Synthesis and SAR of 1,3-disubstituted cyclohexylmethyl urea and amide derivatives as non-peptidic motilin receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2006.04.017

A series of 1,3-disubstituted cyclohexylmethyl urea and amide derivatives were synthesized as motilin receptor antagonists. Starting from known motilin antagonists, 1a and 1b, the cyclopentene scaffold was replaced and the four recognition elements optimi

Full text of DOI:10.1016/j.bmcl.2006.04.017

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3362–3366
Synthesis and SAR of 1,3-disubstituted cyclohexylmethyl urea
and amide derivatives as non-peptidic motilin receptor antagonists
Sigmond G. Johnson,* Joseph W. Gunnet, John B. Moore, William Miller, Pam Wines,
Ralph A. Rivero, Don Combs and Keith T. Demarest
Johnson & Johnson Pharmaceutical Research & Development, L.L.C., 1000 Rt. 202, PO Box 300, Raritan, NJ 08869, USA
Received 16 March 2006; revised 6 April 2006; accepted 6 April 2006
Available online 2 May 2006
Abstract—A series of 1,3-disubstituted cyclohexylmethyl urea and amide derivatives were synthesized as motilin receptor antago-
nists. Starting from known motilin antagonists, 1a and 1b, the cyclopentene scaffold was replaced and the four recognition elements
optimized to arrive at a potent novel series.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Motilin belongs to the ghrelin-motilin G-protein
coupled receptor family of ligands that as a whole regu-
late appetite and gastrointestinal motility.1,2 Motilin, as
a gastrointestinal paracrine hormone, stimulates smooth
muscle contractions affecting the motility and emptying
of both the antral stomach3 and the antroduodenal
region of the small intestine.3–5 Disruption of this coor-
dinated contractile activity may lead to a variety of gas-
tric motility disorders of GI tract origin. As a significant
subset among these, functional disorders, such as irrita-
ble bowel syndrome (IBS) and dyspepsia, are ill-charac-
terized multifactorial diseases with the hallmark of
abnormal gastric function.6 Discordant motilin levels
may play a central role in some patients presenting with
these illnesses.7 Unfortunately, clinical studies with
motilin agonists have thus far not demonstrably im-
proved symptoms in patients.8,9 It has recently been
shown that motilin levels are elevated in IBS patients
relative to healthy volunteers during all phases of motile
activity.10 This evidence would suggest that a motilin
antagonist might be an appropriate course of treatment
for IBS.
series originated from a pharmacophore-model-directed
database search that ultimately focused on mimicking
four key binding amino acids (F1, I4, F5, Y7) of the
motilin peptide. Not surprisingly, when overlaid onto
this model, the peripheral functionality on compounds
1a and 1b overlap with these same four putative binding
amino acid side-chains of motilin. In the further explo-
ration of this series, we focused on changes to the series
that conceptually would maintain the spatial arrange-
ment of the peripheral functionality.
Cl
Cl
Cl
Cl
Cl
O
X
HN
O
HN
Cl
N
HN
N
O
O
N
O
O
X
N
O
N
O
1a X = H
1b X= Cl
7
On the basis of the model overlay and the structure–
activity relationship (SAR) of the lead structure we con-
cluded that the core cyclopentene ring serves largely as a
scaffold for spatially displaying recognition elements.12
Thus, it seemed plausible that the cyclopentene ring
could be replaced without compromising activity. In
considering possible changes to the scaffold we also
wanted to address the constraints imposed by the exist-
ing synthesis route.13 The existing approach was best
suited for SAR development of the urea portion of the
molecule. Benzylic groups were introduced early in the
The discovery of compounds 1a and 1b represented a
fundamental shift away from peptidic and macrocy-
clic-based motilin antagonists.11 This small molecule
Keywords: Motilin; Antagonist.
