Understanding the structural requirements of 4-anilidopiperidine analogues for biological activities at μ and δ opioid receptors

Article abstract of DOI:10.1016/j.bmcl.2007.01.114

New 4-anilidopiperidine analogues in which the phenethyl group of fentanyl was replaced by several aromatic ring-contained amino acids (or acids) were synthesized to study the biological effect of the substituents on μ and δ opioid receptor interactions. These analogues showed broad (47 nM-76 μM) but selective (up to 17-fold) binding affinities at the μ opioid receptor over the δ opioid receptor, as predicted from the message-address concept.

Full text of DOI:10.1016/j.bmcl.2007.01.114

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2161–2165
Understanding the structural requirements of 4-anilidopiperidine
analogues for biological activities at l and d opioid receptors
Yeon Sun Lee,a Joel Nyberg,a Sharif Moye,a Richard S. Agnes,a Peg Davis,b Shou-wu Ma,b
Josephine Lai,b Frank Porreca,b Ruben Vardanyana and Victor J. Hrubyb,*
aDepartment of Chemistry, University of Arizona, Tucson, AZ 85721, USA
bDepartment of Pharmacology, University of Arizona, Tucson, AZ 85724, USA
Received 22 December 2006; revised 25 January 2007; accepted 25 January 2007
Available online 8 February 2007
Abstract—New 4-anilidopiperidine analogues in which the phenethyl group of fentanyl was replaced by several aromatic ring-con-
tained amino acids (or acids) were synthesized to study the biological effect of the substituents on l and d opioid receptor interac-
tions. These analogues showed broad (47 nM–76 lM) but selective (up to 17-fold) binding affinities at the l opioid receptor over the
d opioid receptor, as predicted from the message-address concept.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Since the discovery of fentanyl1 which was shown to
have potent agonist activity at the l opioid receptor with
unique biological properties, many anilidopiperidine
analogues have been synthesized and evaluated for
structure–activity relationships (SAR) to provide insight
into the key structural features for high biological activ-
ity for the l opioid receptor.2–4 In general, the minimal
structural requirements necessary for a ligand to interact
with the l opioid receptor are an appropriate spatially
oriented basic nitrogen and a hydroxylated phenyl ring.5
According to the concept of message-address domains,
the tertiary nitrogen and the phenolic hydroxy group
in nonpeptide opioid ligands, and the free amine and
the hydroxyl groups of tyrosine residue in peptide li-
gands constitute a part of the message domain to anchor
the ligands to opioid receptors.6
substituents on the d and l opioid receptors. The design
of these analogues was based on the message-address
concept. In Figure 1, the upper part of fentanyl moiety
(left), N-phenyl-N-piperidin-4-yl-propionamide, was
considered as an address region, and the lower part of
Dmt-Tic (right), namely the 2,6-dimethyltyrosine
(Dmt) moiety, a message region. By combination of
the two parts shown (center) it was expected that the
analogues would show a broad range of biological activ-
ities, since the modification on the C-terminus of Dmt-
Tic was recognized to result in analogues that display
a variety of l/d selectivities and agonist/antagonist com-
binations.8 Considering the correlation between the
structures of fentanyl and Dmt-Tic, and their bioactivi-
ties, we decided to examine thoroughly other modifica-
tions with the aim of evaluating the structural
requirements for binding to the opioid receptors, and
identifying the variations in the receptor binding pocket.
In this study, eighteen 4-anilidopiperidine analogues
1–18 in which the piperidine nitrogen was coupled with
different kinds of aromatic ring-contained amino acids
(or acids) were prepared and screened for their binding
affinities and functional activities at the d and l opioid
receptors.
