Discovery of GlyT1 inhibitors with improved pharmacokinetic properties

Article abstract of DOI:10.1016/j.bmcl.2009.01.015

Glycine transporter 1 (GlyT1) represents a novel target for the treatment of schizophrenia via the potentiation of glutamatergic NMDA receptors. The discovery of 4,4-disubstituted piperidine inhibitors of GlyT1 which exhibit improved pharmacokinetic prope

Full text of DOI:10.1016/j.bmcl.2009.01.015

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1492–1495
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of GlyT1 inhibitors with improved pharmacokinetic properties
a
a
a
b
Scott E. Wolkenberg a, , Zhijian Zhao , David D. Wisnoski , William H. Leister , Julie O’Brien ,
Wei Lemaire b, David L. Williams Jr. b, Marlene A. Jacobson b, Cyrille Sur b, Gene G. Kinney b,
Doug J. Pettibone b, Philip R. Tiller c, Sheri Smith c, Christopher Gibson c, Bennett K. Ma c,
Stacey L. Polsky-Fisher c, Craig W. Lindsley a, , George D. Hartman a
*
a Department of Medicinal Chemistry, Technology Enabled Synthesis Group, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
b Department of Neuroscience, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
c Department of Drug Metabolism, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Glycine transporter 1 (GlyT1) represents a novel target for the treatment of schizophrenia via the poten-
tiation of glutamatergic NMDA receptors. The discovery of 4,4-disubstituted piperidine inhibitors of
GlyT1 which exhibit improved pharmacokinetic properties, including oral bioavailability, is discussed.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 24 November 2008
Revised 6 January 2009
Accepted 7 January 2009
Available online 10 January 2009
Keywords:
GlyT1
Schizophrenia
NMDA pharmacokinetics
Cassette dosing
Mounting evidence suggests that the long established dopa-
mine hyperfunction model1 for schizophrenia may inadequately
account for the symptoms of this widespread disease.2,3 Although
both typical and atypical antipsychotics target dopamine receptors
and address the positive symptoms of schizophrenia (hallucina-
tions, paranoia, and other delusions), negative symptoms (blunted
affect, withdrawal) and cognitive deficits are not satisfactorily ad-
dressed with these treatments. A growing body of evidence4 indi-
O
NH2
O
Cl
N
H
N
N
H
Cl
Cl
N
O S
N
O S
O
O
1
2
Figure 1. 4,4-Disubstituted piperidine GlyT1 inhibitors.
cates that hypofunction of N-methyl-D-aspartate (NMDA)
glutamatergic receptors may contribute to the etiology of the dis-
ease, yet direct agonists of NMDA receptors are neurotoxic.5 Gly-
Furthermore, 1 significantly enhances prepulse inhibition in DBA/
2J mice without impairing basal startle amplitude indicating an
antipsychotic effect without sedation.7
While 1 and 2 were effective in animal models after being dosed
subcutaneously, they exhibit poor pharmacokinetic properties
including low bioavailability. In addition to optimizing these prop-
erties to enable oral dosing, a further objective was developing
compounds which demonstrate high transporter occupancy
in vivo. Herein we report the achievement of these goals employ-
ing an iterative analogue library approach.
cine transporter
1 (GlyT1) has emerged as a promising
alternative target due to existing reports of its potentiation of
NMDA receptor activity by modulating the local concentrations
of the NMDA co-agonist glycine.6
Recently, we disclosed potent and selective inhibitors of GlyT1
based on a 4,4-disubstituted piperidine lead structure. Exemplified
by 1 and 2 (Fig. 1), these compounds exhibit potent (<10 nM),
selective (versus GlyT2, taurine transporter) inhibition of GlyT1
and selectively elevate glycine levels in rat prefrontal cortex.7,8
A variety of strategies were pursued in order to improve the
properties of 1 and 2, including modification of the piperidine C4
substituent and substitution of the piperidine N-sulfonamide. Aryl
group replacements were prepared according to the route de-
scribed in Scheme 1a. Quenching the lithium anion of nitrile 3 with
a variety of electrophiles (alkyl halides, epoxides), followed by Ra-
* Corresponding author. Tel.: +1 215 652 0821; fax: +1 215 652 6345.
