(E)- and (Z)-7-arylidenenaltrexones: Synthesis and opioid receptor radioligand displacement assays

Article abstract of DOI:10.1021/jm960573f

The E-isomer of 7-benzylidenenaltrexone (BNTX, 1a) was reported by Portoghese as a highly selective δ-opioid antagonist. The corresponding Z- isomer 1b was not readily available through direct aldol condensation of naltrexone (6) with benzaldehyde. Using

Full text of DOI:10.1021/jm960573f

J . Med. Chem. 1997, 40, 749-753
749
(
E)- a n d (Z)-7-Ar ylid en en a ltr exon es: Syn th esis a n d Op ioid Recep tor
Ra d ioliga n d Disp la cem en t Assa ys
Robert B. Palmer, Alana L. Upthagrove, and Wendel L. Nelson*
Department of Medicinal Chemistry, School of Pharmacy, University of Washington, Seattle, Washington 98195
X
Received August 5, 1996
The E-isomer of 7-benzylidenenaltrexone (BNTX, 1a ) was reported by Portoghese1,2 as a highly
selective δ-opioid antagonist. The corresponding Z-isomer 1b was not readily available through
direct aldol condensation of naltrexone (6) with benzaldehyde. Using the photochemical
3
methods employed by Lewis to isomerize cinnamamides, we have obtained Z-isomer 1b in
good yield from E-isomer 1a . A series of (E)- and (Z)-7-arylidenenaltrexone derivatives was
prepared to study the effect of larger arylidene groups on opioid receptor affinity in this series.
By aldol condensation of naltrexone (6) with benzaldehyde, 1-naphthaldehyde, 2-naphthalde-
hyde, 4-phenylbenzaldehyde, and 9-anthracaldehyde, the (E)-arylidenes were readily obtained.
Photochemical isomerization afforded the corresponding Z-isomers. These compounds were
evaluated via opioid receptor radioligand displacement assays. In these assays, the Z-isomers
generally had higher affinity and were more δ-selective than the corresponding E-isomers.
The (Z)-7-(1-naphthylidene)naltrexone (3b) showed the greatest selectivity (δ:µ ratio of 15)
and highest affinity δ-binding (K ) 0.7 nM). PM3 semiempirical geometry optimizations
i
suggest a significant role for the orientation of the arylidene substituent in the binding affinity
and δ-receptor selectivity. This work demonstrates that larger groups may be incorporated
into the arylidene portion of the molecule with opioid receptor affinity being retained.
Sch em e 1a
Naltrexone (6) is an important opioid antagonist that
binds at all classes of opioid receptors but has a small
preference for the µ- and κ-receptors over the δ-recep-
tors.4 In a search for receptor subtype specific antago-
5
nists, it has been found that naltrindole and (E)-7-
benzylidenenaltrexone (BNTX, 1a ) are agents with
significant δ-receptor selectivity.2 It has been postu-
lated that the similar activities of these two compounds
are due, at least in part, to sharing the same confor-
mational space by their aromatic moieties in the ligand-
receptor interaction.2 Rigidly fixing the location of the
aryl group, as through the olefin in 1a , allows the
isolation of that component of binding influenced by the
conformational space occupied by the aryl group.
a
Reagents: (a) R-CHO, NaOH, MeOH, 0 °C for 12 h; (b) R-CHO,
Though the influence of the direction of the aryl group
with regard to the rest of the structure has been
partially addressed, its size limitations have not. Using
87% aq KOH, MeOH, reflux for 2-4 h.
2
spatial requirements at C-7 of δ-selective opioid receptor
BNTX (1a ) as a starting point, a series of 10 7-
arylidenenaltrexone derivatives was synthesized. Both
the E- and Z-isomers of the 7-benzylidene compound
binding.
Resu lts a n d Discu ssion
(
1a ,b) were obtained via the standard aldol condensa-
Ch em istr y. Compounds 1a ,b-5a ,b were prepared
by aldol condensation of the appropriate aldehyde with
naltrexone (6) (Scheme 1). Liquid aldehydes were
converted to their 7-arylidene derivatives using a pro-
tion of naltrexone and benzaldehyde; however, the
Z-form was incompletely characterized due to the very
1
,2
low Z/E ratio of products (∼2/∼98).
This difficulty has
now been overcome through the use of a photochemical
isomerization process to convert E-olefins to the corre-
sponding Z-olefins. Using the methods employed by
cedure analogous to that for the formation of 7-benz-
1,2
ylidenenaltrexones 1a ,b.
Solid aldehydes were con-
densed with naltrexone using an adaptation of the
3
Lewis to photoisomerize cinnamamides, Z-isomer 1b
6
procedure reported by Rice. In all cases of these
was readily obtained from E-isomer 1a . Using the same
approach, the E- and Z-isomers of the aldol condensa-
tion products of naltrexone with a series of aldehydes
nonstereospecific aldol condensations, more of the steri-
cally less-encumbered E-isomer was produced than the
Z-isomer. E:Z isomeric ratios were on the order of >90:
<10. The yield of aldol condensation product decreased
with increasing size of the aromatic aldehyde, presum-
ably due to steric effects and perhaps due to the
solubility of the aldehyde in the reaction medium. The
photoisomerizations (Scheme 2) of the E- to Z-com-
pounds also showed a less-favorable E- to Z-photoequi-
(1-naphthaldehyde, 2-naphthaldehyde, 4-phenylbenzal-
dehyde, and 9-anthracaldehyde) were prepared. This
series is a useful battery of compounds to probe the
*
To whom correspondence should be addressed.