Recently our group has been developing several piperi-
dine derivatives which can be ligated with various kinds
of amino acids to create a new series of 4-anilidopiperi-
dine analogues.7 Here new analogues in which the phen-
ethyl group of fentanyl was replaced by several aromatic
ring-contained amino acids (or acids), N-(1-substituted
piperidin-4-yl)-N-phenyl-propionamide, were designed
and synthesized to study the biological effect of these
All analogues were prepared by coupling various kinds
of aromatic ring-contained amino acids (or acids) to
19 (see Note 9 for the synthesis of 19) using BOP/
HOBt/NMM in solution (Scheme 1). After deprotection
of Na-Boc group with TFA, crude compounds were
purified by preparative RP-HPLC to give >95% pure
Keywords: 4-Anilidopiperidine analogues; Fentanyl; Dmt-Tic; Opioid
receptors; Analgesic effects.
*
Corresponding author. Tel.: +1 520 621 6332; fax: +1 520 621
8407; e-mail: hruby@u.arizona.edu
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.01.114

2162
Y. S. Lee et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2161–2165
O
O
N
N
N
N
address
region
COOH
NH2
N
NH2
O
O
R2
X
R2
R1
R1
message
region
R2
R2
fentanyl
X
R1, R2=Me, X=OH,1(S-form),2(R-form)
R1=H, R2=Me, X=OH, 3
R1, R2=H, X=OH, 4
R=Me, Tmt-Tic
R=H, Dmt-Tic
R1, R2=H, X=H, 5
Figure 1. Opioid ligands.
O
dimethyltyrosine (Tmt)11 residue rather than Dmt
showed the highest agonist activity in the GPI assay
with no antagonist function, but low agonist activity
in the MVD with some antagonist function (5-fold
[3H]DPDPE shift at 1 lM), but showed low binding
affinities at both d and l opioid receptors in the
300 nM range. Although there is no evidence that com-
pounds 2 and 3 have potent antagonist activity, it is sug-
gested that their partial antagonist activity could be
responsible for its reduced agonist activity in the same
assay. A stereo difference was found for the opioid activ-
ities of 1 [R1 = (S)-NH2, R2 = (S)-Me] and 2 [R1 = (S)-
NH2, R2 = (R)-Me]. Compound 2 showed higher bind-
ing affinities for both opioid receptors and was more
biologically active for the l opioid receptor. Our results
are consistent with the known fact that the preferred ste-
reo configuration of the Tmt residue for the d opioid
receptor interaction is 2S and 3R,11 even though signif-
icant d opioid agonist activity was not shown due to its
potential interruption by some antagonist activity. It
was also shown that R-configuration of R1 substituents
was not tolerated in the structure 6 for the d opioid
receptor (Fig. 2).
R4
( )n
O
O
N
R2
Ar
a, b
N
N
+
Ar
OH
O
19
R3
R1
1-18
N
H
n = 0, 1; R1 = NH2, H; R2 = NH2, Me, H; R3= NHBoc, H; R4 = NHBoc, Me, H
Ar = 2, 6-dimethyl-4-hydroxyphenyl, 4-hydroxyphenyl, phenyl, 4-chlorophenyl,
4-methylphenyl, 4-methoxyphenyl, 4-thiazoyl, 3-benzothiophenyl, 2-indanyl,
1-naphtalenyl, 3-chloro-4-methoxyphenyl, 3-bromophenyl
Scheme 1. Synthesis of 4-anilidopiperidine analogues. Reagents and
conditions: (a) 1.1 equiv BOP, 1.1 equiv HOBt, 2.2 equiv NMM, rt, 2–
4 h, 85–95%; (b) TFA, 0 ꢁC, 20 min, quantitative.
analogues in high yields (see Table 1 for analytical data).
