The discovery of a series of N-substituted 3-(4-piperidinyl)-1,3- benzoxazolinones and oxindoles as highly brain penetrant, selective muscarinic M1 agonists

Article abstract of DOI:10.1016/j.bmcl.2010.07.097

A series of N-substituted 3-(4-piperidinyl)-1,3-benzoxazolinones and oxindoles are reported which were found to be potent and selective muscarinic M₁ agonists. By control of the physicochemical characteristics of the series, particularly the li

Full text of DOI:10.1016/j.bmcl.2010.07.097

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5434–5438
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of a series of N-substituted 3-(4-piperidinyl)-1,3-benzoxazoli-
nones and oxindoles as highly brain penetrant, selective muscarinic M1 agonists
*
Dale J. Johnson , Ian T. Forbes, Steve P. Watson, Vincenzo Garzya, Graeme I. Stevenson, Graham R. Walker,
Harminder S. Mudhar, Sean T. Flynn, Paul A. Wyman, Paul W. Smith, Graham S. Murkitt, Adam J. Lucas,
Claudette R. Mookherjee, Jeannette M. Watson, Jane E. Gartlon, Andrea M. Bradford, Fiona Brown
GlaxoSmithKline, New Frontiers Science Park, 3rd Avenue, Harlow, Essex CM19 5AW, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of N-substituted 3-(4-piperidinyl)-1,3-benzoxazolinones and oxindoles are reported which were
found to be potent and selective muscarinic M1 agonists. By control of the physicochemical characteris-
tics of the series, particularly the lipophilicity, compounds with good metabolic stability and excellent
brain penetration were identified. An exemplar of the series was shown to be pro-cognitive in the novel
object recognition rat model of temporal induced memory deficit.
Received 22 June 2010
Revised 22 July 2010
Accepted 23 July 2010
Available online 30 July 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Muscarinic M1 agonist
Cognitive enhancer
Allosteric
Central nervous system penetration
Benzoxazolinone
Oxidole
The acetylcholine (ACh) activated muscarinic family of G-pro-
tein coupled receptors (GPCRs) are widely distributed throughout
allosteric site.8 The evolutionary pressure on the receptor away
from the binding site of the endogenous ligand is much reduced,
leading to a much greater heterogeneity across M1–5. This offers
the opportunity to achieve selectivity over the M2–5 sub-types.
Compound 3 was soon followed by further ligands such as TBPB 4.9
We have previously reported the identification of a series of 20-
biaryl amides, exemplified by compound 5, which were identified
from a high throughput screen (HTS) of the GlaxoSmithKline cor-
porate compound collection.10,11 A further hit identification initia-
tive utilised virtual screening of the compound collection against a
pharmacophore of AC-42 3. This led to the discovery of a series of
benzimidazolinones which, although closely related in structure to
TBPB 4, showed diverse structure activity relationships (SAR),
especially for substitution of the aromatic ring. This programme
of work led to the identification of compound 6, which demon-
strated a pro-cognitive effect in the rat passive avoidance model
and other models of cognition.12 The benzimidazolinones offered
excellent pharmacological and pharmacokinetic profiles, but the
modest brain penetration for 6 in rat suggested the potential
involvement of active efflux out of the brain (e.g., via P-glycopro-
tein (P-gp) transporter). Herein we report the identification of a
series of oxindoles 7 and benzoxazolinones 8, describing strategies
to prepare compounds, which are selective M1 agonists, with
excellent brain penetration and without P-gp liability.
both the central and peripheral nervous systems. There are five
1
known receptor sub-types, M1–5
.