Abstract published in Advance ACS Abstracts, February 1, 1997.
X
S0022-2623(96)00573-0 CCC: $14.00 © 1997 American Chemical Society

7
50 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Palmer et al.
Sch em e 2
Ta ble 1. Ki’s of Synthetic 7-Arylidenenaltrexones in
Radioligand Displacement Assaysa
Ki (nM)
selectivity for
test ligand
total
µ
δ
κ
δ-sites (δ:µ ratio)
(
(
(
(
(
(
(
(
(
(
E)-benzyl (1a )
Z)-benzyl (1b)
E)-biphenyl (2a )
Z)-biphenyl (2b)
E)-1-naphthyl (3a ) 110
Z)-1-naphthyl (3b) 22
E)-2-naphthyl (4a ) 138
Z)-2-naphthyl (4b) 190
E)-anthracyl (5a ) 150
100
120
210
750
26
33
50
210
36
11
23
51
110
34
6.2 48
3.7 34
4.0
8.6
0.6
4.2
7.5
15
0.6
3.0
6.4
14
75
47
190
400
4.6 120
0.7 37
39
16
17
74
180
170
Z)-anthracyl (5b)
50
2.4 100
naltrexone (6)
2.6
0.8 11
1.1
0.1
a
Results are calculated based on the Cheng-Prusoff estima-
tion8 and the experimental determinations of KD reported herein.
IC50’s were calculated from duplicate samples ((10-15%) at nine
concentrations of 1.0-1000 nM test ligand. The displacing radio-
librium with increasing size of the 7-arylidene group.
The photoreactions went to photostationary states with
the E:Z ratios being 60:40 to 25:75. In all cases, only
trace amounts of the (E)-7-arylidenenaltrexone decom-
posed, allowing for reisolation of nonisomerized E-
isomer.
3
ligands were as follows: (-)-[9- H]bremazocine (0.5 nM) (KD )
3
3
2
4
1
.1 nM), total sites; [ H]DAMGO or [ H][D-Ala ,MePhe ,Gly-
5
3
ol ]enkephalin (1.0 nM) (KD ) 2.7 nM), µ-sites; [ H]DPDPE or
[ H][D-Pen ,D-Pen ]enkephalin (1.0 nM) (K ) 2.3 nM), δ-sites; and
3
2
5
D
3
3
[ H]U69,593 or [ H]-(5R,7R,8â)-(-)-N-methyl-N-(1-pyrrolidinyl-1-
oxaspiro[4.5]dec-8-yl)benzeneacetamide (1.0 nM) (KD ) 7.0 nM),
κ-sites. The slight differences in the KD’s reported herein and those
Stereochemical characterization of the individual
isomers was based on the chemical shifts of the signals
16
3
3
reported by Schultz ([ H]DAMGO ) 1 nM, [ H]DPDPE ) 4.7
nM, [3H]U69,593 ) 1.35 nM) are attributed to the differences in
1
of the vinyl protons in the NMR. In the H NMR spectra
membrane preparations from guinea pig brain homogenate versus
calf frontal cortex or calf caudate tissue.
of R,â-unsaturated carbonyl compounds, the chemical
shift of a â-vinyl proton of a Z-isomer appears upfield
7
very similar affinity profiles between 1a and 1b, with
Z-isomer 1b showing slightly greater δ-affinity and
δ-selectivity.
of the signal for the corresponding E-isomer. This
difference was observed in these E/Z isomeric pairs in
every case.
Op ioid Recep tor Affin ity. The affinities of the 10
The most δ-selective affinity profile was exhibited by
Z)-1-naphthyl compound 3b. This compound provided
(
7
-arylidene E/ Z pairs of compounds (1a ,b-5a ,b) at the
the lowest Ki’s of all four receptor affinities assayed. In
the total sites (bremazocine) and µ-sites (DAMGO)
assays, 3b has a 2-20-fold higher affinity than the other
nine test compounds. However, its affinity at these sites
is about 10-15 times less (total and µ-sites, respectively)
than that of naltrexone (6). In the assay for κ-sites
three opioid receptor sites (µ, δ, and κ) were determined
in a crude membrane preparation from guinea pig brain.
The radioligand displacement assays were determined
3
3
using [ H]bremazocine (total opioid sites), [ H]DAMGO
3
3
(
µ-sites), [ H]DPDPE (δ-sites), and [ H]U69,593 (κ-sites).
Naltrexone (6) was included as a standard in the assays.
