limits of accuracy of this
method we deliberately added a mixture containing the
opposite isomer together with chiral reagent in steps of
0.5% (wt%) to each sample. Addition of 0.5% of the
less predominant enantiomer to the predominant isomer
did not cause a detectable change in the signal. After
addition of 1% a weak signal of the minor isomer,
which was clearly distinguishable from background
noise could first be observed (Fig. 2, traces D and E).
Continued addition of the minor isomer in 0.5% wt
increments produced an increased intensity of its peak.
These findings indicate that the detection limit of
enantiomeric impurity was between 0.5 and 1%. Thus,
our resolved compounds were at least 99% optically
pure. In order to avoid misinterpretation of the spec-
troscopic data and to obtain reproducible and accurate
results with this method, it is essential to use optically
pure chiral derivatization agent. The absence of optical
impurities prevents the formation of peaks due to arti-
facts in the reagent that would occur when these impu-
rities reacted with optically pure enantiomers. This
inexpensive spectroscopic method should prove useful
for easily establishing the optical purity of certain
opioids and other molecular structures that contain
phenolic secondary and possibly primary amino
functions.
We utilized an NMR method based on spectroscopic
enantiodiscrimination that we had previously employed
with other phenolic amines.20 The method relies on
derivatization of N-norbremazocine enantiomers to give
diastereomers selectively on nitrogen (vs phenolic
1
hydroxyl) that show separation of peaks in the
H
NMR spectrum. We used optically pure R-(+)-1-
phenylethylisocyanate for the reaction with (+)-8 and
(
ꢀ)-8 in CDCl to give the corresponding ureas 10
3
(Scheme 4). The quantitative and rapid formation of
diastereomers with this reagent offers an excellent alter-
native to other commonly used reagents such as chiral
lanthanide shift reagent or chiral solvating agents. It
bears the advantage of rendering both chromatography
and isolation of the diastereomers unnecessary.
Well-separated peaks between 1.1 and 1.6 ppm were
1
observed in the H NMR spectrum (300 MHz) of the
diastereomeric urea mixture 10 (the significant peaks are
shown in Fig. 2, trace A). The urea of each enantiomer,
formed by reaction with the chiral reagent, displayed a
signal that overlapped some of those observed in the
diastereomeric ureas of the racemic mixture (Fig. 2,
traces B and C). This occurred at d 1.50, 1.47 and 1.15
ppm for the (ꢀ)-isomer and at d 1.49, 1.46 and 1.14 ppm
for the (+)-isomer in the selected area. Little if any
enantiomeric contamination was present in the enantio-
3. Conclusion
Our improved synthesis of the two optically pure
N-norbremazocine enantiomers combined with the
facile and rapid determination of their optical purity
and the greatly simplified introduction of the N-sub-
stituent considerably facilitates the availability of bre-
mazocine enantiomers for use as pharmacological tools
for the in vitro and in vivo exploration of the k-opioid
receptor system and for other possible applications. In
addition, the increased availability of the N-nor inter-
Scheme. 4. Diastereomeric urea formation employing R-(+)-1-
phenylethylisocyanate.
Figure 2. 1H NMR spectra of ureas of (+)-8 and (ꢀ)-8: (a) two well separated doublets (d 1.48 and 1.47 ppm, 2 d, J=6.9 Hz, 3H) and two singlets
(
(
d 1.15 and 1.14 ppm, 2 s, 6H) in the spectrum of diastereomeric urea mixture 10; (b) and (c) peaks for enantiopure ureas of (+)-8 and (ꢀ)-8; (d) and
e) continued addition of the minor isomer produced an increased intensity of its peak.

236
E. Greiner et al. / Bioorg. Med. Chem. 12 (2004) 233–238
mediates will further enable future structure–activity
studies of novel pharmacological effects. This is parti-
cularly important in the opioid area where the pharma-
cological profile is largely determined by the N-
substituent.
saturated NaHCO solution and drying with Na SO
3 2 4
the solvent was removed and the product 4 crystallized
ꢂ
from 95% EtOH, (total yield, 120.5 g, 70.6%) mp 60 C
ꢂ
1
4
(lit. 58–60 C).