*
Corresponding author. Tel.: +1 908 704 4347; fax: +1 908 203
8109; e-mail: sgjcej@juno.com
Present address: High Throughput Chemistry, GlaxoSmithKline,
Collegeville, PA, USA.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.04.017

S. G. Johnson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3362–3366
3363
Table 2. Motilin inhibitory activity of 1,3-cyclohexylmethyl ureasa
synthesis through Grignard reagent addition to 1-eth-
oxycyclohexen-3-one. The stereogenic center was subse-
quently set by an aza-Claisen rearrangement. This
rearrangement step was particularly limiting in scope
as it necessitated the inclusion of the stereogenic center
in all analogs. Consequently, the significance of the qua-
ternary center was unknown, though the clear binding
preference for one enantiomer over the other was infor-
mative. In addition to conforming to the spatial require-
ment then, a new scaffold would ideally lend itself to a
facile synthetic route for substitution at all four recogni-
tion elements sites and provide some insight as to the
role of the stereogenic center. We reasoned all of these
objectives might be achieved by moving the benzylic
group in a peptoid-like fashion from the quaternary ste-
reogenic center in compounds 1a and 1b to the nitrogen
of the trichloroacetamide, initially giving derivatives
such as 7.
3
R
O
X
O
N
N
N
H
1
R
O
Compound18 R1
(racemic)
R3
X
Binding assay19
IC50 (lM)
7b
7c
7d
7e
7f
Morpholino
Morpholino
Morpholino
Morpholino
Morpholino
Morpholino
Morpholino
Morpholino
Morpholino
CCl3
CCl3
CCl3
CCl3
H
Cl
0.957
23%
CF3 58%
NO2 0.035
3-NO2Ph
2-Furyl
H
H
H
H
H
H
1.16
7g
7h
7i
0.656
0.730
3%
2-Ph(ethyl)
4-Biphenyl
tert-Butyl
7j
50%
7k
1a
1b
Pyrrolidin-1-yl tert-Butyl
0.021
0.020
0.010
The compounds found in Tables 1–3 are representative
of the many analogs synthesized and were primarily
chosen to illustrate the major SAR advancements. All
of the compounds, with the exception of 13, were
synthesized according to Scheme 1.14 Starting from
cis-3-aminocyclohexanecarboxylic acid, 2, the amine
was protected with trityl chloride to give 3.15 The
protected amino acid was subsequently coupled with a
variety of substituted anilines, the first recognition
element, to arrive at amide intermediates 4 in good yield.
LAH reduction of the amide group then provided a
secondary amine that was capped with phenyl isocya-
nate or 4-fluorobenzoyl chloride to give compounds 5,
incorporating the second recognition element. After
trityl deprotection, Lewis acid-mediated reductive
amination proved to be a general method for introduc-
ing benzaldehydes, the third recognition element.16
The resulting secondary amines, 6, were then capped
with an acid chloride, the fourth recognition element,
to provide compounds 7a–s. The enantiomers 7n and
a IC50 values are means of at least two determinations; CV was 24%.
% inhibition of 125I-motilin binding measured at 1 lM.
7o were separated from a racemic mixture by chiral
chromatography. Compound 8 was prepared by an
analogous route starting from cis 4-aminocyclohexyl-
carboxylic acid. The des-benzylic analog 7t was synthe-
sized directly from intermediate 5 by deprotection of the
amine and capping with 3-methoxybenzoyl chloride.
The methylene spacer and the carbocyclic ring in the
scaffold were transposed according to Scheme 2. Reduc-
tive amination of 3-cyanopentanone with the substituted
aniline 9 gave the 1,3-substituted cyclopentane ring, 10.
This aniline was then reacted with phenyl isocyanate to
produce the highly congested trisubstituted urea 11 in
low yield. Reduction of the nitrile provided the primary
amine 12 for further substitution. A second reductive
amination and acylation gave the desired compound
13 as a 1:1 mixture of diastereomers.
We synthesized a variety of carbocyclic ring structures,
A, as cyclopentene ring replacements (Table 1). Among
these analogs only the 1,3-disubstituted carbocyclic
ring derivatives had appreciable activity. As an exam-
ple, the 1,3-disubstituted cyclohexyl ring systems were
uniformly more potent than their 1,4-disubstituted
counterparts (7a vs 8).17 It was also found that the
location of the methylene spacer was critical as trans-
posing it with the carbocyclic ring, as in 13, led to a
series devoid of activity. Each of the four recognition
elements on the carbocyclic core contributed to optimal
potency. For instance, removal of the benzyl group
from the amide diminished the little activity we had
seen, 7t versus 7a. Conversely, removal of the acyl
group had a similar effect, 6a versus 7a.
