Identification of N-(2-(azepan-1-yl)-2-phenylethyl)-benzenesulfonamides as novel inhibitors of GlyT1

Article abstract of DOI:10.1016/j.bmcl.2010.06.085

A novel series of glycine transporter 1 (GlyT1) inhibitors is described. Scoping of the heterocycle moiety of hit 4-chlorobenzenesulfonamide 1 led to replacement of the piperidine with an azepane for a modest increase in potency. Phenyl sulfonamides proved superior to alkyl and non-phenyl aromatic sulfonamides, while subsequent ortho substitution of the 2-(azepan-1-yl)-2- phenylethanamine aromatic ring yielded 39 (IC₅₀ 37 nM, solubility 14 μM), the most potent GlyT1 inhibitor in this series. Favorable brain-plasma ratios were observed for select compounds in pharmacokinetic studies to evaluate CNS penetration.

Full text of DOI:10.1016/j.bmcl.2010.06.085

Bioorganic & Medicinal Chemistry Letters 20 (2010) 4878–4881
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of N-(2-(azepan-1-yl)-2-phenylethyl)-benzenesulfonamides as
novel inhibitors of GlyT1
Jeffrey G. Varnes, Janet M. Forst, Tiffany N. Hoerter, Christopher R. Holmquist, Deidre E. Wilkins,
Gaochao Tian, Gerald Jonak, Xia Wang, William M. Potts, Michael W. Wood, Cristóbal Alhambra,
*
Todd A. Brugel, Jeffrey S. Albert
CNS Discovery Research, AstraZeneca Pharmaceuticals, 1800 Concord Pike, Wilmington, DE 19850, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of glycine transporter 1 (GlyT1) inhibitors is described. Scoping of the heterocycle moiety
of hit 4-chlorobenzenesulfonamide 1 led to replacement of the piperidine with an azepane for a modest
increase in potency. Phenyl sulfonamides proved superior to alkyl and non-phenyl aromatic sulfona-
mides, while subsequent ortho substitution of the 2-(azepan-1-yl)-2-phenylethanamine aromatic ring
Received 21 April 2010
Revised 13 June 2010
Accepted 15 June 2010
Available online 19 June 2010
yielded 39 (IC50 37 nM, solubility 14 lM), the most potent GlyT1 inhibitor in this series. Favorable
brain–plasma ratios were observed for select compounds in pharmacokinetic studies to evaluate CNS
penetration.
Keywords:
GlyT1
NMDAr
Ó 2010 Elsevier Ltd. All rights reserved.
NMDA
Glycine
Transporter
Inhibitor
Schizophrenia
Schizophrenia affects approximately 1% of the world’s popula-
tion; however, current therapies are associated with unwanted side
affects and/or do not provide sufficient relief across all symptom
domains.1 Based on pre-clinical models and clinical data, the
O
S
O
Cl
N
HN
N-methyl-D-aspartate receptor (NMDAr) hypofunction hypothesis
of schizophrenia correlates disease symptomology with glutama-
tergic neurotransmission dysfunction.2 Hence, increasing NMDAr
neurotransmission represents a promising approach for the devel-
opment of antipsychotics. Direct NMDAr agonists and elevated
levels of its endogenous ligand glutamate are known to be neu-
rotoxic. In contrast, targeted elevation of synaptic glycine, an
NMDAr coagonist, represents a viable opportunity for neurotrans-
mission enhancement. Synaptic glycine levels are regulated by gly-
cine transporter 1 (GlyT1), making the blockade of GlyT1-mediated
glycine reuptake an attractive method for enhancing NMDAr
neurotransmission. Considerable effort has been invested in discov-
ering GlyT1 inhibitors and the area has been extensively revie-
wed.2b,3 Herein, we describe a novel series of GlyT1 inhibitors as a
potential therapy for schizophrenia.
1
Our lead-finding strategy relied on a virtual screen to identify
compounds that had pharmacophoric similarity with validated
GlyT1 inhibitors from the literature. The computational screening
model employed in-house developed 2D topological fingerprints.
They encoded the distribution of various pharmacophore features
(e.g., aromaticity, lipophilicity, H-bond donors and acceptors, posi-
tive and negative charges) up to distances of 12 bonds. Compounds
that met our similarity criteria to literature inhibitors were clus-
tered and a representative sample of members from each cluster
was filtered for suitable CNS-drug like properties (e.g., molecular
weight, lipophilicity) and the absence of toxic or reactive func-
tional groups. A final set of 15,000 candidates was submitted for
biological screening. These were tested in a GlyT1 uptake inhibi-
tion assay4 at 10
The most potent hit was 4-chlorobenzenesulfonamide
(IC50 1.4 0.1 M) which contains elements that are related to
lM.