4.3. 3-Amino-2,2-dimethyl-4-(p-methoxyphenyl)butyrate
(
5)
4
. Experimental
To a solution of 4 (51.2 g, 0.2 mol) in MeOH (600 mL)
was added Pd/C (6.0 g, 10%) under N . The mixture
All melting points were determined on a Thomas-Hoover
1
2
melting point apparatus and are uncorrected. The H
NMR spectra were recorded on a Varian XL-300 instru-
was brought to reflux at which time a solution of
ammonium formate (37.8 g, 0.6 mol) in H O (75 mL)
2
ment using DMSO-d or CDCl as solvent, d values in
was added. The mixture was heated to reflux for 2 h,
cooled, and the catalyst removed by filtration through a
pad of Celite. The oil obtained upon removal of solvent
was partitioned between ether and saturated NaHCO3
and the aqueous layer was again extracted with ether.
The combined ether portions were washed with brine,
dried over Na SO and evaporated. Crystallization
6
3
ppm (TMS as internal standard), J (Hz) assignments of
1
H resonance coupling. Chemical-ionization mass spectra
CIMS) were obtained using a Finnigan 1015 mass spec-
(
trometer. Electron ionization (EIMS) mass spectra were
obtained using a VG-Micro Mass 7070F mass spectro-
meter. Elemental analyses were performed by Atlantic
Microlabs, Inc., Norcross, GA, USA.
2
4
ꢂ
from hexane afforded 5 (34.1 g, 90.5%), mp 46.5 C
ꢂ
1
4
(
lit. 46–49 C).
4
.1. 2,2-Dimethyl-4-(p-methoxyphenyl)acetoacetate (2)
0
4.4. 3-(p-Methoxybenzyl)-2,2-N-trimethyl-3,3 -iminodi-
propionate (6)
A solution of methyl 2-bromoisobutyrate (583.0 g, 3.2
mol) and methyl p-methoxyphenylacetate (1) (450.0 g,
2
.5 mol) in THF (435 mL) was added to a refluxing
suspension of granulated Zn (189 g, 2.9 mol) in THF
435 mL) that was activated with I (0.8 g, 3.5 mmol)
A mixture of 5 (62.6 g, 0.25 mol) and methyl acrylate
(24.9 mL, 0.3mol) was stirred vigorously at 50 C.
ꢂ
Acetic acid (6.23mL) was added rapidly and the tem-
(
2
ꢂ
taneously to 100 C. After 1 h, the excess of methyl
under argon atmosphere. After the reaction had started,
the addition was continued at such a rate that refluxing
was maintained with minimal application of external
heat. Stirring was continued at reflux for 5 h. The reac-
tion mixture was cooled in an ice bath and H SO (1.25
L, 4 N) was added slowly under vigorous stirring. The
organic layer was separated and after filtration from
solid material, the aqueous layer was extracted with
ether. The combined organic layers were first washed
perature was raised to 90 C at which time it rose spon-
ꢂ
acrylate and acetic acid were removed at 50 C under
ꢂ
aspirator vacuum. Formic acid (49 mL, 95%) and aqu-
eous formaldehyde (49 mL, 35%) were added dropwise
successively over a period of 30 min while cooling the
2
4
reaction with H O. Subsequently, the reaction mixture
2
ꢂ
was warmed to 40 C for 1 h, then stirred at room tem-
ꢂ
perature for 16 h and finally heated to 60 C for 1 h. The
with H O until the pH was neutral, then with saturated
2
low boiling material was evaporated and the resulting
oil was treated with an access of a concentrated aqueous
solution of K CO and CHCl (300 mL). The aqueous
NaHCO , brine and dried over Na SO . Evaporation of
4
3
2
the solvent yielded the crude product 2 that was purified
ꢂ
2
3
3
by distillation in vacuo (576.0 g, 92%), bp 135 C/0.7
ꢂ
layer was extracted with CHCl3 and the combined
organic extracts were washed with brine and dried over
MgSO . Removal of solvent provided 6 as an oil (87.7
1
4
mm (lit. 145 C/1 mm).
4
g) that was used without further purification for the
next reaction.
4.2. 3-Benzylamino-2,2-dimethyl-4-(p-methoxyphenyl)-
butyrate (4)
TiCl (262.0 mL, 0.3mol, 1 M in CH Cl ) was added
4
2
2
4.5. 2-(p-Methoxybenzyl)-4-oxo-1,3,3-trimethyl-5-piper-
idine-carboxylate hydrochloride (7)
slowly to a mixture of 2 (125.0 g, 0.5 mol) and benzyl-
amine (166.0 g, 1.6 mol) in dry CH Cl (3L) under N
2
2
2
at such a rate as to keep the reaction mixture warm, but
not boiling (45–60 min for addition). Stirring under N2
was continued for 48 h, then a solution of NaCNBH3
Compound 6 (87.7 g, 9 mmol) was added to a suspen-
sion of NaH (18 g, 0.75 mol, 60% dispersion in oil) in
toluene (600 mL, dried by azeotropic distillation over-
night). The mixture was gradually heated to reflux and
EtOH (3mL) was added to start the reaction after the
(
90.0 g, 1.4 mol) in MeOH (1.2 L) was cautiously added.