Table 1. Motilin inhibitory activity of alternative scaffoldsa
R'
O
N
R4
A
N
N
H
O
N
O
Compound18
(stereo)
A
R0
H
R4
Binding
assay19 IC50
(lM)
6a (rac)
7a (rac)
7t (rac)
8 (rac)
3-ClBn 28%
0.650
26%
3-MeOPhCO 3-ClBn
3-MeOPhCO
3-MeOPhCO 3-ClBn
Cl3CCO 3-ClBn
H
After exploring more than a dozen alternative scaffolds,
we focused our SAR development on 1,3-disubstituted
cyclohexyl ring systems (Table 2). When directly com-
pared with the original series, these derivatives showed
relatively poor inhibition, 7b–c versus 1a and 1b. Never-
theless, we did learn some important SAR from these
6%
13 (dia)
4%
a IC50 values are means of at least two determinations; CV was 24%.
% inhibition of 125I-motilin binding measured at 1 lM.

3364
S. G. Johnson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3362–3366
Table 3. Inhibitory and functional activity of select motilin antagonists
F
R3
O
3
1
N
N
O
O
R1
X
Compound18
(racemic)
R1
R3
X
Binding assay19
Tissue assay20
Human21 recombinant cellular
a
b
c
IC50 (lM)
IC50 (lM)
assay IC50 (lM)
7l
7m
Pyrrolidin-1-yl
Morpholino
tert-Butyl
tert-Butyl
tert-Butyl
tert-Butyl
CCl3
H
H
H
H
H
0.029
0.772
0.010
0.072
0.058
0.058
0.063
0.014
0.024
—
0.990 0.264
—
7n (1S,3R)
7o (1R,3S)
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Piperidin-1-yl
Pyrrolidin-1-yl
0.008
0.638
0.027
0.036
0.021
0.072
0.338 0.059
2.48 0.737
—
7p
7q
7r
7s
CCl3
CCl3
3-Cl
3-Cl
0.267 0.170
0.228 0.043
0.052
CCl3
3-NO2
a IC50 values are means of at least two separate determinations; CV was 24%.
b IC50 values are means of at least two separate determinations; CV was 30%.
c IC50 values are means SE of 2–6 determinations.
O
O
O
TrtHN
TrtHN
R
c
R2
b
OH
NH
N
77-86%
63-79%
R1
R1
O
O
2 R=NH2
4
5
a
85%
3 R=NHTrt
f
62%
R3
O
O
O
H
N
N
R4
N
R2
O
d
N
R2
e
R1
R1
X
O
6
7a-t
Scheme 1. Reagents and conditions: (a) (1) TMSCl, DCM–CH3CN; (2) TrtCl, TEA; (b) substituted anilines, PyBopTM, DIEA, DCM; (c) (1) LAH,
THF; (2) PhNCO, DCM or 4-FPhCOCl, DIEA, DCM; (d) (1) 10% TFA, DCM; (2) ArCHO, Ti(i-OPr)4, DCM; (3) EtOH, NaBH3CN; (e) R3COCl,
DCM; (f) (1) 10% TFA, DCM; (2) 3-MeOPhCOCl, DCM.
analogs. As already shown by example 7a in Table 1, the
trichloroacetyl group found in 7b–c could be replaced
with a variety of aromatic and heteroaromatic contain-
ing carbonyls, 7f–h, without substantially altering
potency. It was only once multiple aromatic rings, such
as biphenyl 7i, or fused rings (data not shown) were
introduced that there was a marked reduction in inhibi-
tion. On the benzyl portion of the molecule, the addition
of electron-withdrawing groups to the meta-position
was tolerated, 7b–d, but only the introduction of a nitro
group significantly improved potency, 7e. The most dra-
matic improvement in activity resulted from replacing
the morpholino group with other tertiary amines, for
example, compound 7j versus 7k. This was a surprising
finding given that tertiary amines were interchangeable
in the original series.13
amide. The results with this substitution in our series
are shown in Table 3. As a general rule this substitution
produced compounds with similar activity; for instance,
the amide 7l (IC50 = 29 nM) versus the urea 7k
(IC50 = 21 nM, Table 2). The nature of the basic side-
chain plays a critical role in these derivatives as well,
7l versus 7m. Activity was not strictly limited to pyrrol-
idine-containing molecules. The analogous piperidine
derivatives, absent the oxygen heteroatom of the mor-
pholino group, were nearly as active (7q vs 7r). With
the preferred pyrrolidine side-chain in place, the intro-
duction of a nitro group to the benzylic ring now had
a more modest 4-fold improvement in binding inhibi-
tion, 7p versus 7s. Once again, no similar improvement
was seen from electron-withdrawing halogen substitu-
ents (7q vs 7p) in the binding assay. In all cases the chi-
rality of the compounds plays a considerable role in
binding, with the (1S,3R) enantiomers consistently
showing 5- to 10-fold greater potency than their antipo-
Data from the original series suggested that the phenyl-
urea group could be replaced with 4-fluorophenyl

S. G. Johnson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3362–3366
3365
Ohning, G.; Miner, P. B.; Mathias, J. R.; Fumagalli, I.;
Staessen, D.; Mack, R. J. Aliment. Pharmacol. Ther. 2000,
14, 1653; (b) Netzer, P.; Schmitt, B.; Inauen, W. Aliment.