Their opioid binding affinities at the human d opioid
receptor (hDOR) and the rat l opioid receptor (rMOR)
were determined by competition analyses against
[3H]DPDPE (d) and [3H]DAMGO (l) using membrane
preparations from transfected HN9.10 cells that consti-
tutively express the respective receptors. Functional
activities for d and l opioid receptors were evaluated
in stimulated isolated mouse vas deferens (MVD, d)
and guinea pig isolated ileum (GPI, l) bioassays, respec-
tively, as previously published.10
Among the analogues, 16 [n = 0, R1 = (S)-NH2] and 17
[n = 0, R1 = (R)-NH2] which have one less carbon be-
tween the piperidine and substituted aromatic rings than
the others [n = 1] lost their binding affinities at the d
opioid receptor, while 18 [n = 0, R1 = (S)-NH2, Ar =
2-indanyl] retained its binding affinity at the same recep-
tor. The losses of binding affinities at the d opioid recep-
tor could be caused by the shortened length of the
molecule interrupting the ability (potential flexibility)
to access the binding pocket. On the other hand, the
2-indanyl group conferred greater binding to the d opi-
oid receptor by extending the length of molecule
through the cyclopentenyl moiety, which could play an
important role in holding the piperazine ring as a rigid
spacer. It is apparent that the length of molecules, espe-
cially between the piperazine and substituted aromatic
rings, can be considered to be an important factor which
contributes to the binding. Hydrophobic aromatic ring
In Table 2, the synthesized analogues showed a broad
range (47 nM to 76 lM) but selective (up to 17-fold)
binding affinities for the l opioid receptor over the d
opioid receptor. Their selectivities for the l opioid
receptor were predicted from the address region of the
fentanyl structure (Fig. 1). Opioid agonist activities in
the MVD and the GPI assays were very low and did
not correlate with the binding affinities at both recep-
tors. One of the analogues, 3, in which Dmt was coupled
to the piperidine ring showed high binding affinities at
both d and l opioid receptors (Ki = 50 nM and 47 nM,
respectively) with very low agonist activities in the
MVD and the GPI assays. Interestingly, this compound
showed some l opioid antagonist activity by its ability
to shift the dose response of PL-017 twofold in the
GPI assay. Compound 2 which had a b–methyl-20,60-
substituted analogues,
9
[R1 = (S)-NH2, R2 = H,
Ar = naphthyl] and 10 [R1 = (S)-NH2, R2 = Ph,

Y. S. Lee et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2161–2165
Table 1. Analytical data of the 4-anilidopiperidine analogues
2163
Compound
Molecular formula
LowMSa
HRMSb
HPLCc (tR, min)
Observed
Observed
Calcd
1
2
C
C
26H35N3O3
26H35N3O3
438.3
438.4
424.5
396.2
380.5
414.1
387.4
436.5
430.5
456.6
414.4
394.5
410.5
443.3d
429.4d
366.2
366.2
406.5
438.2741
438.2763
424.2595
396.2296
380.2342
414.1949
387.1851
436.2047
430.2489
456.2671
414.1967
394.2489
410.2437
443.1339d
429.1939d
366.2185
366.2179
406.2501
438.2757
438.2757
424.2600
396.2287
380.2338
414.1948
387.1855
436.2059
430.2495
456.2651
414.1948
394.2495
410.2444
443.1334d
429.1945d
366.2182
366.2182
406.2495
18.5
16.0
13.6
18.5
19.1
17.3
12.3
17.1
16.4
19.0
17.6
18.7
18.0
22.0
24.5
14.6
16.6
20.3
3
C25H33N3O3
C23H29N3O3
C23H29N3O2
4
5
6
C23H28ClN3O2
7
C20H26N4O2S
C26H29N3O2S
C27H31N3O2
8
9
10
11
12
13
14
15
16
17
18
C29H33N3O2
C23H28ClN3O2
C24H31N3O2
C
C
24H31N3O3
23H27BrN2O2
C24H29ClN2O3
C22H28N3O2
C
C
22H28N3O2
25H31N3O2
a (MꢀTFA+H)+, ESI method [Finnigan, Thermoelectron, LCQ classic].
b (MꢀTFA + H)+, FAB-MS (JEOL HX110 sector instrument) or MALDI-TOF.
c Performed on a Hewlett Packard 1100 [C-18, Vydac, 4.6 mm · 250 mm, 5 lm, 10–90% of acetonitrile containing 0.1% TFA within 40 min and up to
100% within additional 5 min, 1 mL/min].
d (M+H)+.