After 30 min the mixture was filtered through Celite and
evaporated to give an oil. This oil was stirred with
EtOAc (900 mL) and again filtered through Celite. The
yellow gelatinous residue was washed with EtOAc and
the filtrate and washings evaporated to an oil under
vacuum. Crystallization from 95% EtOH (250 mL)
provided product (97.2 g). An additional 23.3 g of pro-
duct was obtained by evaporating the mother liquor
from the first crystallization and dissolving it in EtOAc
ꢂ
temperature had reached 50–60 C. After about 15–30
min, a vigorous evolution of hydrogen occurred. Heat-
ing was continued for 3h. Upon cooling the reaction
mixture was washed with half saturated NaHCO and
3
brine and dried over Na SO . After removal of solvent,
2
4
the residual oil was dissolved in EtOH and made acidic
with HCl in ether. Crystallization was induced by cool-
ꢂ
dec. (lit. 163–164 C, dec).
ꢂ
ing to ꢀ70 C. Yield 52 g (59% yield from 5), mp 169 C,
1
4
ꢂ
(
800 mL). After washing the solution twice with half

E. Greiner et al. / Bioorg. Med. Chem. 12 (2004) 233–238
237
0
4
.6. (+) and (ꢀ)-9,9-Dimethyl-5-ethyl-2 -hydroxy-6,7-
J=7.5 Hz), 0.98 (t, 3H, J=7.5 Hz); MS m/z (FAB)
+
benzomorphan [(+)-norbremazocine, (+)-8 and (ꢀ)-nor-
372 M+1] . Anal. calcd for C H NO : C, 71.13; H,
2
2
29
4
bremazocine, (ꢀ)-8]
7.87; N, 3.77. Found: C, 70.88; H, 8.14; N, 3.69.
(
ture in 5 steps from 7.
+)-8 and (ꢀ)-8 were prepared according to the litera-
4
.12. (ꢁ) - 2 - (1 - Acetoxy - cyclopropylcarbonyl) - 9,9 - di-
1
4
0
methyl-5-ethyl-2 -hydroxy-6,7-benzomorphan [(ꢁ)-8a]
This material was prepared in 86% yield using the pro-
1
4
ethyl-2 -hydroxy-2-(1-phenylethyl)carbamoyl-6,7-benzo-
.7. Preparation of H NMR samples of 9,9-dimethyl-5-
0
ꢂ
1
cedure described for (ꢀ)-8a; mp 196–198 C; H NMR
(
(
DMSO-d ) d 9.08 (s, 1H, OH), 6.91–6.55 (m, 3H), 1.83
q, 2H, J=7.5 Hz), 0.99 (t, 3H, J=7.5 Hz).
6
morphan from (+)-8 and (ꢀ)-8
R-(+)-1-Phenylethylisocyanate (2.4 mL, 0.017 mmol)
2
5
ꢂ
MeOH)} (4.0 mg, 0.016 mmol) in CDCl (0.6 mL). The
was added to a solution of (+)-8 {[a] =+75 (c 2,
D
0
cyclopropylmethyl)-6,7-benzomorphan
4
.13. (ꢀ)-9,9-Dimethyl-5-ethyl-2 -hydroxy-2-(1-hydroxy-
3
hydrochloride
1
HNMR spectrum of this mixture was determined.
Subsequently, a mixture consisting of the other isomer,
.
[
(ꢀ)-bremazocine hydrochloride, (ꢀ)-9 HCl]
(
ꢀ)-8, (2.4 mg, 0.01 mmol) and R-(+)-1-phenylethyl-
H SO (7.66 g, 75 mmol, 100%) was added dropwise
2 4
isocyanate (1.44 mL, 0.01 mmol) in exactly 2.5 mL
CDCl3 was added in 0.5% (20.8 mL) increments by
weight relative to (+)-8 and analyzed by H NMR
to a solution LAH in THF (150 mmol, 150 mL of 1 M
ꢂ
LAH in THF) at 0 C under argon and the mixture
1
was allowed to stir at room temperature. After 1 h (ꢀ)-
8a (5.6 g, 15 mmol) in dry THF (200 mL) was added in
a dropwise manner. After 2 h the reaction was stopped
spectroscopy. An analogous procedure was followed for
2
5
D
ꢂ
(
ꢀ)-8 {[a] =ꢀ75 (c 2, MeOH)}.