The KD value for each of the radioligands determined
in the guinea pig brain homogenate are reported (Table
(U69,593), 3b has approximately equal affinity as the
7
-benzylidene compounds 1a ,b and 10 times higher
affinity than the least potent κ-ligand, the (Z)-4-biphenyl
compound 2b. At δ-receptors, 3b is bound with greater
affinity (3-fold) relative to the next closest compound,
1
). Both Scatchard plots and saturation curves were
derived and provided analogous results (not shown). An
average value of these two determinations is reported
as the KD. The derived KD’s from guinea pig brain
homogenate preparations were then used with the IC50’s
in the Cheng-Prusoff estimation to derive Ki’s for each
of the test ligands (Table 1). Comparison of the results
(Z)-9-anthracylidene 5b, and with more than 100 times
the affinity of the least potent δ-ligand, (E)-4-biphenyl
compound 2a . The δ-receptor affinity of 3b is also 15
times higher than that of naltrexone (6) and 5-9 times
higher than the (Z)- and (E)-benzylidene compounds
1b,a , respectively.
8
of these Ki determinations with those previously re-
ported for naltrexone (6) shows consistent values.
9
A
slight difference for the IC50 value for κ-sites was noted
The δ:µ selectivity demonstrated by the synthetic
compounds is greater by 2-7-fold for each of the
Z-isomers versus its corresponding E-isomer in all five
cases (Table 1). The (Z)-1-naphthyl analog 3b shows
the highest selectivity of the series with a δ:µ Ki ratio
of about 15. Benzylidene analogs 1a ,b exhibit only 25-
50% the δ:µ selectivity of 3b. Compound 3a , the (E)-
1-naphthylidene analog of 3b, shows only 50% of the
δ:µ selectivity of 3b. Interestingly, the analog with the
largest substituent, (Z)-anthracyl 5b, is nearly as
δ-selective as 3b, and its E-isomer (5a ) also shows
roughly 50% of its δ:µ selectivity.
but may be attributed to the use of different radio-
3
ligands (1.0 nM [ H]U69,593 in this work versus 0.5 nM
3
[
H]bremazocine in the presence of 100 nM DAMGO and
00 nM DPDPE in the earlier experiment).
1
None of the new compounds had an overall receptor
affinity as high as was observed for naltrexone (6). With
the exceptions of (E)-biphenyl compound 2a and (E)-2-
naphthyl compound 4a , all of the synthetic compounds
showed the highest affinity at the δ-sites. This is in
contrast to naltrexone (6) which showed lower affinity
at the δ-receptors than at the other opioid receptor
types. Though high δ-selectivity of the (E)-benzylidene
compound 1a in a guinea pig brain homogenate was
previously reported,1 we did not observe the same
result in our guinea pig brain preparation. We observed
The δ:κ selectivity ratios for all of the ligands were
generally higher than the δ:µ ratios (by almost 3-fold)
because all of the ligands showed less affinity for
κ-receptors than for µ-receptors, usually by about 3-fold.
,2

(E)- and (Z)-7-Arylidenenaltrexones
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 751
One exception was Z-isomer 1b, which showed a δ:κ
ratio approximately equal to its δ:µ ratio.
Com p u ta tion a l Ch em istr y. To gain some insight
about the expected conformations of the aryl groups
with respect to the olefin-carbonyl system in these
compounds, geometry optimization studies were per-
formed. In a previous study of 1a ,b, a molecular
2
mechanics force field method was used. We report the
use of a semiempirical computational approach that
includes a priori examination of electronic effects of the
conjugated system. The geometries of naltrexone and
the 10 arylidene derivatives were optimized with the
PM3 semiempirical method10 using SPARTAN 4.0 (Wave-
function, Inc., Irvine, CA). Comparison of the resulting
structures for relative conformational space sampling
was then accomplished by overlaying the rigid epoxy-
morphone moieties and examining that space sampled
by the aromatic groups at C-7. The minimized energies
of the Z-isomers were higher than those of their corre-
sponding E-counterparts by 2.8-5.6 kcal/mol. Although
the magnitude of the error associated with these cal-
culations ((∼1 kcal/mol) precluded analysis of specific
energy differences, a clear trend that larger aryl groups
tend to have larger E/Z energy differences is observed.
F igu r e 1. PM3-minimized structures of 3a (green) and 5b
(yellow) showing the different dispositions of the 7-arylidene
rings. The conformation exhibited by compound 5b appears
to be preferred for binding affinity and δ-selectivity.
similar in conformation and also possess the highest
binding affinity and greatest δ-selectivity suggest that
an aryl group fixed in this position is significant. The
difference in binding affinities and δ-selectivities be-
tween the benzyl (1a ,b) and extended 4-biphenyl (4a ,b)
compounds suggests that additional substitution on the
distal portion of the benzyl ring is not tolerated, despite
strong structural similarity.
In the minimized structures, as expected, none of the
aryl rings is coplanar with the olefin-carbonyl system.