E-mail address: scott_wolkenberg@merck.com (S.E. Wolkenberg).
Present address: Departments of Pharmacology and Chemistry, Vanderbilt
Program in Drug Discovery, Vanderbilt Institute of Chemical Biology, Vanderbilt
University Medical Center, Vanderbilt University, Nashville, TN 37232, USA.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.015

S. E. Wolkenberg et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1492–1495
1493
Table 1
A
Piperidine 4-position replacements
CN
R1
CN
R1
O
Cl
NH2
c
a
b
R1
N
H
N
N
N
Cl
Boc
Boc
Boc
4
3
N
S
O
O
O
O
R1
R1
N
H
R
N
H
R
Compound
R1
GlyT1
GlyT2
TauT
d, e
IC50, nMa
IC50, nMa
IC50, nMa
N
Boc
N
S
O
O
O
R2
5
2.9
>30,000
>30,000
>30,000
B
O
6
7
26.5
2.8
>30,000
>30,000
CN
R1
CN
R1
CN
R2
b
a
d, e
>30,000
ndb
N
Boc
N
N
S
Boc
O
3
OH
O
O
R1
R1
8
9
17.7
nd
NH2
N
H
R
c
OAc
N
N
S
N
S
O
O
O
O
R2
38.1
30.0
nd
nd
R2
Scheme 1. Reagents and conditions: (a) LHMDS, THF, 25 °C, then electrophile R1–X;
(b) H2, Raney Ni; (c) carboxylic acid, PS-DCC, HOBt, CH2Cl2, 25 °C; (d) 4 M HCl/
dioxane; (e) sulfonyl chloride, i-Pr2NEt, 25 °C.
10
>30,000
>30,000
a
Values are means of at least three experiments.
Not determined.
b
ney Ni-catalyzed reduction provided mono-Boc protected diamine
4. Acylation (PS-DCC, HOBt, or carboxylic acid chloride), acidic
deprotection of the Boc group, and sulfonamide formation fur-
nished the GlyT1 inhibitors listed in Table 1. Alternatively, the se-
quence could be modified (Scheme 1b) such that amide formation
furnished the final products. Use of both routes enabled the facile
generation of both amide and sulfonamide libraries; >300 ana-
logues were prepared according to this procedure and selected
data are presented in Table 1.9
Generally, replacement of the aryl group was well tolerated,
with tetrahydropyran (5) and hydroxyl (7) functionality providing
potency equivalent to 1 and 2. These substitutions had no effect on
selectivity versus GlyT2 and taurine transporter (TauT)—no activity
was observed at micromolar concentrations. This level of selectiv-
ity (>30,000-fold) is notable among non-sarcosine derived GlyT1
inhibitors.10 Despite the variety of structures incorporated at the
piperidine C4 position and the removal of likely sites of metabo-
lism (aromatic oxidation and N-oxide formation), compounds in
Table 1 exhibited uniformly poor pharmacokinetic properties
(dog Cl > 15 mL/min/kg, t1/2 < 2 h).
Evaluation of propylsulfonamide modifications was therefore
undertaken; a focused sulfonamide library was prepared according
to Scheme 1 and results are shown in Table 2. Metabolite profiling
of 2 indicated that oxidation occurs on the propyl chain, and poten-
tial metabolite 11 lost 20-fold in potency. The corresponding fluo-
ride (12), while only 2Â less potent than 2, demonstrated poor
pharmacokinetics. Alkyl sulfonamides (13–17) were 10- to 50-fold
less potent than 2, with the truncation and extension of the
straight alkyl chain by even a single methylene unit having a sig-
nificant negative effect (14 and 15, respectively). Sulfamide (18)
and trifluorinated alkyl chains (19 and 20) were poorly tolerated.