The M1 receptor is highly ex-
pressed in the hippocampus and cerebral cortex, areas of the brain
linked with memory and learning function.2 This makes it an
attractive target for pharmaceutical intervention in diseases asso-
ciated cognitive deficits, such as schizophrenia and Alzheimer’s
disease.3
The first generation of muscarinic agonists mimicked ACh and
utilised the ACh binding site. Due to the high sequence homology
of this orthosteric binding site, these ligands were non-selective
for the different muscarinic sub-types.4 Molecules such as xanom-
eline 15 and sabcomeline 26 (Fig. 1) did progress into the clinic;
however, studies were discontinued, despite some promising signs
of pharmacological action, due to intolerable dose limiting side ef-
fects.7 These side effects are thought to be mediated primarily by
the M2 and M3 receptor sub-types, highlighting the need for a
selective ligand.
A second generation of selective M1 ligands was heralded by the
disclosure of AC-42 3 by Acadia, which was shown to bind at an
Abbreviations: Ach, acetylcholine; GPCRs, G-protein coupled receptors; HTS,
high throughput screen; SAR, structure activity relationship; P-gp, P-glycoprotein.
* Corresponding author. Tel.: +44 1992 411300.
The common right hand intermediates 9a–d were synthesised
starting from either the protected cyclohexanol 10 or cyclohexa-
E-mail address: dalejames.johnson@gmail.com (D.J. Johnson).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.097

D. J. Johnson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5434–5438
5435
Figure 1. Muscarinic agonists.
none 11 (Scheme 1). Reductive etherification, catalysed by either
bismuth tribromide or iron trichloride, installed the desired cap-
ping ether, with saponification using aqueous sodium hydroxide
giving the carboxylic acid. For intermediates starting from cyclo-
hexanol 10 the quaternary methyl group was installed using lith-
ium diisopropylamide (LDA) and methyl iodide. Separation of the
cis and trans isomers was achieved using thionyl chloride.13 The
isomerically pure acid 12 was converted to amine 13 by a Curtius
rearrangement using diphenyl phosphoryl azide (DPPA), followed
by hydrolysis of the isocyanate using aqueous hydrochloric acid.
The final piperidinone ring was formed by reaction of amine 13
with piperidinone salt 14 to give the right hand ketone 9.
The oxindoles 7a–c were prepared using the method described
by Forbes, (Scheme 2).14 SNAr displacement using di-t-butyl mal-
onate gave the nitro aromatic compound 15. Reduction to the ani-
line 16 was followed by reductive amination with the right hand
component 9 to give the substituted aniline 17. Treatment with
para-toluenesulfonic acid (p-tsa), in toluene at reflux, produced
the final cyclisation and decarboxylation to give the oxindole 7.15
The benzoxazolinones 8a–k were prepared by a two step proce-
dure, starting from the substituted aminophenol 18 (Scheme 3).
Reductive amination with the ketone 9, using polymer supported
cyanoborohydride, was followed by cyclisation with either
triphosgene or carbodiimidazole (CDI) to give the target benzoxaz-
olinone 8.
We started from a series of benzimidazolinones related to 6,
with replacement of the tetrahydropyranyl (THP) capping group
by alkoxy substituted cyclohexyl groups giving an improved M1
potency (cf. 19 and 6, Table 1).16,17 The strategy focused on replac-
ing the benzimidazolinone, as the presence of the hydrogen bond
donating N-H group has been implicated in P-gp interaction.18
Oxindoles 7a–c were prepared and 7a and 7b maintained an excel-
lent potency for the M1 receptor, with a pEC50s of 9.3 and 9.4 (Table
1). However, the use of the methoxy group (7c) led to a reduction
in potency of just under a log unit. Pharmacokinetic profiling of 7a
in rat revealed excellent brain penetration (Br:Bl ratio of 5.7). Fur-
thermore, the free concentrations in the brain and blood (2.5 nM
and 2.6 nM, respectively) suggested passive diffusion across the
brain blood barrier with no transporter efflux (Table 2).19 However
7a had a high estimated blood clearance in rat of 85 mL/min/kg. It
was presumed this was due to metabolic instability driven by the
increased lipophilicity of 7a compared to 6.