Table 2. Bioactivities of 4-anilidopiperidine analogues at the d and l opioid receptors10
O
R2
1-18
N
N
O
R1
Ar
R1
R2
Ar
Binding Kia (lM)
Bioassay IC50 (lM)
b
Compound
n
hDOR (d) rMOR(l) MVD(d)
GPI(l)
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
(S) NH2 (S) Me 2,6-Dimethyl-4-hydroxyphenyl 1.7
(S) NH2 (R) Me 2,6-Dimethyl-4-hydroxyphenyl 0.26
0.50
0.30
0.047
7.9
39% at 1 lMc
13% at 1 lMc
2% at 1 lMc
18% at 1 lMc
0% at 1 lMc
28% at 1 lMc
8% at 1 lMc
0% at 10 lMc
26% at 1 lMc
48% at 10 lMc
12% at 1 lMc
27% at 10 lMc
19% at 1 lMc
0% at 1 lMc
0% at 1 lMc
7% at 1 lMc
12% at 1 lMc
19% at 10 lMc
3% at 1 lMc
3.3
3
(S) NH2
(S) NH2
(S) NH2
(R) NH2
(S) NH2
(S) NH2
(S) NH2
H
H
H
H
H
H
H
2,6-Dimethyl-4-hydroxyphenyl 0.050
4-Hydroxy phenyl
Ph
45% at 30 lMc
8% at 1 lMc
2% at 1 lMc
0% at 1 lMc
3% at 1 lMc
28% at 10 lMc
8% at 1 lMc
73% at 10 lMc
5% at 1 lMc
24% at 10 lMc
3% at 1 lMc
18% at 1 lMc
33% at 10 lMc
3% at 1 lMc
1% at 1 lMc
59% at 10 lMc
4
16
5
6.9
nc
3.9
6.9
6
4-Chlorophenyl
4-Thiazoyl
7
76
15
9.7
2.6
8
3-Benzothiopheneyl
2-Naphthyl
Ph
9
3.3
6.5
2.4
2.9
9.0
12
1.1
10
11
12
13
14
15
16
17
18
(S) NH2 Ph
H
0.40
0.99
3.0
NH2
4-Chlorophenyl
4-Methylphenyl
4-Methoxyphenyl
3-Bromophenyl
3-Chloro-4-methoxyphenyl
Ph
H
NH2
NH2
H
H
9.7
1.9
H
H
H
6.3
nc
3.8
6.4
(S) NH2
(R) NH2
(S) NH2
Ph
2-Indanyl
nc
15
5.1
2.6
(2S,3R)-Tmt-Tic11
Fentanyl12
0.0093
0.57
0.00097
35
547-fold shift at 1 lM 30% at 30 lMc
0.0059
0.0046
—
38.9ED50
5.2
500ED50
Dmt-Tic13
a Competition analyses against radiolabeled ligand ([3H]DPDPE for d, [3H]DAMGO for l) were carried out using membrane preparations from
transfected HN9.10 cells that constitutively express the respective receptor types.
b Concentration at 50% inhibition of muscle contraction at electrically stimulated isolated tissues.
c Inhibition % of muscle contraction at electrically stimulated isolated tissues. nc: no competition at 10ꢀ5 M.

2164
Y. S. Lee et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2161–2165
bond of phenethyl group) and h2 = gauche (C–Npiperidine
a
b
c
bond).3 This also suggested that minor differences in
the conformations of substituted fentanyls may have
a significant impact on ligand binding. In comparing
fentanyl with the classic l opiate morphine, certain
similarities are apparent. Both compounds possess a
nitrogen which can be easily protonated and aromatic
groups that are commonly thought to mimic the N-ter-
minal tyrosine moiety of opioid peptides. However,
SAR studies of N-phenolic derivatives of fentanyl have
shown that hydroxyl substitution into the aromatic
ring does not influence the morphinomimetic potency
of these compounds. In our study, it was also shown
that the phenolic hydroxyl group is not a necessary
part for the opioid receptor interactions. Analogues 4
and 5 had no significant differences in their biological
activities and both showed completely superimposed
lowest energy conformations with the two aromatic
rings faced perpendicularly in computer modeling
experiments.
