Identification of novel glycine sulfonamide antagonists for the EP1 receptor

Article abstract of DOI:10.1016/j.bmcl.2006.12.060

A high-throughput screen targeting the EP₁ receptor identified non-acidic glycine sulfonamide derivative 2a with a pK_i of 6.2. Analogue synthesis allowed a thorough investigation of the structure-activity relationship (SAR) and led to a 100-fold increase in recombinant potency.

Full text of DOI:10.1016/j.bmcl.2006.12.060

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1750–1754
Identification of novel glycine sulfonamide antagonists
for the EP1 receptor
b,
a
b
Stephen C. McKeown,a,* Adrian Hall, Richard Blunt, Susan H. Brown,
*
Iain P. Chessell,b Anita Chowdhury,b Gerard M. P. Giblin,b Mark P. Healy,b
Matthew R. Johnson,b Olivier Lorthioir,a Anton D. Michel,b Alan Naylor,b
Xiao Lewell,a Shilina Roman,b Stephen P. Watson,a
Wendy J. Winchesterb and Richard J. Wilsona
aMolecular Discovery Research, GlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow, Essex CM19 5AW, UK
bNeurology and Gastrointestinal Centre of Excellence for Drug Discovery GlaxoSmithKline, New Frontiers Science Park,
Third Avenue, Harlow, Essex CM19 5AW, UK
Received 26 October 2006; revised 15 December 2006; accepted 16 December 2006
Available online 22 December 2006
Abstract—A high-throughput screen targeting the EP1 receptor identified non-acidic glycine sulfonamide derivative 2a with a pKi of
6.2. Analogue synthesis allowed a thorough investigation of the structure–activity relationship (SAR) and led to a 100-fold increase
in recombinant potency.
Ó 2006 Elsevier Ltd. All rights reserved.
Prostaglandin E2 (PGE2) is one of a number of prosta-
glandins generated through the metabolism of arachi-
donic acid, and acts locally both in the periphery and
CNS. PGE2 mediates a multitude of pharmacological
actions, acting predominantly through four identified
G-protein coupled receptors (GPCRs, EP1–4).1 The
EP1 receptor has been of interest to the pharmaceutical
industry due to the preclinical effects seen with small
molecule antagonists in conditions such as inflammatory
and neuropathic pain.2–5
There have been several reports of carboxylic acid-con-
taining EP1 receptor antagonists.8–11 However, there
have been relatively few reports of non-acidic antago-
nists, such as those from Searle12 (1a, SC-51322) and
Merck Frosst6 (1b), Figure 1.
A high-throughput screen of the GlaxoSmithKline com-
pound collection, using a FLIPR based assay,13 identi-
fied the glycine sulfonamide derivative 2a with a
functional pKi of 6.2, and at least 10-fold selectivity over
the EP3 receptor, Figure 2.
In this paper, we describe the identification of a novel
series of non-acidic glycine sulfonamides as potent EP1
antagonists. We were particularly interested in these
hit molecules because they do not contain the common
acidic motif displayed in many EP1 antagonists which,
if present, would generally limit exposure to the CNS.6
We wanted to explore molecules with CNS penetration
since there is some evidence that centrally expressed
EP1 receptors are involved in certain types of pain.7
Compound 2a represented a novel EP1 ligand that was
chemically tractable, had low molecular weight and
Cl
S
H
O
Cl
N
H
O
N
N
O
N
S
O
H
O
O
Keywords: EP1 antagonist; Pain; Pyrazole; Prostanoid.
*
1a12, SC-51322; EP1: pA2 8.1
1b6, EP1: pKi 8.4
Corresponding authors. Tel.: +44 1279 627558; fax: +44 1279 87 5414
(S.C.M.); tel.: +44 1279 623464 (A.H.); e-mail addresses:
Steve.C.McKeown@gsk.com; Adrian.2.Hall@gsk.com
Figure 1. Known non-acidic EP1 antagonists.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.060

S. C. McKeown et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1750–1754
1751
O
S
Fixing the newly discovered 3,4-dimethoxyphenethyl
amide moiety, we next sought to explore the SAR
around the aniline moiety (ring A), Table 2. These com-
pounds were synthesized according to Scheme 1 (using
route B).