Although the truncated alkyl sulfonamides (13, 14, and 16) lose
potency versus 1 and 2, they constitute a breakthrough in terms of
pharmacokinetic properties (Table 2). These compounds exhibit
low clearance (<2 mL/min/kg) and long half lives (>7.5 h) in dogs.
The data cited in Table 2 were generated in a series of dog iv
cassette experiments, which was an attractive strategy due to the
number of compounds in the library for which pharmacokinetic
evaluation was desired.11 Although there are myriad potential
complications associated with dosing in cassette format,12 within
this structural series we noted excellent agreement between iv
clearance values determined via cassette and single dose experi-
ments. This strong correlation led to a high degree of confidence
in these data, and cassettes were followed with single dose exper-
iments evaluating promising compounds. Based on the dramatic
improvement in dog clearance afforded by the shortened alkyl sul-
fonamides, efforts were focused on enhancing potency with these
groups in place.
Previous experience gained during the development of 1 sug-
gested that incorporation of a chiral (S) methyl group alpha to
the amide nitrogen could confer increases in potency up to 10-
fold.7 The methylated analogue of 14 was therefore prepared
according to Scheme 2. Following incorporation of the 2-pyridyl
group via a nucleophilic aromatic substitution reaction, the methyl
group was installed through Grignard addition (MeMgBr, 25 °C)
and reduction (NaBH4) of the resulting imine. Following resolution
using chiral chromatography, acylation of the amine, piperidine N-
Boc deprotection, and sulfonamide formation provided optically
pure 21.13 Gratifyingly, 21 exhibited improved GlyT1 potency
(11 nM) while maintaining complete selectivity versus GlyT2 and
TauT; this analogue was not a substrate for human or rat P-glyco-
protein (PgP) and displayed excellent passive permeability
(37 Â 10À6 cm/s).14 Moreover, 21 retained a favorable dog pharma-
cokinetic profile (see Fig. 3 and Table 3). Data contained in Fig. 3

1494
S. E. Wolkenberg et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1492–1495
Table 2
CN
Propylsulfonamide replacements
CN
N
N
NH2
a
b, c
Cl
O
Cl
N
N
N
Boc
Boc
Boc
N
N
H
Cl
N
O
R2
N
N
H
d, e, f
R2
GlyT1
GlyT2
TauT
Dog PKb
Cl
Compound
N
O S
IC50, nMa
IC50, nMa
IC50, nMa
Clc
t1/2
d
O
21
2
4.4
>30,000
nde
>30,000
nd
15
2.3
O S
O
Scheme 2. Reagents and conditions: (a) 2-fluoropyridine, LHMDS, THF, 25 °C; (b)
i—MeMgBr, toluene, 25 °C; ii—NaBH4, MeOH; (c) ChiralPak AD; (d) 2,4-dichloro-
benzoylchloride, i-Pr2NEt, 25 °C; (e) 4 M HCl/dioxane; (f) EtSO2Cl, i-Pr2NEt, 25 °C.
OH
F
11
12
13
14
15
16
17
18
19
86
O S
O
Table 3
Selected dog pharmacokinetic data
9.7
>30,000
>30,000
>30,000
nd
>30,000
>30,000
>30,000
nd
O S
O
b
b
Compound
Cl (cass)a
Cl (single)b
t1/2
Vdss
F (%)b
21
22
3.6
0.84
3.5
2.3
5.0
11.8
1.37
2.1
25
38
O S
O
300
57
1.0
1.4
8.4
8.0
a
Dosed as part of an iv cassette, 0.25 mg/kg; see text for a discussion.