Benzoxazolinone analogues 8 were investigated as an alterna-
tive to the oxindole 7. With a simpler synthetic scheme and only
a moderate reduction in potency (cf. 8a with 7c), the benzoxazoli-
none core was selected to probe the effects of aromatic substitu-
tion on potency and physicochemical properties. No substitution
led to a marked reduction in M1 potency (8b), while small substit-
uents were tolerated at R4 and R5 (see 8c and 8d). The electronic
nature of R6 had little impact on muscarinic potency, with the

5436
D. J. Johnson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5434–5438
Scheme 1. Synthesis of right hand capping group 9. Reagents and conditions: (a) Y-CHO, BiBr3 or FeCl3, Et3SiH, MeCN, 87%; (b) concd NaOH, MeOH/THF, 96%; (c) Y-OTBS,
BiBr3 or FeCl3, Et3SiH, MeCN, 83%; (d) (i) 2 equiv LDA; (ii) MeI, 87%; (e) SOCl2, PhMe, 85 °C, 34–40%; (f) (i) DPPA, Et3 N, PhMe 90 °C, 94–100%; (ii) 5 M HCl, THF, 53–100%; (g)
K2CO3, H2O, EtOH, 90 °C, 45–63%.
Scheme 2. Synthesis of oxindole analogues. Reagents and conditions: (a) di-t-butyl malonate, NaH, DMF, 29–56%; (b) H2, 10% Pd/C, EtOH, 96–100%; (c) PS-B(CN)H3, AcOH,
DCM, 100 °C microwave; (d) p-tsa, PhMe, reflux, 62–89% over two steps.
electron donating methoxy analogue 8e and electron withdrawing
nitrile analogue 8f having very similar M1 potencies and selectivi-
ties. Introducing the highly electron withdrawing methyl sulfone
substituent (8g) did lead to a reduction in M1 potency, and the

D. J. Johnson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5434–5438
5437
Scheme 3. Reagents and conditions: (a) PS-B(CN)H3, AcOH, DCM, 100 °C microwave, 29–92%; (b) triphosgene or CDI, DIPEA, DCM, 0 °C, 6–80%.
Table 1
Muscarinic agonist data
R6
R5
R4
c log P
pEC50
a
Compd
X
Y
b
b
b
b
b
M1
M2
M3
M4
M5
6
NH
NH
CH2
CH2
CH2
O
O
O
O
O
O
O
O
O
O
O
N/Ac
n-Pr
n-Pr
Et
Me
Me
Et
Et
Et
Et
Et
Et
Me
Me
Me
Me
Me
Me
H
Me
Me
MeO
CN
MeSO2
EtSO2
MeSO2
Me
F
F
H
H
H
H
H
H
H
H
F
H
H
H
H
H
H
H
1.90
4.18
3.48
2.95
2.56
2.73
2.62
3.26
3.26
2.54
2.05
0.98
1.51
1.51
2.59
1.53
8.3
9.4
9.3
9.4
8.5
8.1
7.6
9.0
8.9
8.6
8.5
7.9
6.7
8.6
8.6
8.0
6.1
7.7
7.2
7.0
6.2
5.9
<4.8
6.3
6.1
6.4
5.6
5.6
<4.8
6.1
6.2
5.6
5.6
7.1
7.2
6.9
6.5
5.7
<4.8
6.1
5.7
5.8
5.6
5.9
6.4
5.5
6.2
5.4
6.6
7.9
7.4
7.5
6.8
6.3
6.9
7.0
6.7
6.8
6.1
5.8
5.8
6.3
6.7
6.1
6.1
7.9
7.7
7.7
6.9
6.4
6.0
7.3
6.8
6.8
6.4
5.9
5.3
6.5
6.8
6.0
19
7a
7b
7c
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
H
H
H
H
H
F
H
H
H
H
H
H
H
H
Et
n-Pr
MeOCH2CH2
MeOCH2CH2
CN
a
b
c
Mean value of at least five test occasions.