O
H
B
N
N
EP1 functional pKi 6.2
EP3 functional pKi <5
C
O
Cl
A
2a
Figure 2. HTS hit compound 2a.
Removal of the 2-methyl substituent present in 2b re-
duced activity by approximately 16-fold (compound 3a
vs 2b). Substitution of the aniline moiety at the 4-posi-
tion generally led to a marked decrease in activity;
electron-withdrawing groups such as F, Cl and Br were
not well tolerated, compounds 3b–d; and similarly, elec-
tron-donating groups such as MeO (3f) and Me (3e)
showed no activity. Further SAR of the aniline moiety
indicated that the 2- and 3-positions were amenable to
substitution. For example, the 3-Me (3g) and 3-Cl (3h)
derivatives displayed an almost 10-fold improvement
in activity relative to the unsubstituted analogue 3a.
The 3-CF3 (3j) and Br (3i) groups showed the highest
activity of the 3-substituents, with functional pKi values
possessed no carboxylic acid or similarly acidic group.
We explored the structure–activity relationship (SAR)
by varying four regions of the hit molecule 2a.
Screening of phenethyl amide analogues (a selection is
shown in Table 1) provided initial SAR around ring
C. These were synthesized according to Scheme 1 (route
A). It was found that the 3,4-dimethoxyphenethyl amide
was one of the most preferred groups, compound 2b,
with a functional pKi of 7.4. The most active analogue
of the phenethyl amides was compound 2g with a func-
tional pKi of 8.4. However, no further derivatives of this
compound were prepared due to its higher molecular
weight and polar surface area. It appears that substitu-
tion on this ring is required since potency of the unsub-
stituted ring, 2h, is reduced to pKi 5.8.
Table 2. SAR for substitution (R1) of the aniline moiety, 3a–3l
B
H
Table 1. SAR of the phenethylamide derivatives, 2a–2h
O
S
N
O
O
N
O
2
C
O
R1
A
B
3
3
4
H
2
O
S
N
3
4
R1
EP1 functional pKi13
N
O
Compound
C
O
A
R3
3a
2b
3b
3c
3d
3e
3f
H
6.23 0.01
7.41 0.23
<5.8
2-Me
4-F
4-Cl
2
Compound
R3
4-Cl
EP1 functional pKi13
5.89 0.23
<5.42
<5.95
2a
2b
2c
2d
2e
2f
6.18 0.30
7.41 0.23
6.77 0.18
6.69 0.05
7.05 0.07
7.18 0.01
8.43 0.06
5.87 0.22
7.9 0.714
4-Br
3,4-DiOMe
4-OMe
3-OH
4-Me
4-OMe
3-Me
3-Cl
<5.8
3g
3h
3i
7.21 0.04
7.21 0.25
7.46 0.32
7.51 0.01
7.87 0.28
7.89 0.19
4-NHSO2Me
2-F–3,4-DiOMe
3-SO2NHMe,4-OH
H
3-Br
2g
2h
1b
3j
3-CF3
2-Et
2,3-diMe
3k
3l
R2
R2
R2
R3
H
N
O
O
O
S
H2N
S
n
OH
S
N
n
N
NH
NH2
Route A
b,c
O
O
O
R3
10
O
O
n=1,2
a
R1
R1
R1
R1
d
2,3,4,5
9
8
7
Cl
H
N
H
O
N
Br
e
Br
or
O
O
O
Route B
1117
1217
Scheme 1. Reagents and conditions: (a) ArSO2Cl, CH2Cl2, pyridine, 16 h; (b) 2 N NaOH, THF, BrCH2CO2Et, 16 h; (c) LiOH, THF, H2O, 2 h; (d)
10, EDCI, HOAT, DMF, CH2Cl2, 16 h; (e) 11 or 12, NaOH, THF, reflux, 16 h.

