1,3-Diaminopropan-2-ol Sulfonamides as potent and selective inhibitors of the glycine transporter type 1

Article abstract of DOI:10.1016/j.bmcl.2006.12.063

High throughput screening led to the discovery of a novel series of 1,3-diaminopropan-2-ol sulfonamides as selective GlyT-1 inhibitors. Structure-activity relationships of this novel series and optimisation of the initial hit that led to the identificatio

Full text of DOI:10.1016/j.bmcl.2006.12.063

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1741–1745
1,3-Diaminopropan-2-ol Sulfonamides as potent and selective
inhibitors of the glycine transporter type 1
Shahzad S. Rahman,a,* Steven Coulton,b Hugh J. Herdon,b
Graham F. Joiner,a Jian Jinc and Roderick A. Porterb
aDiscovery Medicinal Chemistry, GlaxoSmithKline, Harlow, Essex CM19 5AW, UK
bPsychiatry CEDD, GlaxoSmithKline, Harlow, Essex CM19 5AW, UK
cDiscovery Medicinal Chemistry, GlaxoSmithKline, Collegeville, PA 19426, USA
Received 17 November 2006; revised 15 December 2006; accepted 16 December 2006
Available online 22 December 2006
Abstract—High throughput screening led to the discovery of a novel series of 1,3-diaminopropan-2-ol sulfonamides as selective
GlyT-1 inhibitors. Structure–activity relationships of this novel series and optimisation of the initial hit that led to the identification
of (2), a potent and selective GlyT-1 inhibitor, are also presented.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Molecular cloning has revealed the existence in mamma-
lian brains of two classes of glycine transporters, termed
GlyT-1 and GlyT-2. GlyT-1 is found predominantly in
the forebrain and its distribution corresponds to that
of glutaminergic pathways and NMDA receptors.1
Molecular cloning has further revealed the existence of
three variants of GlyT-1, termed GlyT-la, GlyT-1b
and GlyT-1c,2 each of which displays a unique distribu-
tion in the brain and peripheral tissues. GlyT-2, in con-
trast, is found predominantly in the brain stem and
spinal cord, and its distribution corresponds closely to
that of strychnine-sensitive glycine receptors.3 Another
distinguishing feature of glycine transport mediated by
GlyT-2 is that it is not inhibited by sarcosine, as is the
case for glycine transport mediated by GlyT-1. These
data are consistent with the view that, by regulating
the synaptic levels of glycine, GlyT-1 and GlyT-2 selec-
tively influence the activity of NMDA receptors and
strychnine-sensitive glycine receptors, respectively.
Thus, agents that inhibit GlyT-1 and thereby increase
glycine activation of NMDA receptors may be useful
as novel anti-psychotics and anti-dementia agents, and
to treat other diseases in which cognitive processes are
impaired, such as attention deficit disorders and organic
brain syndromes.
High throughput screening of our in-house compound
collection (measuring inhibition of [3H]glycine uptake
in HEK293 cells stably transfected with hGlyT-1c)6,7
identified the 2,4-dimethylpyrrolidine derivative (1)
(Fig. 1), a mixture of four diastereoisomers.8 Compound
(1) has good inhibitory activity against hGlyT-1c (pIC50
6.1) and good selectivity over hGlyT-2 (pIC50 < 5.0). In
this communication, we describe the optimisation of the
1,3-diaminopropan-2-ol derivative (1) leading to the
identification of the substituted piperidine (2). Com-
pound (2) is a potent and selective GlyT-1c inhibitor
and a valuable tool compound for investigating the
scientific rationale for the use of GlyT-1 inhibitors in
NMDA receptors are critically involved in memory and
learning4 and, furthermore, decreased function of
NMDA-mediated neurotransmission appears to under-
lie, or contribute to, the symptoms of schizophrenia.5
O
O
O
O
(R)
OH
(S)
S
S
N
N
H
N
N
H
Me
OH
Me
(1)
(2)
Keywords: GlyT-1; Schizophrenia; NMDA, Transporter; 1,3-Diami-
nopropan-2-ol.
pIC50 6.1 (hGlyT-1c),
<5.0 (hGlyT-2)
pIC50 7.0 (hGlyT-1c),
<5.0 (hGlyT-2)
*
Corresponding author.
shahzad.rahman-1@gsk.com
Tel.:
+44
1279
627781; e-mail:
Figure 1. Structures of HTS hit (1) and optimised analogue (2).
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.063