1
l
some reported GlyT1 inhibitors. For example aryl sulfonamides
* Corresponding author. Tel.: +1 302 886 4771; fax: +1 302 885 1806.
E-mail address: Jeffrey.albert@astrazeneca.com (J.S. Albert).
also appear in the 1,3-diaminopropanol series from GSK,5 and a
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.06.085

J. G. Varnes et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4878–4881
4879
Table 1
related benzyl linked amine comprises the basic region in
SSR504734 from Sanofi.2d Intrigued by the potential for rapid
structural modification, we embarked on a campaign to scope the
structure–activity relationship of the phenethyl diamine scaffold
of 1 with respect to GlyT1 inhibition.
Uptake inhibition and aqueous solubility data for 1 and 7–13
O
Compounds were prepared from commercially available
2-amino-2-phenylethanamines via sulfonylation or acylation.
Where said phenylethanamines could not be purchased, they were
obtained via a Strecker reaction between an amine and aldehyde
followed by an alane reduction.6,7 For example, as shown in
Scheme 1, the aldehyde used to prepare compound 2 was derived
from 4-mercaptobenzoic acid (3) by first alkylating with propyl io-
dide and then oxidizing with OxoneÒ to afford sulfone 4. Reduction
of the ester to the corresponding alcohol was followed by oxidation
to give aldehyde 5. A Strecker reaction between azepane and 5 pro-
vided the corresponding nitrile, which was reduced with alane to
furnish amine 6. Subsequent sulfonylation yielded 2. If amide ana-
logs of 2 were desired, they were obtained via acylation (ArCOCl,
DIPEA, CH2Cl2) or amide coupling (ArCOOH, HOBt, TBTU, DIPEA,
DMF).8
R
HN
S
O
Cl
b
Compda
R
IC50
(
l
V)
Sol
(lV)
1
7
8
Piperidin-1-yl
1.4 0.1
1.4 0.1
>20
>20
0.46 0.05
13
10
9.7 0.7
3
<1
16
240
1
<1
4-Me-piperidin-1-yl
Morpholin-4-yl
4-Et-piperazin-1-yl
Azepan-1-yl
3-Azabicyclo[3.2.2]nonane-1-yl
Diethylamin-1-yl
9
10
11
12
13
3
1
300
47
Pyrrolidin-1-yl
a
Compounds are racemic.
Equilibrium solubility (pH 7.4).9
b
Table 2
We started our investigation by evaluating the impact of
changes to the piperidine moiety of 1 (Table 1). As demonstrated
with morpholine 8 and piperazine 9, attenuation of nitrogen basi-
city resulted in complete loss of activity, suggesting that a basic
heterocyclic nitrogen is important for potent inhibition.
Uptake Inhibition and aqueous solubility data for 14–22
O
Changes to the size of the piperidine ring of 1 gave mixed
results. Expansion from a six- to a seven-membered ring (1 and
10, respectively) was favored while conversion of 10 to the corre-
sponding azabicyclo[3.2.2]nonane 11 resulted in a 10-fold loss of
potency. Ring truncation (diethylamine 12) or contraction
(pyrrolidine 13) also decreased in vitro glycine uptake inhibition
by approximately 10-fold. These results suggest that there is a
narrow SAR for the heterocycle moiety of the phenethyl diamine
scaffold, and only certain ring sizes and substitutions (e.g., 7) main-
tain or improve potency.
To determine what impact modifications might have on both
potency and aqueous solubility, we examined alkyl, benzyl and
alternative heterocyclic sulfonamides for comparison with simple
phenyl analog 14 (Table 2). Initial data suggested that an aromatic
moiety is required for activity. For example, while both compounds
N
HN
S
O
R
b
Compda
R
IC50
(
l
V)
Sol
(lV)
14
15
16
17
18
19
20
21
22
Phenyl
Ethyl
cPropyl
Benzyl
3,5-Cl-benzyl
Pyridin-3-yl
Quinolin-8-yl
4
>20
>20
1
67
470
371
195
1
143
2
14.8 0.7
1.5 0.1
>20
4
2
2-Br-thiophen-5-yl
3,5-Me-isoxazol-4-yl
0.93 to 0.9
8.3 0.3
9
37
a
Compounds are racemic.