1752
S. C. McKeown et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1750–1754
Table 3. SAR for substitution (R2) of the aryl sulfone moiety, 4a–4l
of 7.5, indicating the preference of a bulky substituent at
this 3-position.
R2
B
Further exploration of the 2-position revealed that the
methyl group could be replaced by an ethyl group
(3k), leading to an increase in activity of approximately
2-fold. Combining the 2- and 3-substitution resulted in
excellent activity as displayed by the 2,3-dimethyl deriv-
ative 3l with a functional pKi of 7.9, Table 2.
H
O
N
O
O
S
N
O
C
R1
O
A
4
Compound
R1
R2
EP1 functional pKi13
Next we investigated the SAR around the aryl sulfon-
amide, ring B. These compounds were synthesized using
solid-phase methodologies described in Scheme 2. A
selection of the results is shown in Table 3.
3k
2b
4a
4b
4c
4d
4e
4f
2-Et
4-Me
4-Me
H
7.87 0.28
7.41 0.23
6.13a
2-Me
2-Me
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
2-Et
4-CF3
4-Br
7.72 0.10
7.54 0.01
7.50 0.16
7.16 0.19
6.78 0.14
6.77 0.02
6.62 0.04
6.15 0.27
5.78 0.06
5.70a
4-CN
4-OCF3
2-Me
4-OMe
3-CN
3-OCF3
3-Me
2-Cl
Substitution at the 4-position appears to be preferred
with substitution at the 2- or 3-positions resulting in a
loss of potency. Introduction of a methoxy group at
the 4-position (4g) leads to a slight reduction in potency
(pKi 6.8). Compound 4a with hydrogen in place of meth-
yl is less potent with a pKi of 6.1. Generally, no potency
enhancement was observed over 3k with any of these
analogues.
4g
4h
4i
4j
4k
4l
2-CN
<5.5
a Single experiment.
Further profiling of compound 3k, Figure 3, indicated
13
excellent selectivity for the EP1 receptor over EP3
(functional pKi < 5). When screened in a [3H]PGE2
binding assay in EP1-transfected CHO cell mem-
branes,15 3k showed a pIC50 of 8.3. Schild analysis in
the reporter gene assay16 revealed 3k to have a pA2 of
8.3 with a slope of 0.9, indicating it to be a competitive
antagonist.
O
O
H
N
S
O
O
EP1 binding pIC50 8.3
EP1 functional pKi 7.7
EP3 functional pKi < 5
N
O
3k
Figure 3. Profile of lead compound 3k.
Having completed the initial round of optimization, we
next sought to expand the SAR by returning to further
optimize the phenethyl amide moiety. Interestingly, we
found that it was possible to truncate the phenethyl
amide to the corresponding benzylamide with less than
10-fold decrease in activity, compound 5a, Table 4.
SAR was then obtained on a range of benzylamides.
The compounds in Table 4 were synthesized according
to Scheme 1 (route A).
Single substitution at the 3-position led to a decrease in
potency, for example, 5d and 5e. However, single meth-
oxy substitution at the 4-position proved to be more suc-
cessful. Compound 5f showed good activity in both
functional and binding assays. Replacement of the 4-
methoxy group by Cl led to a further gain in activity,
compound 5g. The 4-Br analogue 5h was found to be
MeO
MeO
NH
2
OMe
OMe
OMe
OMe
O
14
13
b
a
N
N
H
H
15
Br
c
16
OMe
O
O
O
=
NH
2
H
1
OMe
O
R
Polystyrene Resin
7
OMe
OMe
OMe
OMe
OMe
HN
O
N
OMe
R1
N
S
H
e
d
N
O
O
O
O
N
S
N
R1
Ar
O
18
3,4
R1
Ar
O
17
Scheme 2. Reagents and conditions: (a) 13, NaBH(OAc)3, AcOH, CH2Cl2, NMP, 16 h; (b) bromoacetic acid, DIC, DMF, 16 h; (c) 7, NMP, 120 °C,
20 min, microwave; (d) ArSO2Cl, DIPEA, CH2Cl2,16 h; (e) TFA/H2O, 2 h.