1742
S. S. Rahman et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1741–1745
the treatment of schizophrenia. Structure–activity rela-
tionships of this novel series of Gly-1 inhibitors are also
presented.
cially available (S)- and (R)-glycidyl nosylates, respec-
tively. In both cases activity resided mainly with the
(2R)-isomer (Table 1). In subsequent optimisation only
the preferred (2R)-hydroxyl isomer was used.
1,3-Diaminopropan-2-ols (6) were prepared by the pre-
viously reported procedure9 (Scheme 1). Both (S) and
(R) glycidyl nosylates (3a) and (3b) were reacted with
substituted secondary amines10 affording chiral epoxides
(4a and 4b). Epoxide ring-opening was achieved using
lithium azide (20% aq soln) to give the hydroxyazides
(5a and 5b), which were hydrogenated at atmospheric
pressure in the presence of Pearlmans catalyst to afford
the 1,3-diaminopropan-2-ols (6a and 6b). The sulfona-
mides (7) were prepared by reaction of (6) with sulfonyl
chlorides in the presence of Amberlyst IRA 93 resin.
We next turned our attention to substitution in the pyr-
rolidine ring (Table 2). All compounds were prepared as
a mixture of isomers and separated by HPLC.11 Substi-
tution at the ortho position of the heterocycle is required
for good affinity. Compounds with isopropyl (9) or tert
butyl (10) groups are optimal with potency decreasing
for the cyclopropyl (11), ethyl (12), methyl (8), or the
unsubstituted (13) analogues. Compounds substituted
with 2-(methyl)-propyl (14) or cyclohexyl (15) substitu-
tions are detrimental suggesting that larger groups are
not preferred.
The optimal stereochemistry around the hydroxyl group
was first investigated by preparing the two possible 2-hy-
droxy isomers of 2,4-dimethylpyrrolidine (1a and 1b)
and 2-methylpyrrolidine (8a and 8b) from the commer-
We then investigated ring size by replacing the pyrroli-
dine ring with a piperidine (Table 3). Whilst most com-
parable analogues were equipotent, trends are not
mirrored. Ethyl substitution (2) was identified as the
best group for activity in the piperidine series. The 2-al-
kyl diastereoisomers of pyrrolidine (Table 2) and piper-
idine analogues (2), (8–12), (14), (15) and (21) were
separated by HPLC.11 Both isomers have affinity for
GlyT-1c but the (2S)-alkyl-(2R)-hydroxy stereoisomer
was shown to be the favoured12 isomer. Small alkyl sub-
stitution at the 3 position of the piperidine does not offer
any advantage. The most potent isomers of the dimethyl
and monomethyl pyrrolidine derivatives (1) and (8),
respectively, are equipotent. Alicyclic analogues (16)
and (17) are active but potency decreases a little with re-
spect to similar cyclic analogues.
O2N
O
S O
O
(b)
(a)
O
O
R1
(4a, b)
(3a) (S)-nosylate
(3b) (R)-nosylate
NH2
R1
(c)
R1
N3
OH
OH
(6a, 6b)
(5a, 5b)
We then examined the SAR of the aromatic sulfon-
amide fragment. A diverse set of 50 monocyclic, bicy-
clic, aryl and heteroaryl sulfonamides (7) (keeping the
2,4-dimethyl pyrrolidine group as constant) were pre-
pared by parallel synthesis.7,13 SAR around this posi-
tion (Table 4) is tight with 1-naphthyl substituent,
present in the original HTS hit (1), being optimal.
2-Naphthyl analogue (24) and isosteric replacements
of naphthyl ring such as dichlorophenyl (25) are active
O
R2
(d)
R1
N S
H
O
OH
(7a) (2S-)
(7b) (2R-)
Scheme 1. Reagents and conditions: (a) Amine R1H,10 KH/THF, rt,
argon; (b) 20% aq LiN3, THF, 75 ꢁC; (c) H2/Pd(OH)2, MeOH,
atmospheric pressure, rt; (d) R2-SO2-Cl, DCM, Amberlyst IRA 93.
Table 1. Optimal stereochemistry of C-2 hydroxyl group
O
3
1
2
S
R
N
O
H
OH
a
)
Compound
R
Isomer
GlyT-1c (pIC50
5.1
1a
2S
N
N
1b
2R
6.1
8a
8b
2S
4.9
6.0
2R
a Mean of at least three determinations with standard deviation of < 0.3.