Ref. 16.
THP cap.
gave compound 8i with a c log P of 1.51 and an increased M1 po-
tency of 8.6. Pharmacokinetic analysis of 8i gave a low estimated
rat blood clearance of 11 mL/min/kg, demonstrating that good
metabolic stability could be achieved in the series (Table 2). This
encouraging result was tempered by a 2.25-fold difference be-
tween free brain concentration of 8i compared to the free blood
concentration, suggesting 8i was a substrate for P-gp.19
This result led us to target further analogues with a c log P of
ꢀ1.5, replacing the sulfone with alternative substituents at R6
and varying alkoxide Y. It was anticipated that the reduced hydro-
gen bond acceptor character of the molecule would prevent P-gp
recognition. The combination of a nitrile substituent with the
insertion of a second oxygen atom in the capping alkoxy group
Table 2
Pharmacokinetic data
a
a
a
Compd
Est. Clb
Br:Bl
BTBb
(%)
WBBb
(%)
Cfree,
(nM)
Cfree,
(nM)
brain
blood
(mL/min/kg)
7a
8i
85
11
5.7c
0.8a
94
74
80
60
2.5
168
2.6
378
a
b
c
Estimated from 3 mg/kg po dose.
Brain tissue binding (BTB); Whole blood binding (WBB) Ref. 20.
Determined from 1 mg/kg iv dose.
use of the more lipophilic ethyl sulfone (8h) caused a further loss
in activity at M1, suggesting a steric limit to the binding pocket.
Combining the methyl sulfone with the n-propoxy capping group

5438
D. J. Johnson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5434–5438
Table 3
Pharmacokinetic data for 8k
References and notes
a
a
a
1. (a) Caulfield, M. P.; Birdsall, N. J. M. Pharmacol. Rev. 1998, 50, 279; (b) Eglen, R.
M. Prog. Med. Chem. 2005, 43, 105; (c) Hulmes, E. C.; Birdsall, N. J. M.; Buckley,
N. J. Annu. Rev. Pharmacol. Toxicol. 1990, 30, 633.
Compd
Est. Clb
Br:Bla
1.7
BTBb
(%)
WBBb
(%)
Cfree,
(nM)
Cfree,
(nM)
brain
blood
(mL/min/kg)
2. Eglen, R. M. Auton. Autacoid Pharmacol. 2006, 26, 219.
8k
23
61
42
261
265
3. (a) Raedler, T. J.; Bymaster, F. P.; Tandon, R.; Copolov, D.; Dean, B. Mol.
Psychiatry 2007, 12, 232; (b) Fisher, A. Neurotherapeutics 2008, 5, 433.
4. (a) Langmead, C. J.; Fry, V. A. H.; Forbes, I. T.; Branch, C. L.; Christopoulos, A.;
Wood, M. D.; Herdon, H. J. Mol. Pharmacol. 2006, 69, 236; (b) Heinrich, J. N.;
Butera, J. A.; Carrick, T.; Kramer, A.; Kowal, D.; Lock, T.; Marquis, K. L.; Pausch,
M. H.; Popiolek, M.; Sun, S.; Tseng, E.; Uveges, A. J.; Mayer, S. C. Eur. J. Pharmacol.
2009, 605, 53.
a
Estimated from 3 mg/kg po dose.
Ref. 20.
b
5. Sauerberg, P.; Olesen, P. H.; Nielsen, S.; Treppendahl, S.; Sheardown, M. J.;
Honore, T.; Mitch, C. H.; Ward, J. S.; Pike, A. J.; Bymaster, F. P.; Sawyer, B. D.;
Shannon, H. E. J. Med. Chem. 1992, 35, 2274.
6. Loudon, J. M.; Bromidge, S. M.; Brown, F.; Clark, M. S.; Hatcher, J. P.; Hawkins, J.;
Riley, G. J.; Noy, G.; Orlek, B. S. J. Pharmacol. Exp. Ther. 1997, 283, 1059.