S. C. McKeown et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1750–1754
Table 4. SAR of substituted benzylamide compounds 5a–5k Table 6. SAR of substituted benzylamide compounds 6a–6i
1753
3
B
B
Cl
4
5
2
R3
H
C
6
H
C
O
O
S
N
S
N
N
N
O
O
6
O
O
Ar
A
Compound Ar
EP1 functional EP1 binding
15
pIC50
5
pKi13
Compound
R3
EP1 functional
pKi13
EP1 binding
15
pIC50
5g
6a
6b
6c
6d
6e
6f
2-Ethylphenyl
3-Pyridyl
2-Pyridyl
7.94 0.22
<5.84
<5.84
8.0 0.3
<6
<6
5a
5b
5c
5d
5e
5f
3,4-DiOMe
H
6.55 0.05
6.65 0.36
6.86 0.47
6.37 0.02
5.85 0.01
7.45 0.24
7.94 0.22
7.64 0.67
6.68 0.03
6.87a
7.6 0.1
7.2 0.2
7.1 0.1
n/t
2-Me-3-pyridyl
3-Me-2-pyridyl
6-Me-3-pyridyl
6-Me-2-pyridyl
5-Me-2-pyridyl
6.17a
6.7 0.3
6.1 0.1
<6
<5.84
<5.84
3-OMe
3-OCF3
3-Cl
<5.84
<5.84
<6
<6
6.6 0.1
7.6 0.1
8.0 0.3
6g
6h
6i
4-OMe
4-Cl
4-Br
1-Me-5-pyrazolyl 6.23 0.06
1-Et-5-pyrazolyl 7.14 0.07
6.4 0.1
6.8 0.1
5g
5h
5i
a Single experiment.
4-CF3
2-F
2-MeO
5j
5k
5.97a
ment of the phenyl ring by an unsubstituted pyridine
abolished activity, 6a and 6b. However, when a 2-methyl
group was added to the 3-pyridyl isomer, moderate
activity was observed, 6c. None of the other pyridine
analogues synthesized showed appreciable activity, 6d–
g, even when incorporating the preferred 2- or 3-methyl
substituents observed in Table 2. The most active com-
pound from this effort was the 1-ethyl-5-pyrazolyl ana-
logue 6i which displayed a functional pKi of 7.1 and a
binding pIC50 of 6.8. The corresponding Me analogue
6h was slightly less active. However, the intrinsic clear-
ance for 6i remained high (30 and 42 mL/min/g liver in
human and rat microsomes, respectively). Hence, it
was generally observed that replacement of the A ring
with heteroaryls led to less potent derivatives.
a Single experiment.
slightly less active, and the 4-CF3 analogue 5i was signif-
icantly less active. Interestingly, the SAR around the
phenethylamides (Table 1) and benzylamides (Table 4)
appear to differ. For example, 4-Cl appears optimal
and more potent than 3,4-dimethoxy substituent in the
benzylamide series, whereas the reverse potency order
is observed in the phenethylamide series.
A profile of 5g is shown in Table 5 together with its
selectivity against other prostanoid receptors. Com-
pound 5g is approximately 100-fold selective over EP3
and is poorly active against the TP and FP receptors.
It shows inhibition of CYP450 isoforms tested with
the highest inhibition being against 3A4 (pIC50 6.6).
The high microsomal intrinsic clearance (CLi) observed
for 5g was typical of potent compounds described in Ta-
bles 1–4 (>50 mL/min/g liver in both rat and human
microsomes).
The molecules described in this paper were either syn-
thesized using solution chemistry (Scheme 1) or solid-
phase methodologies (Scheme 2) or alternatively were
commercially available.