Pharmacol. Ther. 2002, 16, 1481.
NC
R
H2N
O
NH
N
H
N
b
a
O
37%
67%
´
10. Simren, M.; Bjo¨rnsson, E. S.; Abrahamsson, H. Neuro-
gastroenterol. Motil. 2005, 17, 51–57.
O
O
N
11. Beavers, M. P.; Gunnet, J. W.; Hageman, W.; Miller, W.;
Moore, J. B.; Zhou, L.; Chen, R. H.; Xiang, A.; Urbanski,
M.; Combs, D. W.; Mayo, K. H.; Demarest, K. T. Drug
Des. Discov. 2001, 17, 243.
O
N
N
O
O
c
9
10
12. Cyclohexene ring systems have also given promising
activity, see Chen, R. H.; Xiang, M. A. WO 0168621.
13. Chen, R. H.; Xiang, M.; Moore, J. B.; Beavers, M. P. WO
9921846.
11 R=CN
12 R=CH2NH2
78%
14. Complete experimental protocols for Schemes 1 and 2 may
be found in: Johnson, S. G.; Rivero, R. A. WO 01085694.
The yields in Scheme 1 for steps d and e were typically
20–50%.
15. Barlos, K.; Papaioannou, D.; Theodoropoulos, D. J. Org.
Chem. 1982, 47, 1324.
O
NH
O
Cl
d
Cl
N
N
O
27%
Cl
N
16. Mattson, R. J.; Pham, K. M.; Leuck, D. J.; Cowen, K. A.
J. Org. Chem. 1990, 55, 2525.
O
Cl
17. The corresponding 1,3-substituted cyclopentyl derivatives
gave a very similar SAR, see Ref. 14.
13
18. In general, reactions were run in parallel and com-
pounds purified on semi-prep HPLC to >95% purity.
Compounds in Tables 1–3 were characterized by 1H
NMR and LC/MS.
Scheme 2. Reagents and conditions: (a) (1) 3-cyanopentanone, 1%
AcOH, MeOH; (2) NaBH3CN; (b) phenylisocyanate, THF; (c) LAH,
THF, ꢀ78 ꢁC to rt; (d) (1) 3-chlorobenzaldehyde, 1% AcOH, MeOH;
(2) NaBH3CN; (3) Cl3CCOCl, DCM.
19. Motilin receptor preparations were made from the
colons of New Zealand white rabbits (Covance Research
Products, Inc., Denver, PA). The tissues were homog-
enized in Tris–HCl buffer, pH 7.5, with a Polytron
(Brinkmann Instruments, Westbury, NY) and then
centrifuged. The pellet was resuspended and stored at
ꢀ80 ꢁC. The binding assay was performed in HEPES
buffer (pH 7.0) containing 15 mg/mL BSA, protease
inhibitors, 125I radiolabeled porcine motilin (50,000–
70,000 cpm; specific activity 2000 Ci/mmol [Amersham
Pharmacia, England]), test compound, and membrane
preparation. The assay was incubated for 60 min at
30 ꢁC with membrane-bound tracer separated by centri-
fugation. The assay signal-to-noise ratio was ꢁ4:1 with
1 lM unlabeled porcine motilin (Peninsula Laboratories,
Belmont, CA) used to measure nonspecific binding. IC50
values were determined using Kaleidograph software
(Synergy Software, Reading, PA). Unlabelled porcine
motilin displaced tracer from rabbit colon with an IC50
value (mean SE) of 0.9 0.3 nM.
des (7n vs 7o). Analogs within the series were confirmed
as antagonists by agreement between the binding, cellu-
lar, and functional tissue data.