by cautiously adding a mixture of THF/H O (15 mL,
2
1:1) followed by concd NH OH (22 mL). The mixture
was filtered and the precipitate washed well with
CHCl . The combined filtrates were evaporated and
3
4
4
.8. Carbamate from (+)-8
1
H NMR (CDCl ) d 7.39–6.56 (m, 8H), 1.47 (d, 3H,
3
J=6.9 Hz), 1.13 (s, 3H), 1.02 (t, 3H, J=7.5 Hz), 0.95 (s,
3
replaced with MeOH (80 mL). After the pH of the
solution was adjusted to 3with ethereal HCl, the pro-
duct was crystallized by boiling the solution and add-
ing EtOAc to keep the volume constant until
crystallization occurred. Yield (3.7 g, 70%), mp 255–
H).
4
.9. Carbamate from (ꢀ)-8
1
ꢂ
14
ꢂ
25
ꢂ
H NMR (CDCl ) d 7.39–6.58 (m, 8H), 1.49 (d, 3H,
256 C, (lit.
1
242–246 C), [a] =ꢀ107 (c 0.95,
3
D
J=6.9 Hz), 1.15 (s, 3H), 1.02 (t, 3H, J=7.2 Hz), 0.95 (s,
3
MeOH); H NMR (DMSO-d ) d 9.26 (s, 1H, OH), 8.45
6
H).
(s, 1H, NH), 6.97 (d, 1H, J=8.1 Hz), 6.70 (d, 1H,
J=2.4 Hz), 6.63(dd, 1H, J=8.1 Hz, J=2.4 Hz), 6.07 (s,
+
1H, OH); MS m/z (FAB) 316 M+1] . Anal. calcd for
C H NO Cl: C, 68.26; H, 8.59; N, 3.98. C, 68.20; H,
4
.10.
(ꢀ)-2-(1-Acetoxy-cyclopropylcarbonyl)-9,9-di-
0
20 30
2
methyl-5-ethyl-2 -hydroxy-6,7-benzomorphan [(ꢀ)-8a]
8.66; N, 3.98.
EDCI (8.97 g, 46.8 mmol) was added portionwise to a
stirred mixture of (ꢀ)-8 (4.42 g, 18 mmol) and 1-acet-
oxycyclopropane-1-carboxylic acid (5.19 g, 36 mmol) in
0
cyclopropylmethyl)-6,7-benzomorphan [(+)-bremazocine
4
.14. (+)-9,9-Dimethyl-5-ethyl-2 -hydroxy-2-(1-hydroxy-
ꢂ
anhyd CH Cl (70 mL) at 0 C. After stirring for 30 min
2
2
.
hydrochloride, (+)-9 HCl]
ꢂ
perature and stirring was continued overnight. CH Cl2
at 0 C the solution was allowed to come to room tem-
(+)-9 was prepared in 67% yield using the procedure
2
ꢂ
14
ꢂ
was added and the mixture was washed with H O, with
2
described for (ꢀ)-9, mp 254–256 C, (lit. 242–246 C);
ꢂ
6
2
5
1
brine and dried over Na SO . The solvent was removed
2
[a] =+110.4 (c 0.90, MeOH); H NMR (DMSO-d )
4
D
under reduced pressure and the crude product was fil-
tered through a short column of silica gel (acetone/n-
hexane, 3/5) to yield (ꢀ)-8a as colorless crystals (5.6 g,
d 9.26 (s, 1H, OH), 8.44 (s, 1H, NH), 6.97 (d, 1H, J=8.4
Hz), 6.70 (d, 1H, J=2.4 Hz), 6.63(dd, 1H, J=8.4 Hz
J=2.4 Hz), 6.07 (s, 1H, OH); MS m/z (FAB)
ꢂ
25
D
ꢂ
+
8
4%), mp 168–171 C; [a] =ꢀ194.2 (c 1.04, MeOH);
316 M+1] . Anal. calcd for C H NO Cl: C, 68.26; H,
20 30
2
1
H NMR (DMSO-d ) d 9.07 (s, 1H, OH), 6.91–6.55 (m,
8.59; N3.98. C, 68.37; H, 8.62; N3.98.
6
3
MS m/z (FAB) 372 M+1] . Anal. calcd for
H ar), 1.83 (q, 2H, J=7.5 Hz), 0.99 (t, 3H, J=7.5 Hz);
+
0
4
.15. (ꢁ)-9,9-Dimethyl-5-ethyl-2 -hydroxy-2-(1-hydroxy-
C H NO : C, 71.13; H, 7.87; N, 3.77. C, 71.12; H,
2
2
29
4
cyclopropylmethyl)-6,7-benzomorphan [(ꢁ)-bremazocine,
8
.11; N, 3.70.