Dosed singly, 0.5 mg/kg iv; 1.0 mg/kg po; Cl in mL/min/kg; t1/2 in h; Vdss in L/kg.
b
O S
O
An alternative approach to improving pharmacokinetic profile
O S
O
120
200
220
77
while retaining high GlyT1 potency was also pursued. Libraries
were prepared in which potency enhancing piperidine C4 substit-
uents were combined with clearance lowering truncated alkyl
sulfonamides. The result was the discovery of 22 (Fig. 2). Unlike
the 4-pyridylpiperidine series, where truncation of the propylsulf-
onamide to an ethylsulfonamide results in a >10-fold decrease in
potency (compare 2 and 14 in Table 2), within the 4-cyclo-
propylmethylpiperidine series the propyl and ethyl sulfonamides
are equipotent (for 10 and 22 GlyT1 IC50 30 and 26 nM,
respectively).
In the 4-cyclopropylmethylpiperidine series incorporation of a
chiral methyl group did not confer an increase in potency (data
not shown). Despite these differences, 22 retains the favorable pro-
file of the 4-pyridylpiperidine series. Compound 22 is not a sub-
strate for human or rat PgP. Like 21, compound 22 exhibits low
clearance when dosed in either iv cassette (Fig. 3) or single format
(iv and po, Table 3). Compound 22 was suitable for evaluation in
the in vivo transporter occupancy assay and has a plasma Occ50
of 260 nM.
In conclusion, the pharmacokinetic profile of GlyT1 inhibitors
within the 4,4-disubstituted piperidine series has been dramati-
cally improved through modification of the sulfonamide alkyl
chain. Accompanying losses in potency could be recovered through
the installation of a chiral methyl group alpha to the amide nitro-
gen and through exchange of the 2-pyridyl group with a cyclopro-
pylmethyl substituent. The improvements include an increase in
bioavailability which allowed, for the first time, the measurement
O S
O
>30,000
nd
>30,000
nd
1.7
7.6
1.9
4.9
O S
O
O S
O
nd
nd
11.8
N
>5000
1070
nd
nd
O S
O
CF3
20
O S
O
nd
nd
2.5
CF3
a
Values are means of at least three experiments.
Determined in dog iv cassettes, 0.25 mg/kg—see text for a discussion.
b
c
Clearance in mL/min/kg.
Half life in hours.
Not determined.
d
e
are derived from a dog iv cassette experiment and illustrate the
utility of the approach. Simple visual inspection indicates the com-
parative stability of 21 and 22 (vide infra) versus 1, and while iv
clearance values were calculated, in general only a rank order of
stabilities for compounds in cassettes was generated. Based on this
order, compounds were prioritized for evaluation in single com-
pound pharmacokinetic studies from which quantitative parame-
ters were derived (Table 3). As noted previously, within this
series excellent agreement of cassette and single-compound clear-
ance rates was found.
O
Cl
N
H
Compound 21 was evaluated in a rat in vivo transporter occu-
pancy assay via oral dosing using the previously described proto-
Cl
N
O S
col.15
A plasma Occ50 of 150 nM was measured, and this
O
analogue represents the achievement of the desired profile with
respect to in vitro potency and selectivity, pharmacokinetics, and
in vivo occupancy.
22
Figure 2. Optimized GlyT1 inhibitor.

S. E. Wolkenberg et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1492–1495
1495
1000
100
10
thank Carl Homnick for performing chiral separations, Joan Mur-
phy and Charles W. Ross, III, for HRMS measurements, and Nancy
Tsou for X-ray crystal analysis of 21.
1
21
22
References and notes
1. (a) Seeman, P.; Lee, T.; Chau-Wong, M.; Wong, K. Nature 1976, 261, 717; (b)
Seeman, P.; Lee, T. Science 1975, 188, 1217.
2. Lindsley, C. W.; Shipe, W. D.; Wolkenberg, S. E.; Theberge, C. R.; Williams, D. L.;
Sur, C.; Kinney, G. G. Curr. Top. Med. Chem. 2006, 6, 771.