A nearly perpendicular relationship between the aryl
group and the olefin-carbonyl system is found in each
case. The aryl ring is forced out of planarity with the
double bond and the carbonyl group by unfavorable
steric interactions. The degree of nonplanarity of the
Con clu sion
We have shown that the E-isomers of 7-arylidenen-
altrexone derivatives can be readily converted photo-
chemically to their corresponding Z-isomers without
degradation. Furthermore, the results obtained from
the opioid receptor binding assays demonstrate that the
Z-isomers of the 7-arylidene series are generally more
δ-selective than are the E-isomers in the radioligand
displacement profile in guinea pig brain homogenate
receptors. PM3 semiempirical geometry calculations
suggest a singificant role for the orientation of the
7
-arylidene group from the olefin-carbonyl plane was
quantitated in the minimized structures by measuring
the dihedral angle formed by C-7-benzyl carbon-first
aromatic carbon-second aromatic carbon. Planarity is
defined by angles of 0° or (180°, while the largest
difference is at (90°. This difference from planarity for
the E-compounds 1a -5a ranges from 68° in the case of
(E)-4-biphenyl compound 2a to 80° for (E)-9-anthracyl
compound 5a . The planarity deviations for the Z-
isomers 1b-5b ranged from -33° for (Z)-2-naphthyl
compound 4b to -81° for (Z)-9-anthracyl compound 5b,
the sign of the angle reflecting the E versus Z disposi-
tion of the 7-arylidene group. In the case of the
7
-arylidene substituent in the binding of these ligands.
The results also suggest that the spatial requirements
of the receptor will allow an extension to the lateral
portion of the benzylidene moiety (as in a naphthyl or
anthracyl group) to bind without difficulty. However,
the placement of an additional phenyl substituent on
the benzylidene in the position distal to the olefin, as
in a 4-biphenyl compound, is not tolerated. Overall, the
highest affinity and greatest δ-selectivity was observed
in (Z)-7-(1-naphthylidene)naltrexone (3b).
9
-anthracylidene compounds, the differences are virtu-
ally identical: Z-isomer 5b, -81°, and E-isomer 5a ,
80°. In all other cases, the Z-isomer has a smaller
-
deviation from planarity. It is also noteworthy that the
deviations are essentially identical in the benzyl series
1
6
a ,b and the 4-biphenyl series 2a ,b. Values of 69° and
8° were obtained for E-isomers 1a and 2a , respectively,
Exp er im en ta l Section
and -55° and -54° for Z-isomers 1b and 2b, respec-
tively. Deviations of 71° and -33° were obtained for
the (E)- (4a ) and (Z)-2-naphthylidenes (4b), respectively,
while deviations of 77° and -68° were obtained for the
Gen er a l Meth od s. Melting points were determined on a
Thomas-Hoover capillary melting point apparatus and are
uncorrected. Infrared spectra were recorded as thin film
depositions on NaCl plates with a Perkin-Elmer 1600 series
FTIR. Absorptions are expressed in wavenumbers (cm ). H
NMR spectra were recorded at 300 MHz on a Varian VXR-
-
1
1
(
E)- (3a ) and (Z)-1-naphthylidenes (3b), respectively.
The (E)-1-naphthyl (3a ) and (Z)-9-anthracyl (5b)
3
00 or Bruker AMX-300 spectrometer. Chemical shifts are
compounds show the spatial extremes in the minimized
structures. The aryl group in Z-isomer 5b extends
above and away from the epoxymorphinone nucleus,
while in E-isomer 3a , the aryl ring is in an orientation
forming a “pocket” with the epoxymorphinone (Figure
expressed in parts per million (δ) downfield from Si(CH ) .
3
4
Spectral assignments were supported by proton decoupling.
Mass spectra were obtained on VG-7070 and VG-70SEQ mass
spectrometers by direct-insertion probe. Analytical thin-layer
chromatography (TLC) was performed on Analtech silica gel
HLF TLC plates (0.25 mm thickness), and compounds were
visualized using a short wave UV lamp. TLC eluent was 95:
1
). The Z-isomer of the 1-naphthyl series (3b) occupies
a similar area of conformational space as the (Z)-9-
anthracyl compound 5b. That these two compounds are
5
:0.5 CH
2 2 4
Cl :MeOH:NH OH, unless otherwise indicated. Mer-
ck silica gel 60 (230-400 mesh) was used for preparative flash

7
52 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Palmer et al.
chromatography. All reactions were performed under an argon
atmosphere unless otherwise noted.
C-19′/20′ H); yield 64% as a yellow solid, mp 179-181 °C; FTIR
-
1
(neat) 1692 (CdO), 1620 (CdC) cm ; HRFABMS calcd for
Syn th etic P r oced u r es: Ald ol Con d en sa tion s. All liq-
uid aryl aldehydes were condensed with naltrexone using
adaptations of a previously reported method (procedure a,
C
(C33
33
H
H
32
O
4
N
4
(M
N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(E)-(1-n a p h t h -
+ H) 506.2331, obsd 506.2313. Anal.