(
ꢁ)-9]
(
ꢁ)-9 was prepared in 68% yield using the procedure
4.11. (+)-2-(1-Acetoxy-cyclopropylcarbonyl)-9,9-di-
0
methyl-5-ethyl-2 -hydroxy-6,7-benzomorphan [(+)-8a]
ꢂ
1
described for (ꢀ)-9, mp 247-249 C. H NMR (DMSO-
d ) d 9.27 (s, 1H, OH), 8.45 (s, 1H, NH), 6.97 (d, 1H,
6
This material was prepared in 79% yield using the pro-
ꢂ
J=8.1 Hz), 6.70 (d, 1H ar, J=2.1 Hz), 6.64 (dd, 1H,
J=8.1, Hz J=2.1 Hz), 6.07 (s, 1H, OH). Anal. calcd for
C H NO Cl: C, 68.26; H, 8.59; N 3.98. C, 68.13; H,
cedure described for (ꢀ)-8a. Mp 168–170 C;
2
5
ꢂ
1
[
a] =+189.1 (c 0.97, MeOH); H NMR (DMSO-d )
D
6
20 30
2
d 9.07 (s, 1H, OH), 6.91-6.54 (m, 3H ar), 1.83 (q, 2H,
8.56; N3.97.

238
E. Greiner et al. / Bioorg. Med. Chem. 12 (2004) 233–238
8. Zernig, G.; Burke, T.; Lewis, J. W.; Woods, J. H. J.
Pharmacol. Exp. Ther. 1996, 279, 23 .
Acknowledgements
9
. Collins, S. L.; Kunko, P. M.; Ladenheim, B.; Cadet,
J. L.; Carroll, F. I.; Izenwasser, S. Synapse 2002, 45,
The authors thank the National Institute on Drug
Abuse, NIH, for partial financial support of our
research program.
153.
0. Mori, T.; Nomura, M.; Nagase, H.; Narita, M.; Suzuki,
T. Psychopharmacology 2002, 161, 17.
1
1
1
1. Lokensgard, J. R.; Gekker, G.; Peterson, P. K. Biochem-
ical Pharmacol. 2002, 63, 1037.
2. Pugsley, M. K. Pharmacol. Ther. 2002, 93, 51.
References and notes
1
2
3
. Jacquet, Y. F.; Klee, W.; Rice, K. C.; Iijima, I.; Minami-
kawa, J.-I. Science 1977, 198, 842.
. Eriksson, T.; Bjorkman, S.; Hoglund, P. Eur. J. Clin.
Pharmacol. 2001, 57, 365.
. Roemer, D.; Buescher, H.; Hill, R. C.; Maurer, R.;
Petcher, T. J.; Welle, H. B. A.; Bakel, H. C. C. K.;
Akkerman, A. M. Life Sci. 1980, 27, 971.
13. Casy, A. F.; Parfitt, R. T. Opioid Analgesics: Chemistry
and Receptors; Plenum: New York, 1986.
14. Akkerman, A. M., Janssen, P. A. J. German
Patent 2,027,077 1970 (CA 74:125486, 1971); see
also US Patent 4,012,392 1977 at http://www.
uspto.gov.
15. Barney, C. L.; Huber, E. V.; McCarthy, J. R. Tetrahedron
Lett. 1990, 31, 5547.
16. Adger, B. M.; O’Farrell, C.; Lewis, N. J.; Mitchell, M. B.
Synthesis 1987, 53 .
17. Akkerman, A. M., van Bakel, H. C. C. K. In European
patent appl. 4960 1983( CA 93:46941, 1980).
18. Salaun, J. R.; Champion, J.; Conia, J. M. Org. Synth.
1977, 57, 3 6.
19. Hartmann, W. Synthesis 1989, 272.
20. Rice, K. C.; Brossi, A. J. Org. Chem. 1980, 45, 592.
4
5
6
7
. Corbett, A. D.; Kosterlitz, H. W. Br. J. Pharmacol. 1986,
89, 245.
. Butelman, E. R.; Baron, S. P.; Woods, J. H. Behav.
Pharmacol. 1993, 4, 57.
. Amarante, L. H.; Duarte, I. D. G. Eur. J. Pharmacol.
2002, 454, 19.
. Meng, F.; Xie, G.; Thompson, R. C.; Mansour, A.;
Goldstein, A.; Watson, S. J.; Akil, H. Proc. Natl. Acad.
Sci. U.S.A. 1983, 90, 9954.