S. S. Rahman et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1741–1745
Table 2. Inhibitory activity of 1-pyrrolidine analogues (9–17)
1743
O
S
R1
N
O
H
OH
a
Compound
R1
GlyT-1c (pIC50
)
Isomer A
retention time (run time in min)b
Isomer B retention time
(run time in min)b
Isomer
Mixture
A + B
—
A
B
iPr
9
6.0
6.7
9.71 (16)
12.78 (16)
*
*
N
N
tBu
10
5.1
6.7
—
24.59 (35)
27.32 (35)
11
12
5.8
5.9
6.3
5.8
—
—
20.76 (32)
19.53 (35)
26.05 (32)
29.66 (35)
*
N
N
Et
*
*
13
14
—
—
5.5
—
N
5.2
5.6
12.70 (18)
19.75 (30)
14.25 (18)
24.44 (30)
*
*
N
15
5.2
5.2
—
N
16
17
—
—
—
—
5.9
6.0
*
*
N
N
Pr
tBu
Pr
a Mean of at least three determinations with standard deviation of < 0.3.
b HPLC retention time using the analytical protocol described in the notes.11
but less potent. Similarly, substituted aryl or heteroaryl
sulfonamide analogues with ortho substitution, such as
compounds (26), (27) and (28), show reduced inhibitory
activity.
bioavailability after subcutaneous administration
(5 mg/kg in rats) was very high (95%), with a moderate
half-life (t1/2 = 0.9 h) and brain to plasma AUC
ratio of 0.4:1 (Cmax 564 ng/ml, AUC0–inf 1731 h ng/
ml, brain level 1 h postdose 564 ng/ml). With this
Optimisation of the HTS hit (1) culminated in the
identification of the (2S)-ethylpiperidine derivative
(2) GlyT-1c (pIC50 7.0). This compound demonstrated
high intrinsic clearance in both rat and human micro-
somes (CLi 8.0 mg/ml/g for both species) and
significant affinity for CYP2D6 (1.2 lM). After oral
administration of 5 mg/kg (free base) in rats, (2)
showed poor oral bioavailability (4%).14 However,
PK profile, diaminopropan-2-ol (2) represents a
valuable tool compound for validating the rationale
for the use of GlyT-1 inhibitors in the treatment of
schizophrenia.
The results of these studies together with the optimisa-
tion of PK parameters of this novel series of GlyT-1
inhibitors will be the subject of a later paper.

1744
S. S. Rahman et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1741–1745
Table 3. Inhibitory activity of 1-piperidine analogues (18–23)
O
S
R1
N
O
H
OH
Isomer A retention time (run time in min)c Isomer B retention time (run time in min)c
a
Compound R1
GlyT-1c (pIC50)
Isomer
Mixture
A + B
5.5
A
B
*
18
—
—
N
N
*
19
—
—
6.5
—
Et
20 and 2
5.1 7.0b
27.04 (40)
20.69 (40)
30.63 (40)
22.71 (40)
*
N
iPr
*
21
5.4 6.2
—
N
N
*
22
—
—
—
—
5.4
6.0
*
23
N
a Mean of at least three determinations with standard deviation of < 0.3.
b This isomer was shown to correspond to the N-{(2R)-3-[(2S)-2-ethyl-1-piperidinyl]-2-hydroxypropyl}-1-naphthalenesulfonamide diastereoisomer.
c HPLC retention time using the analytical protocol described in the notes.11
Table 4. Modifications of the arylsulfonamide group
R
N
N
H
Me
OH
Me
b
Compounda
R
GlyT-1c (pIC50
)
*
6.1
1b
O
S O
*
*
O
24
5.4
S
O
Cl
Cl
O
S
Cl
25
26
5.6
5.5
O
O
S
*
S
O
Cl