Scheme 1 shows the reaction of a substituted aniline 7
with the appropriate aryl sulfonyl chloride generating
sulfonamide 8 which was subsequently alkylated with
ethyl bromoacetate followed by basic hydrolysis to give
the carboxylic acid 9. Amide coupling with benzyl or
phenethyl amine derivatives (10) under standard condi-
tions furnished the desired amides (route A). Alterna-
tively, some final products were prepared directly from
sulfonamide 8 (via route B) by alkylation with bromide
11 or 12.17
In an attempt to improve the poor metabolic stability of
these molecules, we replaced the 2-ethylphenyl moiety
with heterocycles in order to lower the logP, Table 6.
These molecules were synthesized using similar condi-
tions to those described in Scheme 1, route B. Replace-
Table 5. Selectivity, in vitro metabolism and CYP450 profile of 5g
Parameter
Value
EP1 FLIPR13 pKi
EP1 binding15 pKi
EP3 FLIPR13 pKi
TP pIC50
7.94 0.22
8.0 0.3
Scheme 2 shows the solid phase synthetic route to com-
pounds 3 and 4. Resin-bound amine 15 was formed via
reductive alkylation of the phenethylamine 14 with alde-
hyde resin 13 using sodium triacetoxy borohydride.
Bromoacetic acid was then directly coupled to the amines
to give the alkylating agent 16, which was reacted with the
anilines (7) to furnish the resin-bound secondary anilines
17. These were then reacted with the appropriate aryl sul-
fonyl chloride and subsequently cleaved from the resin to
yield the final products.
6.1a
<5.0
FP pIC50
CYP450b, pIC50
5.38 0.05
4.5 at 1A2, 5.2 at 2C9, 5.7 at
2C19, 4.0 at 2D6, 6.6 at 3A4
Microsomal CLi (mL/min/g liver) >50 rat; >50 human
Molecular weight 456
a Tested three times, inactive (<5.0) twice, active (6.1) once.
b In vitro CYP450 assay results using Gentest protocol.

1754
S. C. McKeown et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1750–1754
10. Hall, A.; Atkinson, S.; Brown, S. H.; Chessell, I. P.;
Chowdhury, A.; Clayton, N. M.; Coleman, T.; Giblin, G.
M. P.; Gleave, R. J.; Hammond, B.; Healy, M. P.;
Johnson, M. J.; Michel, A. D.; Naylor, A.; Novelli, R.;
Spalding, D. J.; Tang, S. P. Bioorg. Med. Chem. Lett.
2006, 16, 3657.
11. McKeown, S. C.; Hall, A.; Giblin, G. M. P.; Lorthioir, O.;
Blunt, R.; Lewell, X. Q.; Wilson, R. J.; Brown, S. H.;
Chowdhury, A.; Coleman, T.; Watson, S. P.; Chessell, I.
P.; Pipe, A.; Clayton, N.; Goldsmith, P. Bioorg. Med.
Chem. Lett. 2006, 16, 4767.
Compounds synthesized by either Scheme 1 or Scheme 2
were purified using HPLC mass-directed purification
(fraction collection triggered by mass ion detection).
In summary, we have described the synthesis and SAR
of glycine sulfonamides as novel EP1 receptor antago-
nists. Compound 3k has been shown to be a competitive
antagonist at EP1 and chemical optimization around
this template resulted in 5g having an EP1 functional
pKi of 7.8. Unfortunately, all of the compounds de-
scribed here exhibited poor in vitro metabolic stability
in rat and human microsomes. Further optimization of
this template will be described in due course.
12. Hallinan, E. A.; Stapelfeld, A.; Savage, M. A.; Reichman,
M. Bioorg. Med. Chem. Lett. 1994, 4, 509.
13. The pKi of compounds were calculated from IC50’s
measured using stable either EP1 or EP3 transfected
CHO (Chinese hamster ovary)-K1 whole cells and by
measuring inhibition of intracellular Ca2+ mobilisation in
response to PGE2. Calcium-sensitive dye (Fluo-3) was
Acknowledgment
assessed using
(FLIPR). Values are means of at least two experiments
unless specified.
a Fluorimetric Image Plate Reader
We thank Dave Langley for coordinating the purchas-
ing of compounds.