Equilibrium solubility (pH 7.4).9
b
showed marked improvement in solubility relative to their aro-
matic counterparts, ethyl (15) and cyclopropyl (16) sulfonamides
were inactive. Several results indicated that large lipophilic aro-
matic groups are well tolerated. Thus, while only weak activity
was observed for benzyl sulfonamide 17, bis-chlorination (18) im-
proved potency by 10-fold. Similarly, potency was retained in
going from 14 to quinoline 20. Thiophene 21, with a large bromo
substituent, was the most potent phenyl sulfonamide surrogate
tested. In contrast, an attempt to increase solubility by conversion
of the phenyl sulfonamide of 14 to a more polar pyridine analog
(19) completely eliminated activity.
We concluded from the data of Table 2 that phenyl sulfona-
mides offer the best combination of potency and solubility,
prompting us to aggressively investigate the scope of sulfonamide
phenyl ring substitution (Table 3). In general, meta and para-
substituted phenyl sulfonamides with large hydrophobic substitu-
ents were more active than sulfonamides with alternative substi-
tution patterns or polar substituents. Supporting this hypothesis
are para-substituted bromo and trifluoromethoxy analogs 34 and
35, respectively, which are two of the more potent compounds of
Table 3. In contrast, replacing chlorine with fluorine (e.g., 10 to
33 and 26 to 25) improved solubility but reduced potency. Electron
withdrawing cyano (32 and 36) and electron donating methoxy
(37) moieties gave similar results, while sulfone 38 was completely
O
S
O
S
O
O
HS
a, b
c, d
O
O
O
HO
3
O
4
5
e, f
O
S
O
S
O
O
g
O
S
O
N
HN
Cl
N
6
NH2
2
Scheme 1. Reagents: (a) Iodopropane, K2CO3, DMF, 34%; (b) OxoneÒ, DMF, 90%; (c)
DIBAL-H, THF, 46%; (d) Dess-Martin periodinane, CH2Cl2, 96%; (e) azepane, HOAc,
NaCN, 43%; (f) LiAlH4, H2SO4, THF; Na2SO4. 10 H2O, 51%; (g) 4-Cl-PhSO2Cl, Et3N,
CH2Cl2, 68%.

4880
J. G. Varnes et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4878–4881
Table 3
potent than less lipophilic 43, again suggesting that lipophilic
electron withdrawing groups at the para position are favored.
Ortho substitution had the greatest impact on in vitro potency,
particularly when that substituent was a methoxy group (39 and
46). Not only was there more than a two- to fivefold increase in up-
take inhibition, but a significant increase in solubility was also ob-
served for 39 compared to 35 (Table 5). In contrast, an electron
withdrawing group at the ortho position (40) decreased potency
by two orders of magnitude. With regards to other alkoxy substi-
tuted analogs, dioxolane 44 was potent but less active than 39,
and had little to no aqueous solubility. Bis-substituted 45, which
could potentially capitalize on increased binding interactions via
rotation of the phenyl ring, was weakly active.
Uptake inhibition and aqueous solubility data for 10 and 23–38
O
S
O
R
N
HN
Compda
R
IC50
0.46 0.05
15
2.1 0.7
(
l
V)
Sol
(lV)
b
10
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
4-Cl
2-F
2-OCF3
2,4-diF
2,4-diCl
3-CF3
1
1
108
10
19
1
4
7
8
1
1.8 0.2
0.83 0.04
0.68 0.02
No compounds in this series showed any appreciable activity at
GlyT2 and can thus be considered GlyT1-selective. For 39 the GlyT2
IC50 was >20 lM.
3-Cl
3,5-diCl
3,4-diF
3-Cl, 4-F
0.391 0.009
1.0 0.3
2
To determine CNS penetration of N-(2-(azepan-1-yl)-2-phenyl-
ethyl)-benzenesulfonamides, select compounds were evaluated in
a cassette CNS exposure screen in male Sprague-Dawley rats. Com-
pounds 39, 47, and 48 were selected based on their potencies and
relative solubilities. Appreciable whole brain and plasma concen-
trations were observed one hour post subcutaneous injection.
Brain–plasma ratios ranged from 1.6 to 2.1, indicating good parti-
tioning across the blood–brain barrier and distribution into brain
tissue.
10
17
14
9
1
2
7
30
4
0.6 0.2
3-CN, 4-F
4-F
6
2
2.9 0.1
0.39 0.01
0.17 0.03
4-Br
4-CF3
4-CN
4-OMe
4-SO2Me
5
2
4.6 0.9
>20
a
Compounds are racemic.
b
Equilibrium solubility (pH 7.4).9
In conclusion, starting from hit 4-chlorosulfonamide 1, the aze-
pane amine was identified as the only heterocyclic replacement
more potent than a piperidine. Solubility was quickly identified
as a physical property to be monitored and, if possible, improved.