31
O
1
,2
1
Scheme 1). Solid aldehydes were condensed with naltrexone
using the procedure outlined below (procedure b, Scheme 1).6
To 340 mg (1.0 mmol) of naltrexone (6) free base in 30 mL
of methanol were added 456 mg (8.0 mmol) of KOH and 6.0
mmol of the appropriate aldehyde. The mixture was heated
to reflux under an argon atmosphere for 2-4 h, during which
time the solution turned a deep orange. The solution was
poured into 100 mL of water, extracted with EtOAc (3 × 50
mL), and washed with brine (1 × 75 mL). The organic layers
3
ylid en e)-3,14-d ih yd r oxym or p h in a n (3a ): H NMR (CDCl )
δ 8.18 (s, 1 H, vinyl H), 7.92 (m, 1 H), 7.84 (m, 2 H), 7.50 (m,
2 H), 7.44 (d, J ) 8.2 Hz, 1 H), 7.34 (d, J ) 7.2 Hz, 1 H), 6.76
(d, J ) 8.2 Hz, 1 H, C-2 H), 6.63 (d, J ) 8.3 Hz, 1 H, C-1 H),
4.77 (s, 1 H, C-5 H), 1.67 (d, J ) 13.4 Hz, 1 H, C-15′ H), 0.79
(m, 1 H, C-18 H), 0.50 (d, J ) 7.7 Hz, 2 H, C-19/20 H), 0.09 (d,
J ) 4.8 Hz, 2 H, C-19′/20′ H); yield 64% as a yellow solid, mp
-
1
208-210 °C; FTIR (neat) 1686 (CdO), 1611 (CdC) cm
HRFABMS calcd for C31 N (M + H) 480.2175, obsd
480.2151. Anal. (C31 N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(Z)-(1-n a p h t h -
;
30 4
H O
H O
29 4
were combined and dried over Na
2
SO
4
, and the solvent was
evaporated. Purification of the arylidene isomeric mixture was
effected by column chromatography (SiO ; CH Cl :MeOH:NH
1
2
2
2
4
-
3
ylid en e)-3,14-d ih yd r oxym or p h in a n (3b): H NMR (CDCl )
OH, 95:5:0.5). The final separation of the E- and Z-isomers
was accomplished by using multiple elutions on preparative
TLC plates (SiO , 2.0 mm thickness) in the same solvent
2
system. Yields for the aldol condensation ranged 36-75% with
the Z-isomer present only in trace (<10%) amounts. Reported
yields for the E-isomers represent yields from the aldol
condensations; reported yields for the Z-isomers represent
yields from the photoisomerization reactions (with the balance
of the yield being unconverted E-isomer).
δ 7.99 (m, 1 H), 7.79 (m, 2 H), 7.44 (m, 5 H), 7.19 (s, 1 H, vinyl
H), 6.70 (d, J ) 8.0 Hz, 1 H, C-2 H), 6.61 (d, J ) 8.0 Hz, 1 H,
C-1 H), 4.62 (s, 1 H, C-5 H), 1.56 (d, J ) 10.9 Hz, 1 H, C-15′
H), 0.88 (m, 1 H, C-18 H), 0.57 (d, J ) 7.6 Hz, 2 H, C-19/20
H), 0.17 (d, J ) 4.4 Hz, 2 H, C-19′/20′ H); yield 68% as a yellow
solid, mp 135-138 °C; FTIR (neat) 1704 (CdO), 1620 (CdC)
-1
cm ; HRFABMS calcd for C31
480.2161. Anal. (C31 N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(E)-(2-n a p h t h -
30 4
H O N (M + H) 480.2175, obsd
29 4
H O
1
P h otoisom er iza tion . In a Pyrex test tube (20 mm × 150
3
ylid en e)-3,14-d ih yd r oxym or p h in a n (4a ): H NMR (CDCl )
mm) 100 mg of the appropriate (E)-arylidene ketone was
δ 7.83 (s, 1 H, vinyl H), 7.81 (d, J ) 8.2 Hz, 4 H), 7.42-7.52
(m, 3 H), 6.76 (d, J ) 8.3 Hz, 1 H, C-2 H), 6.66 (d, J ) 7.9 Hz,
1 H, C-1 H), 4.73 (s, 1 H, C-5 H), 1.67 (d, J ) 11.1 Hz, 1 H,
C-15′ H), 0.83 (m, 1 H, C-18 H), 0.53 (d, J ) 8.1 Hz, 2 H, C-19/
20 H), 0.12 (d, J ) 4.9 Hz, 2 H, C-19′/20′ H); yield 57% as a
yellow solid, mp 271-273 °C; FTIR (neat) 1682 (CdO), 1587
2 2
dissolved in 25 mL of reagent grade CH Cl . The test tube
was sealed with a latex septum and purged with dry nitrogen
for 3-5 min. The test tube was then irradiated (Hanovia Hg
Arc lamp) for 12-14 h with cooling such that the reaction
temperature did not exceed 30 °C. The solvent was evaporated
-
1
and the resultant brown oil chromatographed (SiO
MeOH:NH OH, 95:5:0.5). Separation of the E- and Z-isomers
of the 7-arylidene product was accomplished by using multiple
elutions on preparative TLC plates (SiO , 2.0 mm thickness)
in the same solvent system. Photoequilibrium ratios were 40:
0 to 25:75, E:Z (except in the case of the 9-anthracyl
compound 5b, where a ratio of ca. 60:40 E:Z was observed).