3. Andreasen, N. C. Brain Res. Rev. 2000, 31, 106.
4. (a) Javitt, D. C.; Zukin, S. R. Am. J. Psychiatry 1991, 148, 1301; (b) Olney, J. W.;
Farber, N. B. Arch. Gen. Psychiatry 1995, 52, 998; (c) Olney, J. W.; Newcomer, J.
W.; Farber, N. B. J. Psychiatry Res. 1999, 33, 523.
5. Marino, P. J.; Conn, P. J. Curr. Drug Targets CNS Neurol. Disord. 2002, 1, 1.
6. Sur, C.; Kinney, G. G. Expert Opin. Investig. Drugs 2004, 13, 515. and references
therein.
0
5
10
Time (h)
15
20
7. Lindsley, C. W.; Zhao, Z.; Leister, W. H.; O’Brien, J.; Lemaire, W.; Williams, D. L.;
Chen, T.-B.; Chang, R. S. L.; Burno, M.; Jacobson, M. A.; Sur, C.; Kinney, G. G.;
Pettibone, D. J.; Tiller, P. R.; Smith, S.; Tsou, N. N.; Duggan, M. E.; Conn, P. J.;
Hartman, G. D. Chem. Med. Chem. 2006, 1, 807.
8. Zhao, Z.; Leister, W. H.; O’Brien, J. A.; Lemaire, W.; Williams, D. L.; Jacobson, M.
A.; Sur, C.; Kinney, G. G.; Pettibone, D. J.; Tiller, P. R.; Smith, S.; Hartman, G. D.;
Lindsley, C. W.; Wolkenberg, S. E. Bioorg. Med. Chem. Lett. 2009, 19, 1488.
Figure 3. Dog iv cassette data; 0.25 mg/kg; see text for a discussion.
of in vivo transporter occupancy using oral dosing. Compounds 21
and 22 were identified which occupy brain GlyT1 in vivo at plasma
concentrations <300 nM. The use of an iterative analogue library
approach enabled the optimization of both potency and pharmaco-
kinetic properties despite non-additive SAR. Key to the successful
discovery of enhanced inhibitors was the extensive use of cassette
dosing in dog pharmacokinetic experiments which, when
combined with judicious follow up in single dose experiments,
allowed the screening of hundreds of compounds for improved
pharmacokinetics.
9. Compound
9 was prepared by an alternative route employing a Strecker
reaction of N-benzyl-4-piperidone and azetidine.
10. (a) Pinard, E.; Ceccarelli, S. M.; Stalder, H.; Alberati, D. Bioorg. Med. Chem. Lett.
2006, 16, 349; (b) Ceccarelli, S. M.; Pinard, E.; Stalder, H.; Alberati, D. Bioorg.
Med. Chem. Lett. 2006, 16, 354.
11. (a) Berman, J.; Halm, K.; Adkison, K.; Shaffer, J. J. Med. Chem. 1997, 40, 827; (b)
Shaffer, J. E.; Adkison, K. K.; Halm, K.; Hedeen, K.; Berman, J. J. Pharm. Sci. 1999,
88, 313.
12. White, R. E.; Manitpisitkul, P. Drug Metab. Dispos. 2001, 29, 957.
13. The absolute configuration of 21 was confirmed by single X-ray crystal analysis
(anomalous dispersion).
14. Yamazaki, M.; Neway, W. E.; Ohe, T.; Chen, I-W.; Rowe, J. F.; Hochman, J. H.;
Chiba, M.; Lin, J. H. J. Pharmacol. Exp. Ther. 2001, 296, 723.
15. Zeng, Z.; O’Brien, J. A.; Lemaire, W.; O’Malley, S. S.; Miller, P. J.; Zhao, Z.;
Wallace, M. A.; Raab, C.; Lindsley, C. W.; Sur, C.; Williams, D. L. Nucl. Med. Biol.
2008, 35, 315.
Acknowledgments
We are grateful to Audrey Wallace for dog PK experiments and
Ken Anderson and Kristi Hoffman for analytical measurements. We