7. Bodick, N. C.; Offen, W. W.; Levey, A. I.; Cutler, N. R.; Gauthier, S. G.; Satlin, A.;
Shannon, H. E.; Tollefson, G. D.; Rasmussen, K.; Bymaster, F. P.; Hurley, D. J.;
Potter, W. Z.; Paul, S. M. Arch. Neurol. 1997, 54, 465.
8. Spalding, T. A.; Trotter, C.; Skjaerbaek, N.; Messier, T. L.; Currier, E. A.; Burstein,
E. S.; Li, D.; Hacksell, U.; Brann, M. R. Mol. Pharmacol. 2002, 61, 1297.
9. Bridges, T. M.; Brady, A. E.; Kennedy, J. P.; Daniels, R. N.; Miller, N. R.; Kim, K.;
Breininger, M. L.; Gentry, P. R.; Brogan, J. T.; Jones, C. K.; Conn, P. J.; Lindsley, C.
W. Bioorg. Med. Chem. Lett. 2008, 18, 5439.
10. Budzik, B.; Garzya, V.; Shi, D.; Foley, J. J.; Rivero, R. A.; Langmead, C. L.; Watson,
J.; Wu, Z.; Forbes, I. T.; Jin, J. Bioorg. Med. Chem. Lett., doi:10.1016/
j.bmcl.2010.04.128.
11. Budzik, B.; Garzya, V.; Shi, D.; Walker, G. W.; Lauchart, Y.; Lucas, A. J.; Rivero, R.
A.; Langmead, C. L.; Watson, J.; Wu, Z.; Forbes, I. T.; Jin, J. Bioorg. Med. Chem.
Lett., doi:10.1016/j.bmcl.2010.04.127.
12. (a) Garzya, V. Presented at SCI-RSC International Symposium on Medicinal
Chemistry, Cambridge, UK, Sept 2009.; (b) Budzik, B.; Garzya, V.; Shi, D.;
Walker, G.; Woolley-Roberts, M.; Pardoe, J.; Lucas, A.; Tehan, B.; Rivero, R. A.;
Langmead, C. J.; Watson, J.; Wu, Z.; Forbes, I. T.; Jin, J. ACS Med. Chem. Lett.,
doi:10.1021/ml100105x.
Figure 2. NOR results for 8k.
13. Noyce, D. S.; Weingarten, H. I. J. Am. Chem. Soc. 1957, 79, 3093.
14. Forbes, I. T. Tetrahedron Lett. 2001, 42, 6943.
15. For full experimental procedures and NMR and LC/MS characterisation see
Cooper, D. G.; Forbes, I. T.; Garzya, V.; Johnson, D. J.; Stevenson, G. I.; Wyman, P.
A. WO2009/037294.
16. Muscarinic pharmacology characterised using FLIPR (Fluorometric Imaging
Plate Reader) calcium mobilisation assays. Assays utilise Chinese hamster ovary
(CHO) cells stably expressing human M1, M2, M3, M4 or M5 receptors. M2 and M4
cells also stably express Gqi5 chimaeric G-protein. 11 point concentration–
response curves were generated. Values are average of at least five experiments.
17. Antagonist selectivity was not investigated due to agonist induced
desensitisation of the FLIPR assay.
(8k) showed a slightly reduced potency at M1 of 8.0 but a c log P of
1.53. The pharmacokinetic profile in rat of 8k gave a low estimated
blood clearance (23 mL/min/kg) and the free concentrations in the
brain (261 nM) and blood (265 nM) were almost exactly the same
(Table 3). This gave evidence that 8k was not a substrate for trans-
porters in the blood brain barrier.
