Identification of potent phenyl imidazoles as opioid receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2006.01.082

Using previously reported opioid receptor (OR) agonist analogs 4a-c as starting points, the structure-activity relationship (SAR) for their related series has been further refined. This SAR study has led to the identification of 2,6-di-Me-Tyr (DMT) analog

Full text of DOI:10.1016/j.bmcl.2006.01.082

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2505–2508
Identification of potent phenyl imidazoles as opioid
receptor agonists
Henry J. Breslin,* Chaozhong Cai, Tamara A. Miskowski, Santosh V. Coutinho,
Sui-Po Zhang, Pamela Hornby and Wei He
Johnson & Johnson Pharmaceutical Research & Development, L.L.C., Welsh and McKean Roads, PO Box 776,
Spring House, PA 19477-0776, USA
Received 23 November 2005; revised 19 January 2006; accepted 20 January 2006
Available online 17 February 2006
Abstract—Using previously reported opioid receptor (OR) agonist analogs 4a–c as starting points, the structure–activity relation-
ship (SAR) for their related series has been further refined. This SAR study has led to the identification of 2,6-di-Me-Tyr (DMT)
analogs 4h and 4j as the most potent OR agonist within the series. In addition, it was discovered that 4-(aminocarbonyl)-2,6-dimeth-
yl-Phe is a reasonable bioisostere surrogate for the DMT moiety, as supported by the OR activities of compounds 4x and 4y.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Compounds that modulate opioid receptors (ORs) are
well recognized as being useful therapeutic agents for
pain management (e.g., morphine, 1, and fentanyl, 2),
and gastrointestinal (GI) motility regulation (e.g., lop-
eramide, 3).1,2
results for compounds 4a–c, we have further profiled
the structure–activity relationship (SAR) for this new
series. As a starting point, we chose to explore the iden-
tified phenyl moiety highlighted in generic structure 4.
The metrics for this follow-up SAR exploration included
evaluating the minimal functional group requirements
of the phenyl moiety, and also appraising whether the
OR activities of our initial leads could be enhanced
through related modifications. Following are described
the associated experimental results for the preparation
and activities for compounds 4.
O
CH3
NCH3
N
H
HO
N
N
CH3
HO
Cl
N
O
O
2
CH3
3
1
OH
Fentanyl
(Duragesic)
Loperamide
(Imodium)
Morphine
A
The OR has been well characterized, and has been divid-
ed into three major subclasses of receptors, d, l, and j.2
Recently, we reported our initial OR-related findings
around a new series of compounds (e.g., 4a, 4b, and
4c) that exhibit strong binding affinities at the d and/or
l ORs, and also demonstrate significant OR functional
activity as determined by GTPcS assays.3 Compound
4a from this series was further profiled for in vivo activ-
ity, and found to possess significant GI motility modu-
lating effects.3 Based on the initial promising biological
N
N
R
N
H
R'
O
NH2
X
R
4
a R, R'=H; A=Ph, X=OH
b R=H; R'=CH3; A=Ph, X=OH
c R=H; R'=CH3; A=---, X=OH
Initial SAR
Explorations
The preparation of the initial set of phenyl modified
products 4 proved rather straightforward, as commer-
cially available synthetic Phe amino acids4 were em-
Keywords: Opioid receptor ligands; Opioid receptor agonist; DMT;
2,6-Dimethyl-Tyr; 4-(Aminocarbonyl)-2,6-dimethyl-Phe; 2,6-Di-Me-4-
carboxamido-Phe; Phenyl imidazole.