2
; CH
2
Cl
2
:
(CdC) cm ; HRFABMS calcd for C31
obsd 480.2162. Anal. (C31 N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(Z)-(2-n a p h t h -
30 4
H O N (M + H) 480.2175,
4
29 4
H O
1
2
3
ylid en e)-3,14-d ih yd r oxym or p h in a n (4b): H NMR (CDCl )
δ 7.92 (s, 1 H), 7.77 (m, 3 H), 7.46 (m, 3 H), 6.82 (s, 1 H, vinyl
H), 6.73 (d, J ) 7.8 Hz, 1 H, C-2 H), 6.62 (d, J ) 7.8 Hz, 1 H,
C-1 H), 4.83 (s, 1 H, C-5 H), 1.61 (d, J ) 11.8 Hz, 1 H, C-15′
H), 0.87 (m, 1 H, C-18 H), 0.56 (d, J ) 7.3 Hz, 2 H, C-19/20
H), 0.15 (d, J ) 3.6 Hz, 2 H, C-19′/20′ H); yield 70% as a yellow
solid, mp 163-165 °C; FTIR (neat) 1693 (CdO), 1614 (CdC)
6
1
7-(Cyclop r op ylm eth yl)-4,5-r-ep oxy-7(E)-ben zylid en e-
1
3
1
,14-d ih yd r oxym or p h in a n (1a ): H NMR (CDCl
H, vinyl H), 6.74 (d, J ) 8.2 Hz, 1 H, C-2 H), 6.64 (d, J ) 8.2
3
) δ 7.68 (s,
-
1
Hz, 1 H, C-1 H), 4.70 (s, 1 H, C-5 H), 1.67 (d, J ) 11.9 Hz, 1
H, C-15′ H), 0.84 (m, 1 H, C-18 H), 0.55 (d, J ) 8.3 Hz, 2 H,
C-19/20 H), 0.13 (d, J ) 4.9 Hz, 2 H, C-19′/20′ H); yield 75%
as a yellow solid, mp 115-118 °C; FTIR (neat) 1685 (CdO),
cm ; HRFABMS calcd for C31
480.2167. Anal. (C31 N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(E)-(9-a n t h r a -
30 4
H O N (M + H) 480.2175, obsd
29 4
H O
1
3
cyliden e)-3,14-dih ydr oxym or ph in an (5a): H NMR (CDCl )
-
1
1
4
593 (CdC) cm ; HRFABMS calcd for C27
30.2018, obsd 430.2012. Anal. (C27 N) C, H, N.
7-(Cyclop r op ylm eth yl)-4,5-r-ep oxy-7(Z)-ben zylid en e-
H
28
O
4
N (M + H)
δ 8.45 (s, 1 H, vinyl H), 8.27 (s, 1 H), 8.15 (m, 1 H), 7.99 (m,
2 H), 7.52 (m, 3 H), 7.43 (m, 1 H), 7.33 (m, 1 H), 6.78 (d, J )
8.3 Hz, 1 H, C-2 H), 6.58 (d, J ) 8.3 Hz, 1 H, C-1 H), 4.83 (s,
1 H, C-5 H), 1.66 (d, J ) 12.0 Hz, 1 H, C-15′ H), 0.89 (m, 1 H,
C-18 H), 0.42 (d, J ) 7.3 Hz, 2 H, C-19/20 H), 0.22 (d, J ) 4.8
Hz, 1 H, C-19′/20′ H); yield 36% as an orange solid, mp 200-
27 4
H O
1
1
3
2
1
3
,14-d ih yd r oxym or p h in a n (1b): H NMR (CDCl ) δ 7.43 (m,
H), 7.28 (m, 3 H), 6.73 (d, J ) 8.3 Hz, 1 H, C-2 H), 6.66 (s,
H, vinyl H), 6.62 (d, J ) 8.3 Hz, 1 H, C-1 H), 4.78 (s, 1 H,
-1
C-5 H), 1.61 (d, J ) 12.6 Hz, 1 H, C-15′ H), 0.88 (m, 1 H, C-18
H), 0.57 (d, J ) 8.3 Hz, 2 H, C-19/20 H), 0.16 (d, J ) 4.2 Hz,
203 °C; FTIR (neat) 1690 (CdO), 1613 (CdC) cm ; HRFABMS
calcd for C35 N (M + H) 530.2331, obsd 530.2337. Anal.
(C35 N) C, H, N.
17-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(Z)-(9-a n t h r a -
32 4
H O
2
°
H, C-19′/20′ H); yield 75% as a yellow solid, mp 171-172
31 4
H O
-
1
C; FTIR (neat) 1692 (CdO), 1614 (CdC) cm ; HRFABMS
calcd for C27 N (M + H) 430.2018, obsd 430.2014. Anal.
N) C, H, N.
7-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(E)-(4-p h en yl-
1
H
28
O
4
3
cyliden e)-3,14-dih ydr oxym or ph in an (5b): H NMR (CDCl )
(C
27
H
27
O
4
δ 8.40 (d, J ) 8.3 Hz, 1 H), 8.35 (s, 1 H), 7.95 (d, J ) 7.5 Hz,
2 H), 7.35 (m, 5 H), 6.63 (d, J ) 8.1 Hz, 1 H, C-2 H), 6.61 (s,
1 H, vinyl H), 6.56 (d, J ) 8.1 Hz, 1 H, C-1 H), 4.43 (s, 1 H,
C-5 H), 1.76 (d, J ) 12.4 Hz, 1 H, C-15′ H), 0.85 (m, 1 H, C-18
H), 0.60 (d, J ) 8.2 Hz, 1 H, C-19/20 H), 0.19 (d, J ) 4.3 Hz,
1 H, C-19′/20′ H); yield 40% as a brown solid, mp 195-197 °C;
1
ben zylid en e)-3,14-d ih yd r oxym or p h in a n (2a ): 1H NMR
(
6
CDCl
3
) δ 7.70 (s, 1 H, vinyl H), 7.58 (m, 4 H), 7.42 (m, 5 H),
.76 (d, J ) 7.5 Hz, 1 H, C-2 H), 6.64 (d, J ) 7.7 Hz, 1 H, C-1
H), 4.72 (s, 1 H, C-5 H), 1.66 (d, J ) 11.5 Hz, 1 H, C-15′ H),
0
0
-
1
.84 (m, 1 H, C-18 H), 0.53 (d, J ) 7.7 Hz, 2 H, C-19/20 H),
.13 (d, J ) 4.2 Hz, 2 H, C-19′/20′ H); yield 70% as a yellow
FTIR (neat) 1690 (CdO), 1620 (CdC) cm ; HRFABMS calcd
for C35
(C35
H
O
32
O
4
N (M + H) 530.2331, obsd 530.2328. Anal.