S. S. Rahman et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1741–1745
1745
Table 4 (continued)
Compounda
27
b
R
GlyT-1c (pIC50)
5.5
*
O
S
O
Cl
28
5.5
*
O
S
O
Br
a Compounds are diastereoisomeric mixture of 2,4-dimethylpyrrolidines.
b Mean of at least three determinations with standard deviation of < 0.3.
Lipkowitz, K. B.; Lee, M.; Larson, B. R. J. Org. Chem.
1974, 39, 1963; (c) Bortnick, N.; Luskin, L. S.; Hurwitz,
M. D.; Craig, W. E.; Exner, L. J.; Mirza, J. J. Am. Chem.
Soc. 1956, 78, 4039.
Acknowledgments
We thank Matthew Sanders for providing analytical
support with separating the diastereoisomers and Peter
Eddershaw for providing DMPK data.
11. Diastereoisomers were separated by HPLC using the
following protocol: samples were dissolved in EtOH and
run isocratically on Rainin preparative HPLC equipment,
eluting with n-hexane:HiPerSolv EtOH:AnalaR TEA
(90:10:0.25 v/v pre-mixed) mobile phase, on Chiralpak
AS stationary phase (250 mm · 20 mm id). Flow rate 20
mL minꢀ1, ambient temperature, detection was by UV
absorbance at 290 nm.
References and notes
1. Smith, K. E.; Borden, L. A.; Hartig, P. R.; Branchek, T.;
Weinshank, R. L. Neuron 1992, 8, 927.
12. The absolute stereochemistry of (N-{(2R)-3-[(2S)-2-ethyl-
1-piperidinyl]-2-hydroxypropyl}-1-naphthalenesulfona-
mide) (2) was determined by unambiguous synthesis from
(S)-2-ethylpiperidine (Cymerman, J. C.; Pinder A. R.
J. Org. Chem., 1971, 36, 3648.) and (S)-glycidyl nosylate,
using afore-mentioned literature procedures. Optical rota-
2. Kim, K. M.; Kingsmore, S. F.; Han, H.; Yang-Feng, T.
L.; Godinot, N.; Seldin, M. F.; Caron, M. G.; Giros, B.
Mol. Pharmacol. 1994, 45, 608.
3. (a) Liu, Q. R.; Lopez-Corcuera, B.; Mandiyan, S.; Nelson,
H.; Nelson, N. J. Biol. Chem. 1993, 268, 22802; (b) Jursky,
F.; Nelson, N. J. Neurochem. 1995, 64, 1026.
4. Rison, R. A.; Staunton, P. K. Neurosci. Biobehav. Rev.
1995, 19, 533.
5. Olney, J. W.; Farber, N. B. Arch. Gen. Psychiatry 1995,
52, 998.
6. Herdon, H. J.; Godfrey, F. M.; Brown, A. M.; Coulton, S.;
Evans, J. R.; Cairns, W. J. Neuropharmacology 2001, 41, 88.
7. Coulton, S.;Hadley, M. S.;Herdon, H. J.;Jin, J.;Joiner, G. J.;
Porter, R. A.; Rahman, S. S. WO Patent 03/055478 A1, 2003.
8. Pyrrolidine derivative (1) is an isomeric mixture of four
diastereoisomers–activity resides predominantly with one
isomer as confirmed by testing the four diastereoisomers
separated by HPLC.
20
tion: (2) ½aꢁD +25.70 (c 0.3% in MeOH) and (N-{(2R)-3-
20
[(2R)-2-ethyl-1-piperidinyl]-2-hydroxypropyl}-1-naphtha-
lenesulfonamide)(20) ½aꢁD ꢀ9.5ꢁ (c 0.3% in MeOH).
13. General procedure for array synthesis: A solution of (S)-
1-amino-3-(2,4-dimethylpyrrolidin-1-yl)propan-2-ol (6a,
R1 = 2,4-dimethyl pyrrolidin-1-yl) (800 mg; 4.65 mmol)
in anhydrous dichloromethane (38 ml) was prepared,
and aliquots (0.5 ml; 0.06 mmol) were added to solutions
of 50 sulfonyl chlorides (0.075 mmol each) dissolved in
anhydrous dichloromethane (1 ml each). Triethylamine
(0.021 ml; 0.15 mmol) was added to each solution, which
was shaken overnight at ambient temperature. The reac-
tion mixtures were shaken with polymer-bound tris amine
(ꢂ45 mg) for 3 h, then filtered and purified by preparative
HPLC to afford the sulfonamides (7) as the trifluoroac-
etate salts.
9. Dhanoa, D. S.; Parsons, W. H.; Greenlee, W. J.; Patchett,
A. A. Tetrahedron Lett. 1992, 33, 1725.
10. The non-commercially available alkyl pyrrolidines and
piperidines required for the synthesis of compounds (11),
(12), (13), (17) and (21) were prepared using the following
literature methods: (a) De Jong, M.; Wibaut, J. P. Recl.
Trav. Chim. Pays-Bas 1930, 49, 237; (b) Mundy, B. P.;
14. PK data for (2) after i.v. administration (1 mg/kg; rat),
CLp 18 ml/min/kg; t1/2 = 1.0 h, plasma AUC0–inf 952 h ng/
ml, brain AUC0–inf 587 h ng/ml.