ployed as starting materials (SMs) (Table
1
summarizes all products 4 prepared5). As outlined in
Scheme 1, the diverse set of N-protected Phe amino
acids were coupled under standard 1-[3-(dimethylami-
*
Corresponding author. Tel.: +1 215 628 5610; fax: +1 215 628
4985; e-mail: HBreslin@prdus.jnj.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.01.082

2506
H. J. Breslin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2505–2508
Table 1. Final compounds 4 and related d and l OR in vitro binding affinities
Compound
R
R0
X
A
OR binding
Ratio l/d
Ki d (nM)
Ki l (nM)
4a
4b
4c
4d
4e
4f
H
H
–OH
–OH
Fused Ph
Fused Ph
—
0.9
0.3
22
55
21
63
75
H
CH3
CH3
H
H
–OH
OH
1.5
3.8
0.1
0.02
1.7
8
H
—
198
266
236
1130
0.1
0.1
1.9
687
28
H
H
H
Fused Ph
Fused Ph
—
443
1835
18
H
CH3
H
H
4g
4h
4i
H
H
–OH
0.02
3.2
3.2
0.03
19
CH3
CH3
CH3
H
H
Fused Ph
Fused Ph
—
0.3
0.3
CH3
H
–OH
–OH
4j
0.05
12,800
179
857
207
5260
1335
356
912
413
592
5800
23
4k
4l
CH3
CH3
CH3
CH3
H
–F
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
Fused Ph
—
H
–OCH3
–NH2
–NHAc
–Cl
6.4
9
4m
4n
4o
4p
4q
4r
4s
4t
H
93
34
H
6
H
1130
752
342
466
30
4.6
1.8
1
H
H
–CN
H
H
–NHSO2CH3
–CH2OH
–COCH3
–SO2NH2
–CO2H
–CONH2
–CONH2
–CONH2
–CONH2
H
H
2
14
H
H
H
H
174
5200
1.3
65
3.4
1.1
18.4
0.02
24
4u
4v
4w
4x
4y
H
H
H
H
H
H
1.2
1.4
CH3
CH3
H
Fused Ph
—
0.06
14
H
0.13
0.01
Boc-AA-OH
or
Fmoc-AA-OH
A
A
3 steps (for a,c)
2 steps (for b)
O
L
N
N
N
H
Ref 3
EDC, HOBT
HN
W
R'
5
6
a: W=t-Boc; A=Ph; L=OH
b: W=t-Boc; A=Ph; L=H
c: W=Cbz; A= ----; L=OH
a: R'=H; A=Ph
b: R'=CH3; A=Ph
c: R'=H; A= ----
A
N
N
R
TFA (for Boc protected 7)
N
H
R'
4
O
Piperidine (for Fmoc protected 7)
HN
Boc (Fmoc)
X
R
7
Scheme 1.
no)propyl]-3-ethylcarbodiimide/1-hydroxybenzotriazole
(EDC/HOBT) amide forming reaction conditions with
common intermediate amines 6a–c, whose preparations
and chemical characterizations have been previously de-
scribed in detail.3 For the resulting Boc-protected
amides 7, the Boc group was readily removed with triflu-
oroacetic acid (TFA) at 0 ꢁC to yield their respective
products 4. For the Fmoc-protected adducts 7, treat-
ment with 20% piperidine in MeOH at 0 ꢁC cleanly
cleaved the Fmoc group to generate their respective tar-
gets 4.
prepared, follow-up analogs 4x and 4y were targeted
for synthesis. To prepare 4x and 4y, the appropriate
SM amino acid 12 had to be prepared, which was
accomplished via a four-step sequence in good overall
yield (Scheme 2) as we have recently described in detail.6
The Boc-protected amino acid 12 was then sequentially
carried on to respective intermediates 7 and subsequent
final targets 4x and 4y, in a similar manner as outlined in
Scheme 1.
As alluded to in the introduction, the initial follow-up
SAR exploration around preliminary hits 4a–c was to
answer whether a functional group was needed on the
highlighted phenyl moiety of 4 to maintain good OR
As 4-aminocarbonyl-Phe products 4v and 4w proved
biologically promising from this initial set of analogs

H. J. Breslin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2505–2508
2507
CH3
CH3
Pd(OAc)2, CO (g)
DPPF, K2CO3
DMF
(CF3SO2)2NPh,
Et3N, CH2Cl2
CH3
O
CH3
O
O
O
94%
NH-Boc
CH3
NH-Boc
CH3
99%
F3C
HO
O
S
9
8
O
O
PyBop, HOBT
NH4Cl, DIPEA
DMF
CH3
O
CH3
OH
CH3
O
CH3
O
LiOH, H2O
THF
87%
H2N
CH3
O
O
100%
H2N
NH-Boc
CH3
HO
NH-Boc
CH3
NH-Boc
CH3
O
O
12
O
10
11
Scheme 2.
activity, and if so, which functional groups might prove
most promising. The SAR strategy also encompassed
trying to identify compounds with enhanced OR activi-
ties relative to our initial leads 4a–c. The related d and l
OR binding affinities7 used as the initial biological met-
ric for this SAR evaluation are summarized in Table 1.