Modulation of global physicochemical properties of the mole-
cule had furnished a compound with the desired potency, meta-
bolic stability and brain penetration. Compound 8k was therefore
selected for further profiling in the CEREP high-throughput panel,
only showing significant activity at the sigma-receptor (75% inhibi-
18. Seelig, A. Eur. J. Biochem. 1998, 251, 252.
19. Read, K. D.; Braggio, S. Expert Opinion on Drug Metabol. & Toxicol. 2010, 6, 337.
20. Summerfield, S. G.; Stevens, A. J.; Cutler, L.; Carmen-Osuna, M. D.; Hammond,
B.; Tang, S.; Hersey, A.; Spalding, D. J.; Jeffrey, P. J. Pharmacol. Exp. Ther. 2006,
316, 1282.
21. For
a full list of assays tested see http://www.cerep.fr/cerep/users/pages/
tion of control specific binding at 10 l
M).21 With its selective
catalog/Profiles/catalog.asp.
22. The novel object recognition (NOR) model was used to assess the effects on
recognition memory in rats. Ennaceur and Delacour (Behavioural Brain. Res.,
1988, 31, 47) and is a non-rewarded test of spontaneous behaviour based on
the differential exploration of novel and familiar objects. The NOR test
comprised of two sessions, T1 and T2, each lasting 3 min. On the first test
day (T1), following a 3 min habituation to the empty test box, the rat was
placed into the annex and 2 identical objects were placed into the test arena.
The rat was then returned to the test area and allowed to freely explore the
objects for 3 min. On the second day (T2), a similar protocol was adopted,
except that one of the ‘familiar’ objects was substituted for a novel one of the
same colour, material and similar size but different shape. In order to detect a
potential cognition-enhancing effect an inter-trial interval (ITI) of 24 h
between T1 and T2 was used to induce a performance deficit. The objects
comprised of black hardened plastic geometric shapes, such as towers and
pyramids, that were approximately 6.5 cm high ꢁ 6 cm diameter. Exploration
was scored as time spent sniffing or licking the objects. Sitting on or climbing
over the objects was not considered to be exploratory behaviour. Between
trials, objects were cleaned with 70% ethanol to remove olfactory cues. All
trials were recorded via camera. Exploration was scored retrospectively by an
experimenter blind to the novel object. The presentation of objects was
balanced across days, trials and box position (left/right) to minimise bias. Data
were expressed as a d1 index (time spent exploring novel object —time spent
exp exploring familiar object) and d2 index ([time exploring novel object—
time exploring familiar object]/total exploration time). T1 exploration times
were also considered since hyperactivity or sedation may confound the results.
The d1 and d2 index scores were analyzed using a two-way analysis of
variance (ANOVA, object ꢁ treatment) followed by planned comparisons for
between groups analysis using Statistica version 6 (Statsoft Inc., Tulsa). Data
are expressed as mean SEM and were considered significant when p <0.05.
pharmacology demonstrated 8k was advanced into an in vivo cog-
nition model. The novel objection recognition model of temporal
induced memory deficit in rat was selected and 8k was dosed from
0.1–1 mg/kg (Fig. 2).22 It demonstrated a robust improvement in
the memory of rats in the 1 mg/kg dose group compared to vehicle,
as well producing a dose related increase across the dosing groups.
This showed 8k to have a pro-cognitive effect in pre-clinical
species.
In conclusion, we have identified a series of N-substituted
3-(4-piperidinyl)-benzoxazolinones and oxindoles as highly brain
penetrant, selective muscarinic M1 agonists. By modulation of
the physicochemical properties, particularly lipophilicity, we were
able to achieve good metabolic stability without compromising po-
tency or brain penetration. This led to the discovery of compound
8k, which was further profiled in an in vivo model for cognition,
demonstrating a robust pro-cognitive effect with a minimum effi-
cacious dose of 1 mg/kg.
Acknowledgements
The authors thank Elizabeth Conwey and Sebastien Malaterre
for the synthesis of intermediates and Mark Vine for NMR analysis.