Figure 2. (a) X-ray structure of 3. (b) The lowest energy conformer of
3. (c) Superimposition of 3 (black) and Dmt-Tic (blue). H-bond are
represented by red dashed lines. For simplicity, nonpolar hydrogens
are omitted in the structures.
Ar = Ph], showed greater binding affinities at both d and
l opioid receptors than 7 and 8. It also was observed
that an additional phenyl group at the b-position of 10
increased binding affinity at the l opioid receptor, and
functional activities at the MVD and GPI assays. In
comparison with analogues 11–13 [R1 = H, R2 = NH2],
14 and 15 [R1 = H, R2 = H] showed similar binding
affinities at both opioid receptors. This illustrated that
the position and the presence of amino group are not
critical for binding to these receptors. It also was found
that substitutions at the 4-position on the substituted
phenyl ring did not affect the binding affinities for both
opioid receptors.
In conclusion, we have examined the pharmacological
effects of novel 4-anilidopiperidine analogues on the l
and d opioid receptors. In the structural modifications
described here, all analogues bind more effectively to
the l opioid receptor than to the d opioid receptor.
It is likely that the selective binding to the l opioid
receptor is due to the N-phenyl-N-piperidin-4-yl-propi-
onamide moiety. All analogues showed a wide range of
binding affinities at the d and l opioid receptors
depending on their respective structures. From the
SAR studies of these analogues, it is proposed that
the constraints caused by the methyl groups on the
substituted aromatic ring or the b-carbon serve as the
most critical factor in determining molecular confor-
mation along with the length between the piperazine
and the aromatic rings. Overall, neither of the amino
groups led to an increase in binding. The substituents
at the 3-, or 4-positions also did not play a role in
the binding affinities.
In addition to the bioassays, molecular modeling exper-
iments14 using MacroModel 8.1 were carried and yielded
some insights regarding the determinants of opioid
activity and selectivity. We utilized molecular modeling
to probe the topographical similarities of compounds
2–5 to fentanyl and Dmt-Tic structures, and speculated
that these compounds might provide an overall shape
similar to the bioactive conformation of fentanyl or
Dmt-Tic by controlling the orientation of the anilido
moiety. The studies on 2 gave a topographically identi-
cal structure to the same part of fentanyl, which is con-
sidered as an address region, and gave a longer length
between the two ends of the molecule than the others.
A superimposition of the lowest energy conformers of
2 and fentanyl showed that these compounds have
superimposed anilidopiperidine ring conformations, yet
maintain different aromatic ring conformations. On
the contrary, the lowest energy conformer of 3 looked
more similar to the bioactive conformation of Dmt-Tic
in the shape of the turn, even though its two aromatic
rings were not oriented parallel. The lowest energy con-
former fully matched the structure which was obtained
from X-ray crystallography (Fig. 2).15 Superimposition
of 3 and Dmt-Tic indicates that the binding pocket for
the opioid receptors, preferably for the d, and the possi-
ble p–p interactions between two aromatic rings, may be
blocked by the perpendicularly oriented phenyl ring of
compound 3 to result in the loss of activity.
Acknowledgments
The work was supported by grants from the USDHS,
National Institute on Drug Abuse (DA-12394 and
DA-06284). We thank Margie Colie for assistance with
the manuscript.
References and notes
1. Janssen, P. A. J. Br. J. Anaesth. 1962, 34, 260.
2. Casy, A. F.; Huckstep, M. R. J. Pharm. Pharm. 1988, 40,
605.
3. Subramanian, G.; Paterlini, M. G.; Portoghese, P. S.;
Ferguson, D. M. J. Med. Chem. 2000, 43, 381.