An analysis of various aromatic and non-aromatic sulfonamides
identified phenyl sulfonamides as having the best balance between
in vitro potency and solubility. Substitution of the meta- and/or
para-positions of the phenyl sulfonamide with hydrophobic groups
was preferred, with 4-trifluoromethyl sulfonamide 35 offering the
most potent analog. Subsequent exploration of phenyl ring substi-
tution of the 2-amino-2-phenylethanamine scaffold of 35 identi-
fied the 2-methoxy group (39) as a way to improve both uptake
inhibition and solubility. To evaluate the CNS penetration of
N-(2-(azepan-1-yl)-2-(2-methoxy-phenylethyl)-benzenesulfona-
mides, exemplar 4-trifluoromethyl, 4-chloro, and 3-chloro-4-
fluoro-sulfonamides were analyzed via in vivo pharmacokinetic
studies in Sprague-Dawley rats. All three compounds dem-
onstrated good distribution into brain tissue and favorable
brain–plasma ratios.
inactive. Hydrophobic substituents at the ortho position (24) re-
tained only modest activity. Interestingly, compound 31, with po-
lar and non-polar moieties at the para and meta positions,
respectively, represents a successful increase in solubility while
maintaining activity at GlyT1 comparable to 10.
We used 4-trifluoromethyl sulfonamide 35 as a platform to
scope the impact of substitution of the phenyl ring of the 2-(azepan-
1-yl)-2-phenylethanamine scaffold (Table 4). Meta substitution was
only briefly explored and was not well-tolerated (data not shown).
Electron withdrawing groups at the para position were weakly ac-
tive. Of sulfones and sulfonamides 2 and 40–43, only propylsulfone
2 retained any significant potency. Solubility was also low, which
was likely due to the hydrophobic nature of the propyl side chain.
Sulfonamide 42 was weakly active but approximately fourfold more
Table 4
Uptake inhibition and aqueous solubility for 2, 35, and 39–46
Table 5
CNS exposure data for 39, 47 and 48a
R
O
O
N
HN
S
O
CF3
O
S
R
N
HN
O
b
Compda
R
IC50
(lV)
Sol
(lV)
Compdb
R
IC50
(
l
V)
Sol
Brain
Plasma Br/Pld
2
35
39
40
41
42
43
44
45
46
4-SO2Pr
H
2-OMe
2-SO2Me
4-SO2Me
4-SO2N(Me)2
4-SO2NH2
2,3–(OCH2O)
2,6-diOMe
2-OMe, 4-F
0.3 0.2
0.17 0.03
0.037 0.006
3.2 0.8
1
(l
V)c (nmol/kg) (nM)
2
14
5
3
<1
9
<1
215
2
39
47
48
4-CF3
4-Cl
3-Cl, 4-F
0.037 0.006 14
0.15 0.08
0.3 0.1
154
202
168
72
103
105
2.1
1.9
1.6
9
37
10
2
2
2
8.9 0.7
0.12 0.07
16.4 0.9
0.07 0.01
a
Total plasma and whole brain concentrations at 1 h following a 2.5 lmol/kg
subcutaneous dose.
b
Compounds are racemic.
Equilibrium solubility (pH 7.4).9
c
d
a
Brain–plasma ratio; brain concentrations not corrected for 2–3% vascular
Compounds are racemic.
Equilibrium solubility (pH 7.4).9
b
contamination.

J. G. Varnes et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4878–4881
Assay: To the wells of an OptiPlate, 2
4881
l
L of DMSO containing a test compound
Acknowledgments
are spotted. This was followed by addition of 98
million/mL final). After incubating cells with compound for ꢀ15 min, 100
the SPA (200 g/well final) and isotope mixture (30 nM isotope with 10 M cold
glycine, final) was added to initiate the glycine uptake. At 2 h, the plate was read
on TopCount to quantify SPA counts. Results: IC50s are reported as the
l
L of cell suspension (ꢀ1
l
L of
The authors are grateful to Jennifer Van Anda, David Coomber,
and Xiaomei Ye for purification of many of the final compounds.
l
l
a
geomean and their corresponding standard deviation based on an n of 2, 4, or 6.
5. Rahman, S. S.; Coulton, S.; Herdon, H. J.; Joiner, G. F.; Jin, J.; Porter, R. A. Bioorg.
Med. Chem. Lett. 2007, 17, 1741.