14. These data agree well with data published by Merck
Frosst.6
15. The binding assay was conducted using stable EP1
transfected CHO-K1 membranes and [3H]-PGE2 as the
radio ligand. Cell membranes, compounds and [3H]-PGE2
(3 nM final assay concentration) were incubated in a final
volume of 100 lL for 30 min at 30 °C. The radioactivity
retained was measured by liquid scintillation counting in a
Packard TopCount scintillation counter. Values are means
of at least four experiments.
16. Schild analysis was conducted using Chinese hamster
ovary (CHO) cells, stably transfected with the NFAT
reporter gene and human EP1 receptor (pCIN3 vector).
Cells were quiesced 24 h prior to assay. Dose-related
response curves were obtained to PGE2 in the absence
and presence of five concentrations of antagonist. Cells
were incubated for 4 h to allow for transcription and
translation of the reporter gene. The amount of reporter
gene produced was assumed to be directly proportional
to the level of receptor activation and this was measured
by addition of Perkin Elmer Luclite reagent. The plates
were then read on a Packard Topcount and the data
analysed in Graphpad Prism. PGE2 EC50 values in the
absence and presence of antagonist were obtained. Dose
ratios were calculated and then the log (dose ratio À 1)
was plotted against Log (antagonist) concentration to
produce the Schild graph from which the pA2 value and
the slope were determined by linear regression.
References and notes
1. Narumiya, S.; Sugimoto, Y.; Ushikubi, F. Physiol. Rev.
1999, 79, 1193.
2. Maruyama, T.; Koketsu, M.; Yamamoto, H.; Yamamoto,
K.; Yamamoto, L. T.; Hayashida, K.-I.; Ohuchida, S.;
Kondo,K.ProstaglandinsOtherLipidMediat.1999,59,217.
3. Kawahara, H.; Sakamoto, A.; Takeda, S.; Onodera, H.;
Imaki, J.; Ogawa, R. Anesth. Analg. 2001, 93, 1012.
4. (a) Omote, K.; Kawamata, T.; Nakayama, Y.; Kawamata,
M.; Hazama, K.; Namiki, A. Anesth. Analg. 2001, 92, 233;
(b) Omote, K.; Yamamoto, H.; Kawamata, T.; Nakay-
ama, Y.; Namiki, A. Anesth. Analg. 2002, 95, 1708.
5. Sarkar, S.; Hobson, A. R.; Hughes, A.; Growcott, J.;
Woolf, C. J.; Thompson, D. G.; Aziz, Q. Gastroenterology
2003, 124, 18.
`
6. Ducharme, Y.; Blouin, M.; Carriere, M.-C.; Chateauneuf,
A.; Cote, B.; Denis, D.; Frenette, R.; Greig, G.; Kargman,
ˆ ´
S.; Lamontagne, S.; Martins, E.; Nantel, F.; O’Neill, G.;
Sawyer, N.; Metters, K. M.; Friesen, R. W. Bioorg. Med.
Chem. Lett. 2005, 15, 1155.
7. Minami, T.; Nakano, H.; Kobayashi, T.; Sugimoto, Y.;
Ushikubi, F.; Ichikawa, A.; Narumiya, S.; Ito, S. Br.
J. Pharmacol. 2001, 133, 438.
8. Breault, G. A. WO9700863A1, 1997.
9. Watanabe, K.; Kawamori, T.; Nakatsugi, S.; Ohta, T.;
Ohuchida, S.; Yamamoto, H.; Maruyama, T.; Kondo, K.;
Narumiya, S.; Sugimura, T.; Wakabayashi, K. Cancer
Lett. 2000, 156, 57.
17. Alkylating agents 11 and 12 can be prepared by coupling
the appropriate amine in DMF with bromoacetic acid
using DIC.