4. Zaveri, N.; Polgar, W. E.; Olsen, C. M.; Kelson, A. B.;
Grundt, P.; Lewis, J. W.; Toll, L. Eur. J. Pharm. 2001, 428,
29.
An analysis of the binding modes indicated the most
potent fentanyl derivatives possess an extended confor-
mation for the phenethyl group with h1 = trans (C–C
5. Hruby, V. J.; Gehrigh, C. A. Med. Res. Rev. 1989, 9, 343.
6. Portoghese, P. S.; Sultana, M.; Takemori, A. E. J. Med.
Chem. 1990, 33, 1714.

Y. S. Lee et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2161–2165
2165
7. Petrov, R. R.; Vardanyan, R. S.; Lee, Y. S.; Ma, S.-W.;
Davis, P.; Begay, L. J.; Lai, J.; Porreca, F.; Hruby, V. J.
Bioorg. Med. Chem. Lett. 2006, 4946.
8. Bryant, S. D.; Jinsmaa, Y.; Salvadori, S.; Okada, Y.;
Lazarus, L. H. Biopolymers 2003, 71, 86.
9.
TEA/benzene, 1 day, 54%; (h) Pd/C, EtOH, rt, 2.5 days,
61%.
10. Lee, Y. S.; Agnes, R. S.; Davis, P.; Ma, S.; Badghisi, H.;
Lai, J.; Porreca, F.; Hruby, V. J. J. Med. Chem. 2006, 49,
1773.
11. Liao, S.; Lin, J.; Shenderovich, M. D.; Han, Y.; Hasohata,
K.; Davis, P.; Qui, W.; Porreca, F.; Yamamura, H. I.;
Hruby, V. J. Bioorg. Med. Chem. Lett. 1997, 23, 3049.
12. Jagerovic, N.; Cano, C.; Elguero, J.; Goya, P.; Callado, L.
F.; Meana, J. J.; Giron, R.; Abalo, R.; Ruiz, D.; Goicoe-
chea, C.; Martin, I. M. Bioorg. Med. Chem. 2002, 10, 817.
13. Salvadori, S.; Attila, M.; Balboni, G.; Bianchi, C.; Bryant,
S. D.; Crescenzi, O.; Guerrini, R.; Picone, D.; Tancredi,
T.; Temussi, P. A.; Lazarus, L. H. Mol. Med. 1995, 1, 678.
14. Molecular modeling experiments were performed on Mac-
roModel 8.1 equipped with Maestro 5.0 graphical interface
installed on a Linux RedHat 8.0. Molecular structures were
minimized using MM2 force field and Polak–Ribier conju-
gate gradient (PRCG). Optimizations were converged to a
gradient RMSD less than 0.05 kJ/ mol or continued until a
limit of 10,000 iterations was reached.
O
O
N
NH2
N
O
b, c
d
a
N
N
O
O
O
O
N
N
NH
N
N
g
h
e, f
N
H
19
15. Compound 3 was crystallized from MeOH, ether, and
hexane as colorless rods (0.13 mm · 0.15 mm · 0.78 mm).
X-ray crystallography on a Bruker SMART 1000 CCD
detector X-ray diffractometer at 170(2) K and a power
setting of 50 kV, 40 mA showed measurable diffraction to
at least theta = 18.23 deg. Data were collected on the
SMART1000 system using graphite monochromated
.
Synthesis of N-phenyl-N-piperidin-4-yl-propionamide
(19). Reagents and conditions: (a) acrylic acid methyl
ester, MeOH, rt, 2 h, reflux, 4 h, 95%; (b) Na/toluene,
reflux, 2 h; (c) HCl at 0 ꢁC, heat, 3-4 h, 41%; (d) aniline,
acetic acid, toluene, Dean–Stark, 2 h, 73%; (e) NaBH4,
MeOH, reflux, 1 h; (f) NaOH, 81%; (g) propionyl chloride,
˚
MoKa radiation (k = 0.71073 A).