Discovery and optimization of pyridyl-cycloalkyl-carboxylic acids as inhibitors of microsomal prostaglandin E synthase-1 for the treatment of endometriosis

Article abstract of DOI:10.1016/j.bmcl.2019.07.007

Here we report on novel and potent pyridyl-cycloalkyl-carboxylic acid inhibitors of microsomal prostaglandin E synthase-1 (PTGES). PTGES produces, as part of the prostaglandin pathway, prostaglandin E2 which is a well-known driver for pain and inflammatio

Full text of DOI:10.1016/j.bmcl.2019.07.007

Accepted Manuscript
Discovery and optimization of pyridyl-cycloalkyl-carboxylic acids as inhibitors
of microsomal prostaglandin E synthase-1 for the treatment of endometriosis
Marcus Koppitz, Nico Bräuer, Antonius Ter Laak, Horst Irlbacher, Andrea
Rotgeri, Anne-Marie Coelho, Daryl Walter, Andreas Steinmeyer, Thomas M.
Zollner, Michaele Peters, Jens Nagel
PII:
S0960-894X(19)30452-4
https://doi.org/10.1016/j.bmcl.2019.07.007
BMCL 26548
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 May 2019
27 June 2019
2 July 2019
Please cite this article as: Koppitz, M., Bräuer, N., Laak, A.T., Irlbacher, H., Rotgeri, A., Coelho, A-M., Walter, D.,
Steinmeyer, A., Zollner, T.M., Peters, M., Nagel, J., Discovery and optimization of pyridyl-cycloalkyl-carboxylic
acids as inhibitors of microsomal prostaglandin E synthase-1 for the treatment of endometriosis, Bioorganic &
Medicinal Chemistry Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.07.007
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Discovery and optimization of pyridyl-
cycloalkyl-carboxylic acids as inhibitors of
microsomal prostaglandin E synthase-1 for
the treatment of endometriosis
Leave this area blank for abstract info.
Marcus Koppitz, Nico Bräuer, Antonius Ter Laak, Horst Irlbacher, Andrea Rotgeri, Anne-Marie Coelho, Daryl
Walter, Andreas Steinmeyer, Thomas M. Zollner, Michaele Peters and Jens Nagel
O
H
O
5a
HN
O
35 nM
&1
Cl
N

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com
Discovery and optimization of pyridyl-cycloalkyl-carboxylic acids as inhibitors of
microsomal prostaglandin E synthase-1 for the treatment of endometriosis
Marcus Koppitza, , Nico Bräuera, Antonius Ter Laaka, Horst Irlbachera, Andrea Rotgeria, Anne-Marie
Coelhob, Daryl Walterc, Andreas Steinmeyera, Thomas M. Zollnera, Michaele Petersa and Jens Nagela
aBayer AG, Pharmaceuticals R&D, 13342 Berlin, Germany
bEvotec SE, Manfred Eigen Campus, Essener Bogen 7, 22419 Hamburg, Germany
cEvotec (UK) Ltd, 112-114 Innovation Drive, Milton Park, Abingdon, Oxfordshire OX14 4RZ, UK
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Here we report on novel and potent pyridyl-cycloalkyl-carboxylic acid inhibitors of microsomal
prostaglandin E synthase-1 (PTGES). PTGES produces, as part of the prostaglandin pathway,
prostaglandin E2 which is a well-known driver for pain and inflammation. This fact together
with the observed upregulation of PTGES during inflammation suggests that blockade of the
enzyme might provide a beneficial treatment option for inflammation related conditions such as
endometriosis. Compound 5a, a close analogue of the screening hit, potently inhibited PTGES
in vitro, displayed excellent PK properties in vitro and in vivo and demonstrated efficacy in a
CFA-induced pain model in mice and in a rat dyspareunia endometriosis model and was
therefore selected for further studies.
,Keywords:
mPGES-1
PTGES
pain
2019 Elsevier Ltd. All rights reserved.
prostaglandin pathway
PGE2
-
———
 Corresponding author. e-mail: marcus.koppitz@bayer.com