solid, mp 133-136 °C; FTIR (neat) 1686 (CdO), 1639 (CdC)
H
31
4
N) C, H, N.
-1
cm ; HRFABMS calcd for C33
06.2317. Anal. (C33 N) C, H, N.
7-(Cyclop r op ylm et h yl)-4,5-r-ep oxy-7(Z)-(4-p h en yl-
H
32
O
4
N (M + H) 506.2331, obsd
Biologica l Testin g. A Brandel harvestor and FP-100
Whatman GF/B fired filter paper were used for protein
filtration. The filter paper for κ-receptor binding was pre-
treated with aqueous 0.1% poly(ethylenimine) to coat the glass
fibers.12 All glassware used in the affinity assay was silanized
with Prosil-28. Polypropylene culture tubes and scintillation
vials were used in all binding assays.
5
31 4
H O
1
1
ben zylid en e)-3,14-d ih yd r oxym or p h in a n (2b): H NMR
CDCl ) δ 7.58 (d, J ) 8.3 Hz, 2 H), 7.53 (s, 4 H), 7.43 (m, 2
(
3
H), 7.35 (m, 1 H), 6.73 (d, J ) 8.3 Hz, 1 H, C-2 H), 6.68 (s, 1
H, vinyl H), 6.62 (d, J ) 8.3 Hz, 1 H, C-1 H), 4.80 (s, 1 H, C-5
H), 1.60 (d, J ) 11.2 Hz, 1 H, C-15′ H), 0.88 (m, 1 H, C-18 H),
The binding assays were carried out with slight modifica-
1
3
0
.57 (d, J ) 8.1 Hz, 2 H, C-19/20 H), 0.17 (d, J ) 4.9 Hz, 2 H,
tions to the procedures described by Lin and Simon and

(E)- and (Z)-7-Arylidenenaltrexones
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 753
Werling.14 Hartley-VAF Plus guinea pigs (300-350 g) were
sacrificed by decapitation. The brain, less cerebellum, was
removed and homogenized in 6 vol of 0.05 M Trizma buffer
Antagonists. J . Med. Chem. 1988, 31, 281-282. Portoghese, P.
S.; Sultana, M.; Takemori, A. E. Design of Peptidomimetic
δ-Opioid Receptor Antagonists Using the Message Address
Concept. J . Med. Chem. 1990, 33, 1714-1720. Takemori, A. E.;
Portoghese, P. S. Selective Naltrexone-Derived Opioid Receptor
Antagonists. Annu. Rev. Pharmacol. Toxicol. 1992, 32, 239-269.
(
7
pH 7.4) with a Virtishear homogenizer at a control setting of
0, for three 5-s intervals. The homogenate was centrifuged
at 25000g at 4 °C for 20 min. The pellet was resuspended in
vol of aqueous 0.32 M sucrose and stored at -70 °C until
(
6) Bertha, C. M.; Mattson, M. V.; Flippen-Andersen, J . L.; Roth-
man, R. B.; Xu, H.; Cha, X.-Y.; Becketts, K.; Rice, K. C. A Marked
Change of Receptor Affinity of the 2-Methyl-5-(3-hydroxyphenyl)-
morphans upon Attachment of an (E)-8-Benzylidene Moiety:
Synthesis and Evaluation of a New Class of σ Receptor Ligands.
J . Med. Chem. 1994, 37, 3163-3170.
6
needed. Frozen homogenate was thawed at room temperature
and diluted with 0.05 M Trizma buffer (pH 7.4) to give a final
dilution ratio of 1:60 (initial brain weight:total solution
volume). This corresponded to a final protein concentration
of 0.8-1.5 mg of protein/mL of homogenate as determined
(
7) Heathcock, H. C.; Buse, T. C.; Kleschick, A.; Pirrung, C. M.; Soh,
E. J .; Lamp, J . Acyclic Stereoselection. 7. Stereoselective Syn-
thesis of 2-Alkyl-3-hydroxy Carbonyl Compounds by Aldol
Condensation. J . Org. Chem. 1980, 45, 1066-1081.
1
5
according to the method of Ohnishi and Barr by use of a
Sigma Diagnostics Micro Protein Determination Kit.