For the biologically most promising analogs, a second-
ary OR GTPcS functional screen8 was run (Table 2).
ities were significantly enhanced (30- to 180-fold) for the
DMT analogs. The d OR binding affinities were affected
to a lesser extent, with the DMT analogs demonstrating
improved OR binding affinities between 3- and 100-fold
over their sister Tyr analogs [4a (Ki d = 0.9 nM; Ki
l = 55 nM) relative to 4h (Ki d = 0.1 nM; Ki l =
0.3 nM); 4b (Ki d = 0.3 nM; Ki l = 21 nM) relative to
4i (Ki d = 0.1 nM; Ki l = 0.3 nM); 4d (Ki d = 198 nM;
Ki l = 3.8 nM) relative to 4j (Ki d = 1.9 nM; Ki
l = 0.05 nM)]. The associated OR functional GTPcS
activities evaluated were reflective of the binding data,
with the DMT analogs showing enhanced potency rela-
tive to the simple Tyr derivatives. Unfortunately we did
not obtain functional data for analog 4d, so the closest
comparison for 4j was 4c [4a (EC50 d = 19 nM; EC50
l = 2445 nM) relative to 4h (EC50 d = 0.9 nM; EC50
l = 27 nM); 4c (EC50 d = 500 nM; EC50 l = 142 nM)
relative to 4j (EC50 d = 37 nM; EC50 l = 2 nM)].
The SAR starting point was to evaluate the OR binding
affinity contributions of the phenol’s OH moiety for 4a–
c. The Phe analogs 4e, 4f, and 4g were all found to pos-
sess significantly lower relative binding affinities for both
the d (6- to 780-fold) and l (10- to 90-fold) ORs relative
to their sister Tyr analogs 4a, 4b, and 4d, confirming
that the phenolic OH group is a significant contributor
to 4’s OR binding affinities [4a (Ki d = 0.9 nM; Ki l =
55 nM) relative to 4e (Ki d = 266 nM; Ki l = 443 nM);
4b (Ki d = 0.3 nM; Ki l = 21 nM) relative to 4f (Ki
d = 236 nM; Ki l = 1835 nM); 4d (Ki d = 198 nM; Ki
l = 3.8 nM) relative to 4g (Ki d = 1130 nM; Ki l =
18 nM)]. This finding is consistent with the generally
accepted rule for OR ligands related to the endogenous
opioids, that is, that the phenol OH moiety is critical for
good OR binding.2
Related to exploring potential alternate functional
groups in place of the phenol OH moiety, we found that
with one exception, replacing the phenol OH with either
lipophilic or other polar functional groups significantly
diminished both the d and l OR binding affinities. Di-
rect comparisons where we replaced the phenol of 4b
(X = OH; Ki d = 0.3 nM; Ki l = 21 nM) with the other
functionality included analogs 4k (X = F; Ki d =
687 nM; Ki l = 12,800 nM), 4l (X = OMe; Ki d =
28 nM; Ki l = 179 nM), 4m (X = NH2; Ki d = 93 nM;
Ki l = 857 nM), and 4n (NHAc; Ki d = 34 nM; Ki
l = 207 nM). Other functional group analogs examined
that also had disappointing OR binding affinities, when
compared to alternate parent phenol 4a (Ki d = 0.9 nM;
Ki l = 55 nM), were 4o (X = Cl; Ki d = 1130 nM; Ki
l = 5260 nM), 4p (X = CN; Ki d = 752 nM; Ki l =
1335 nM), 4q (X = NHSO2CH3; Ki d = 342 nM; Ki
l = 356 nM), 4r (X = CH2OH; Ki d = 466 nM; Ki l =
912 nM), 4s (X = COCH3; Ki d = 30 nM; Ki l = 413
nM), 4t (X = SO2NH2; Ki d = 174 nM; Ki l = 592 nM)
and 4u (X = CO2H; Ki d = 5200 nM; Ki l = 5800 nM).