In summary we have demonstrated that the core cyclo-
pentene ring found in compounds 1a and 1b could be
replaced by a 1,3-disubstituted cyclohexyl ring system.
Moving the benzylic group to the amide nitrogen elimi-
nated the need for the quaternary stereogenic center.
A significant improvement in activity resulted from
replacing the morpholine side-chain with other tertiary
amines. These key modifications, allowing for rapid syn-
thesis and SAR exploration, led to novel and potent
motilin antagonists.
References and notes
20. Tissue contractility studies were performed using the
longitudinal muscle layer isolated from the first 8 cm of
the duodenum of fasted New Zealand white rabbits of
both sexes. Strips, 3· 30 mm, were attached to force
displacement transducers (FT03, Grass Instruments,
Quincy, MA) in tissue baths containing Krebs solution
gassed with 95% O2, 5% CO2 at 37 ꢁC. Resting tension was
slowly increased to 1g. After equilibration, compounds
diluted with dimethylsulfoxide were added followed 5–
15 min later by porcine motilin (3 nM). Baseline and
response tension levels were expressed as a percent of the
response produced by 100 lM acetylcholine. Acetylcho-
line and porcine motilin had EC50 values of 4.2 0.5 lM
(n = 3) and 1.9 1.5 nM (n = 3), respectively. No test
compound at 10 lM generated any agonist activity.
21. Recombinant human motilin receptor preparations were
made from transfected HEK 293 cells. Cells were trans-
fected with motilin receptor DNA using DMRIE Cꢂ
reagent (Life Technologies, Inc., Grand Island, NY) and
grown in DMEM/F12 media supplemented with 10% fetal
bovine serum, glutamine (Gibco BRL), and geneticin
1. Nogueiras, R.; Tschop, M. Science 2005, 310, 985, and
references cited herein.
2. Asakawa, A.; Inui, A.; Momose, K.; Ueno, N.; Fujino, M.
A.; Kasuga, M. Peptides 1998, 19, 987.
3. Wilmer, A.; Tack, J.; Coremans, G.; Janssens, J.; Peeters,
T.; VanTrappen, G. Gastroenterology 1993, 105, 773.
4. Review: Itoh, Z. Peptides 1997, 18, 593.
5. (a) You, C. H.; Chey, W. Y.; Lee, K. Y. Gastroenterology
1980, 79, 62; (b) Vantrappen, G.; Janssens, J.; Peeters, T.
L.; Bloom, S. R.; Christofides, N. D.; Hellemans, J. Dig.
Dis. Sci. 1979, 24, 497.
6. Camilleri, M. Pharmacogenomics 2005, 6, 491, and refer-
ences cited herein.
7. Review: Chovet, M. Curr. Opin. Chem. Biol. 2000, 4, 428.
8. Russo, A.; Stevens, J. E.; Giles, N.; Krause, G.; O’Dono-
van, D. G.; Horowitz, M.; Jones, K. L. Aliment. Pharma-
col. Ther. 2004, 20, 333.
9. (a) Talley, N. J.; Verlinden, M.; Snape, W.; Beker, J. A.;
Ducrotte, P.; Dettmer, A.; Brinkhoff, H.; Eaker, E.;

3366
S. G. Johnson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3362–3366
(500 lg/mL, Life Technologies). Motilin-induced changes
Molecular Devices, Inc., Sunnyvale, CA). In each exper-
imental run, compound agonist activity and antagonist
activity (against 2 nM porcine motilin) were determined.
Porcine motilin stimulated calcium mobilization with an
in intracellular calcium mobilization were measured using
calcium-sensitive fluorescent dye as previously described in
Biotechnol. Lett. 2001, 23, 2067. Transfected HEK 293
cells were loaded with fluo 3 AMꢂ (5 mM, Molecular
Probes, Inc., Eugene, OR) for 1 h shielded from light at
room temperature. Intracellular fluorescence was mea-
sured using fluorometric imaging plate reader (FLIPRꢂ;
EC50 of 3.0 0.4 nM (n = 4) producing
a maximal
response ꢁ73-fold greater than baseline. No test com-
pound showed agonist activity at concentrations up to
30 lM.