3
Radioligands used were [ H]bremazocine (0.5 nM) for total
(
8) Cheng, Y.-C.; Prusoff, W. H. Relationship Between the Inhibition
3
ligand binding, [ H]DAMGO (1.0 nM) for µ-receptor binding,
Constant (K
i
) and the Concentration of Inhibitor Which Causes
50 per cent Inhibition (IC50) of an Enzymatic Reaction. Biochem.
3
3
[
(
H]DPDPE (1.0 nM) for δ-receptor binding, and [ H]U69,593
1.0 nM) for κ-receptor binding. Synthetic ligands were tested
Pharmacol. 1973, 22, 3099-3108.
3
in duplicate at nine concentrations between 1.0 and 1000 nM.
Nonspecific binding was measured in the presence of 10 µM
naloxone. The samples were incubated for 60 min at 25 °C.
Samples were filtered, rinsed with ice-cold buffer (3 × 2 mL),
and eluted with 10 mL aliquots of Aquasol, and the radioactiv-
ity was counted. Specific binding for each concentration was
calculated, and the data were analyzed by probit transforma-
tion and linear regression to obtain the IC50. The IC50’s were
(9) Klein, P.; Nelson, W. L. O -(2-Carbomethoxyallyl) Ethers of
Opioid Ligands Derived from Oxymorphone, Naltrexone, Etor-
phine, Diprenorphine, Norbinaltorphimine, and Naltrindole.
3
Unexpected O -Dealkylation in the Opioid Radioligand Displace-
ment Assay. J . Med. Chem. 1992, 35, 4589-4594.
(
10) Stewart, J . J . P. Optimization of Parameters for Semiempirical
Methods. I. Method. J . Comput. Chem. 1989, 10, 209-220.
11) These calculated differences appear to be reasonable ones as
related styryl systems cis-stilbene have been shown experimen-
tally to be 2.3 kcal/mol higher in energy than trans-stilbene;
see: Saltiel, J .; D’Agostino, J . T.; Megarity, E. D.; Metts, L.;
Neuberger, K. R.; Wrighton, M.; Zafiriow, O. C. Cis-trans
Photoisomerization of Olefins. Org. Photochem. 1973, 3, 1-113.
Saltiel, J .; Charlton, J . L. In Rearrangements in Ground and
Excited States; de Mayo, P., Ed.; Academic Press: New York,
(
then transformed into the reported K
Prusoff equation.
i
’s using the Cheng-
8
Ack n ow led gm en t. We acknowledge the support of
this work by the National Institute on Drug Abuse
through Training Grant DA07278 and National Insti-
tutes of Health Research Grant DA06675.
1
980; Vol. 3, Chapter 14.
3
(12) Houghton, R. A.; J ohnson, N.; Pasternak, G. W. [ H]-Beta-
Endorphin Binding in Rat Brain. J . Neurosci. 1984, 4, 2460-
2
465.
Refer en ces
(
13) Lin, H.-K.; Simon, E. J . Phospholipase A Inhibition of Opiate
(
1) Portoghese, P. S.; Sultana, M.; Nagase, H.; Takemori, A. E. A
-Opioid Receptor Antagonist: 7-benzylidene-
Receptor Binding Can Be Reversed by Albumin. Nature (London)
Highly Selective δ
1
1
978, 271, 383-384.
naltrexone. Eur J . Pharmacol. 1992, 218, 195-196.
(
14) Werling, L. L.; Zarr, G. D.; Brown, S. R.; Cox, B. M. Opioid
Binding to Rat and Guinea-Pig Neural Membranes in the
Presence of Physiological Cations at 37 °C. J . Pharmacol. Exp.
Ther. 1985, 233, 722-728.
(
2) Portoghese, P. S.; Sultana, M.; Moe, S. T.; Takemori, A. E.
Synthesis of Naltrexone-Derived δ-Opioid Antagonists. Role of
Conformation of the δ Address Moiety. J . Med. Chem. 1994, 37,
5
79-585.
(
15) Ohnishi, S. T.; Barr, J . K. A Simplified Method of Quantitating
(
3) Lewis, F. D.; Elbert, J . E.; Upthagrove, A. L.; Hale, P. D.
Structure and Photoisomerization of (E)- and (Z)-Cinnamamides
and Their Lewis Acid Complexes. J . Org. Chem. 1991, 56, 553-
Proteins Using the Biuret and Phenol Reagents. Anal. Biochem.
1
978, 86, 193-200.
5
61.
(16) Schultz, A. G.; Wang, A.; Alva, C.; Sebastian, A.; Glick, S. D.;
Deecher, D. C.; Bidlack, J . M. Asymmetric Syntheses, Opioid
Receptor Affinities, and Antinociceptive Effects of 8-Amino-5,9-
methanobenzocyclooctenes, a New Class of Structural Analogues
of the Morphine Alkaloids. J . Med. Chem. 1996, 39, 1956-1966.
(
4) Davis, R. D.; Nelson, W. L. Isothiocyanate-Substituted Benzyl
Ether Opioid Receptor Ligands Derived from 6â-Naltrexol. J .
Med. Chem. 1995, 38, 570-579.
(
5) Portoghese, P. S.; Sultana, M.; Nagase, H.; Takemori, A. E.
Application of the Message-Address Concept in the Design of
Highly Potent and Selective Non-Peptide δ-Opioid Receptor
J M960573F