The sole promising functional group replacement found
for the phenolic OH moiety from this exercise was the
carboxamide group. The OR binding affinities for benz-
amide analogs 4v (Ki d = 1.3 nM; Ki l = 23 nM) and 4w
(Ki d = 65 nM; Ki l = 1.2 nM) proved comparable to
their respective parent phenol analogs 4a (Ki
d = 0.9 nM; Ki l = 55 nM) and 4d (Ki d = 198 nM; Ki
l = 3.8 nM). The associated OR functional GTPcS
Finding the phenol OH moiety essential for maintaining
significant OR binding, we prepared 2,6-di-Me-phenyl-
substituted Tyr (DMT) derivatives 4h, 4i, and 4j, again
direct analogs of the Tyr derivatives 4a, 4b, and 4d.
We chose to prepare these di-Me-substituted analogs,
as some other N-terminal Tyr-related endogenous opi-
oids have been reported to show improved OR binding
activities by substituting DMT for Tyr.9 In the three di-
rect comparisons for our series, the l OR binding affin-
Table 2. d and l OR GTPcS in vitro functional activities
Compound
EC50 (nM)
d
l
4a
4c
4h
4j
19
2445
142
27
500
0.9
37
2
4v
4w
4x
4y
3
155
100
161
9
730
22
135

2508
H. J. Breslin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2505–2508
activity for benzamides 4v (EC50 d = 3 nM; EC50 l =
155 nM) and 4w (EC50 d = 730 nM; EC50 l = 100 nM)
also maintained the OR agonist efficacy noted for
respective phenols 4a (EC50 d = 19 nM; EC50 l =
2445 nM) and 4c (EC50 d = 500 nM; EC50 l = 142
nM), although slight variations were noted in terms of
relative potencies. These biologically favorable OR re-
sults for our benzamides reflect the benzamide for phe-
comparing OR activities of benzamides 4v and 4w to
phenols 4a and 4d. Recognizing that benzamides 4v
and 4w were as promising biologically as phenols 4a
and 4d, and that DMT phenols 4h and 4j proved
extremely potent as OR ligands, led to the idea of
hybridizing the two phenyl-substituted moieties to gen-
erate the first reported 4-(aminocarbonyl)-2,6-dimeth-
yl-Phe OR ligands, 4x and 4y, specifically. Benzamides
4x and 4y maintained consistent d and l OR binding
affinities as the DMT analogs 4h and 4j, although the
benzamides proved a bit less potent in terms of OR
functional activity. We are continuing to explore addi-
tional DMT-amide analogs, as well as further examining
the remaining SAR for molecules akin to 4. Related
findings will be reported in subsequent publications.
nol bioisostere correlation that Wentland et al.
10
discovered for opiate-related OR ligands.
Our find-
ings also further validate the more recently reported
extension of the benzamide–phenol bioisostere equiva-
lency for Tyr-related OR ligands,11 which also comple-
mented Wentland’s initial opiate work.
Having found the carboxamide a good bioisostere
replacement for the phenol OH for compounds within
our series, we questioned whether the OR activities of
the 4-aminocarbonyl-Phe derivatives 4v and 4w could
be enhanced by preparing 4-(aminocarbonyl)-2,6-di-
methyl-Phe derivatives 4x and 4y. This hypothesis was
proposed with the hope of mirroring our earlier obser-
vation of improved OR activities for DMT derivatives
relative to simple Tyr derivatives. The outcome of this
exercise proved positive, with the 4-(aminocarbonyl)-
2,6-dimethyl-Phe derivatives 4x (Ki d = 0.06 nM; Ki
l = 1.4 nM) and 4y (Ki d = 14 nM; Ki l = 0.13 nM)
showing improved OR binding affinities relative to the
des-methyl Phe carboxamides 4v (Ki d = 1.3 nM; Ki
l = 23 nM) and 4w (Ki d = 65 nM; Ki l = 1.2 nM). Con-
sistent with our hypothesis, the OR binding affinities for
4x and 4y were more closely aligned with the OR bind-
ing affinities of their respective DMT analogs 4h (Ki
d = 0.1 nM; Ki l = 0.3 nM) and 4j (Ki d = 1.9 nM; Ki
l = 0.05 nM). Although benzamides 4x (EC50
d = 22 nM; EC50 l = 161 nM) and 4y (EC50
d = 135 nM; EC50 l = 9 nM) also both demonstrated
significant d and l GTPcS OR agonist functional activ-
ities, their functional potencies did not prove quite as ro-
bust as for their sister DMT analogs 4h (EC50
d = 0.9 nM; EC50 l = 27 nM) and 4j (EC50 d = 37 nM;
EC50 l = 2 nM).