References and notes
6. General procedure for the Strecker reaction: Acetic acid (1.1 equiv) was added to a
mixture of benzaldehyde (1 equiv) in amine (12 equiv). Sodium cyanide
(1.1 equiv) in 2 mL of water was added, and the resulting solution was stirred
at room temperature until complete based on thin-layer chromatography (TLC)
and/or LCMS analysis. The reaction was diluted with water, extracted with ethyl
acetate (3Â), and the combined organic layers were dried over sodium sulfate,
filtered and concentrated. The resulting residue was purified by flash column
chromatography (SiO2, ethyl acetate in hexanes) to afford the desired nitrile.
Sluggish Strecker reactions or those with significant precipitate were warmed
gently at 40 °C during stirring to achieve homogeneity.
7. General procedure for alane reduction: To a solution of sulfuric acid (2 equiv) in
tetrahydrofuran (0.2 M) at 0 °C was added 2.0 M lithium aluminum hydride in
tetrahydrofuran (4 equiv) rapidly dropwise and the resulting cloudy mixture
was maintained at 0 °C for 30 min. The nitrile (1 equiv) was then added in
tetrahydrofuran (0.2 M) via cannula, and the reaction was warmed to room
temperature. After another 30 min, the reaction was cooled to 0 °C, and excess
sodium sulfate decahydrate was added, which was accompanied by an
exotherm. The resulting white mixture was diluted with ethyl acetate and
stirred for 10 min before being filtered. The filtrate was concentrated and the
resulting amine was carried on to the next step without further purification.
8. When sulfonamides were compared with their amide analogs, solubility
improved; however, a 2- to 10-fold loss of potency was consistently observed
for GlyT1 uptake inhibition (data not shown).
1. (a) Conn, P. J.; Tamminga, C.; Schoepp, D. D.; Lindsley, C. Mol. Interventions 2008,
8, 99; (b) Vandenberg, R. J.; Aubrey, K. R. Expert Opin. Ther. Targets 2001, 5, 507.
2. (a) Roberts, B. M.; Shaffer, C. L.; Seymour, P. A.; Schmidt, C. J.; Williams, G. V.;
Castner, S. A. Neuroreport 2010, 21, 390; (b) Javitt, D. C. Curr. Opin. Drug Discov.
Devel. 2009, 12, 468; (c) Gozzi, A.; Herdon, H.; Schwarz, A.; Bertani, S.; Crestan,
V.; Turrini, G.; Bifone, A. Psychopharmacology 2008, 201, 273; (d) Depoortere, R.;
Dargazanli, G.; Estenne-Bouhtou, G.; Coste, A.; Lanneau, C.; Desvignes, C.;
Poncelet, M.; Heaulme, M.; Santucci, V.; Decobert, M.; Cudennec, A.; Voltz, C.;
Boulay, D.; Terranova, J. P.; Stemmelin, J.; Roger, P.; Marabout, B.; Sevrin, M.;
Vige, X.; Biton, B.; Steinberg, R.; Francon, D.; Alonso, R.; Avenet, P.; Oury-Donat,
F.; Perrault, G.; Griebel, G.; George, P.; Soubrie, P.; Scatton, B.
Neuropsychopharmacology 2005, 30, 1963.
3. (a) Wolkenberg, S. E.; Sur, C. Curr. Top. Med. Chem. 2010, 10, 170; (b) Glycine
Transporters and Their Inhibitors; Gilfillan, R.; Kerr, J.; Walker, G.; Wishart, G. in
Topics in Medicinal Chemistry: Glycine Transporters and Their Inhibitors,
Springer-Verlag: Berlin, 2009; 4, 223; (c) Bridges, T. M.; Williams, R.; Lindsley, C.
W. Curr. Opin. Mol. Ther. 2008, 10, 591; (d) Hashimoto, K. Cent. Nerv. Syst. Agents
Med. Chem. 2007, 7, 177; (e) Harsing, L. G., Jr.; Juranyi, Z.; Gacsalyi, I.;
Tapolcsanyi, P.; Czompa, A.; Matyus, P. Curr. Med. Chem. 2006, 13, 1017; (f)
Lindsley, C. W.; Wolkenberg, S. E.; Kinney, G. G. Curr. Top. Med. Chem. 2006, 6,
1883; (g) Lowe, J. A., III. Expert Opin. Ther. Pat. 2005, 15, 1657; (h) Slassi, A.; Egle,
I. Expert Opin. Ther. Pat. 2004, 14, 201.
4. GlyT1 uptake inhibition assay: The GlyT1 SPA assay measures [3H] glycine uptake
by CHO cells expressing human recombinant Glycine transporter 1 (GlyT1).
9. Alelyunas, Y. W.; Liu, R.; Pelosi-Kilby, L.; Shen, C. Eur. J. Pharm. Sci. 2009, 37, 172.