References and notes
1. Aldrich, J. V. In Burger’s Medicinal Chemistry and Drug
Discovery, 5th ed; Volume 3: Therapeutic Agents, John
Wiley & Sons, 1996; pp 321–441.
2. Fries, D. S. In Principles of Medicinal Chemistry; Foye, W.
O., Lemke, T. L., Williams, D. A., Eds., 4th ed.; Willimams
and Wilkins: Baltimore, MD, 1995; pp 247–269.
3. Breslin, H. J.; Miskowski, T. A.; Rafferty, B. M.;
Coutinho, S. V.; Palmer, J. M.; Wallace, N. H.; Schneider,
C. R.; Kimball, E. S.; Zhang, S.-P.; Li, J.; Colburn, R. W.;
Stone, D. J.; Martinez, R. P.; He, W. J. Med. Chem. 2004,
47, 5009.
4. Starting materials containing the various functional
groups X were commercially available as Boc-AA-OH to
prepare final products 4e–4o, 4q and 4s–4t, and commer-
cially available as Fmoc-AA-OH to prepare final products
4p, 4r, 4v and 4w. The aniline intermediate 7 for
compound 4m was prepared by subjecting its respective
Boc-protected 4-nitro-Phe coupled adduct 7 to a hydra-
zine/Raney Ni hydrogenation. The starting material for
final product 4u was a di-protected AA, that is, N4-
bis[(1,1-dimethylethoxy)carbonyl]-L-Phe, with both pro-
tecting groups of its respective intermediate 7 being
cleaved in the final TFA deprotection step. RSP Amino
Acid Products supplied starting material 8.
5. Final products 4 were analyzed for purity by HPLC
(>98% purity at 214 and 254 nm; Hewlett-Packard Series
1050 HPLC with a 3 lm, 3.3 mm · 50 mm Supelco AZB+
C18 column using an acetonitrile/water/TFA gradient
eluent system), and chemically characterized by MS
(Finnigan 3300) and 300-MHz 1H NMR (Bruker-Biospin,
Inc. DPX-300).
6. Cai, C.; Breslin, H. J.; He, W. Tetrahedron 2005, 61, 6836.
7. Rat brain homogenate was used for both the d and l OR
binding assays. Complete experimental details for these
binding assays are described in Ref. 3.
8. Purchased CHO-hg cell membrane was used for the
GTPcS OR functional assays. Complete experimental
details for these functional assays are described in Ref. 3.
9. Bryant, S. D.; Jinsmaa, Y.; Salvadori, S.; Okada, Y.;
Lazarus, L. H. Biopolymers 2003, 71, 86.
In conclusion, we systematically evaluated the SAR for
substitutions on the highlighted phenyl moiety of gener-
ic structure 4 relative to d and l OR activities. This
exploration verified that for initial leads 4a–c the pheno-
lic OH moiety was a key functional group for both d and
l OR binding and functional activities, as the related
Phe analogs 4e–g proved to be significantly less potent
than 4a, 4b, and 4d. In contrast, the OR-related activi-
ties for compounds 4a, 4b, and 4d could be significantly
enhanced by substituting 2,6-di-methyls on their pheno-
lic moieties to generate DMT compounds 4h–j, the most
potent d and l OR analogs identified within this series.
Relative to exploring various alternative aromatic func-
tional groups to replace the phenolic OH moiety of Tyr
derivatives 4a–c, we found after a limited diversity
search that the carboxamide was the sole functional
group that could replace the phenolic OH group
and maintain similar OR activities, as confirmed by
10. Wentland, M. P.; Lou, R.; Ye, Y.; Cohen, D. J.;
Richardson, G. P.; Bidlack, J. M. Bioorg. Med. Chem.
Lett. 2001, 11, 623.
11. Dolle, R. E.; Machaut, M.; Martinez-Teipel, B.; Belanger,
S.; Cassel, J. A.; Stabley, G. J.; Graczyk, T. M.; DeHaven,
R. N. Bioorg. Med. Chem. Lett. 2004, 14, 3545.