O
O
HN
Endometriosis is a disease defined by the presence of
functional endometrial glands and stroma outside of the uterine
cavity . The symptoms of this estrogen-dependent disease vary
N
N
H
N
H
1
Cl
in their presentation and severity and include chronic pelvic pain,
1
dysmenorrhea, dyspareunia and subfertility2. As
a result
Eli Lilly
endometriosis can be both physically and emotionally debilitating
and significantly reduces the overall quality of life of affected
women. The condition is one of the most commonly encountered
benign problems in gynaecology. It develops predominantly in
women during reproductive age and regresses after menopause.
Endometriosis is estimated to affect up to 10% of women of
reproductive age in the overall population, but due to
misdiagnosis and diagnostic delay the exact prevalence is not
O
Cl
H
N
N
N
H
CHF2
F
N
O
Cl
H
N
2
O
Boehringer Ingelheim
3
known . Several studies provided broadly accepted evidence
Cl
that important processes relevant to the development of
endometriosis are mediated by increased levels of the
prostaglandin PGE24. The pathway leading to PGE2 synthesis is
highly up-regulated in endometriosis and other inflammation
related diseases. It starts with the release of arachidonic acid from
membrane glycerophospholipids by the action of phospholipase
A2; conversion of arachidonic acid to the endoperoxide PGH2 by
cyclooxygenases 1 and 2 (COX-1, COX-2); and by the
O
O
Cl HN
N
H
N
Cl
3
Glenmark
Figure 1. Selected PTGES inhibitors with common pharmacophore.
isomerization of PGH2 to PGE2 by PGE2 synthases (PTGES)4, 5
.
At least three PTGES are known. While PTGES2 and
PTGES3 are constitutively expressed and expected to contribute
mainly to basal homeostatic PGE2 levels, PTGES is an induced
synthase which is massively up-regulated under inflammatory
conditions and in endometriotic lesions4,6. To overcome the
related PGE2 induced pain states, non-steroidal anti-
inflammatory drugs (NSAIDs) that block both COX isoforms,
COX-1 and COX-2, are frequently used, despite clear scientific
evidence for clinical efficacy of chronic treatment on
Since PTGES is a prime target for the treatment of various
different inflammatory conditions, there has been constant and
continued interest in research for the development of selective
PTGES inhibitors. A motif common to many of the disclosed
inhibitors
is
an
ortho-substituted
(het-)arylmethylen
pivaloylamide, a key pharmacophore found in structures from Eli
13,14
Lilly12 (1), Boehringer Ingelheim
Glenmark
(2) and, more recently,
15-18
(3) (Fig. 1). Our own chemical matter was
identified following high-throughput screening using a HTRF-
based biochemical assay monitoring the conversion of PGH2 to
PGE2 by PTGES. Compound 4 was selected as a promising
starting point for further investigations. Beyond classical
optimization goals such as potency and PK properties, and due to
differences between the human and rodent proteins,a further key
parameter to improve upon was rodent potency. This was
essential in order to enable the in vivo characterization of
advanced compounds in relevant rodent models.
7
endometriosis associated pain is lacking . NSAIDs are reported
to be effective in treatment of acute pain e.g. on primary
8
dysmenorrhea , but their safe usage duration is limited due to
their unselective inhibition of PGE2 synthesis, including PGE2
which is involved in homeostasis. Additionally the synthesis of
all other eicosanoids downstream of PGH2 is substantially
affected. Consequently, NSAIDs are not recommended for long
term application due to risk of several complications such as
ulcers or bleeding within the gastrointestinal tract9, kidney
damage, hypertension and asthma, or an increased risk of
Initial SAR studies investigated the terminal pyridyl group
(R1). It quickly became evident that a pyridyl residue, in
particular the meta-pyridyl moiety was the preferred motif for
activity. Further meta substituents on the pyridine ring increased
potency significantly. Amongst these, chloro (5) and
trifluoromethyl (8) were of particular note, boosting potency by a
factor of around 10 and we therefore selected compound 5
containing the meta-chloropyridyl as a starting point for further
modifications (Fig. 2).
cardiovascular events 10
.
PTGES inhibition, being more selective in its effects, might
therefore constitute a safer alternative for the various treatment
paradigms of endometriosis such as chronic pelvic pain,
inflammatory pain, proliferation of endometriotic tissue, and
further processes related to inflammation11.
Figure 2. Screening hit compound 4 and advanced compound 5 with
assignment of R-positions

O
O
F
Selected SAR data are shown in Table 1. (For additional
information, please see the supporting info.) In the R2 position
we found that the para-biaryl motif was mandatory for potency
(cf. 15, 16). Small substituents such as fluoro, chloro or methyl in
the 3-position of the aryl ring , provided, beyond H-substitution,
highest activities (cf. 17, 18).
H O
H O
39
40
41
44
R4
R4
R3
R3
69
46
O
Table 1. SAR of Pyridyl-cycloalkyl-carboxylic acids
HN
O
h PTGES
Cmpd
Modified Area
Modified Element
72
IC50 [nM]
>20000
9260
15
6
7
R1
R1
R1
R1
R1
N
HN
O
4490
N
CF3
HN
O
O
8
45
48
49
R3
R3
R3
1960
N
O
10
14
161
N
N
Cl
>20000
>20000
276
O
NH
15
16
R2-R1
R2-R1
5860
N
>20000
In position R4 the chain length and the attachment point of the
carboxylic acid sidechain were found to be crucial to activity (see
23 – 25). Furthermore the carboxylic acid group could not be
replaced by amides (26) or acid bioisosteres such as tetrazoles
were inferior in the end (27). Incorporation of O or N
heteroatoms into the 3 position of the linker chain was possible
(28). Furthermore, methyl and ethyl substituents in the 3-position
were nicely tolerated as was the 2,3-cyclopropyl derivative which
was highly active as illustrated by compounds 30 and 31 (all
racemic mixtures). A single 2-methyl substituent did not,
however,further increase activity (32). The substitution in the 2-
position of the aromatic ring was also very important for good
activity. We found that activity increased in this order for close
analogues OCF3 < OCH3 < CF3 (33) < CH3. A shift of the methyl
group to the other three possible positions around the aromatic
ring was significantly worse (data not shown). With 2-methyl
group in place, we found that further substituents such as OCH3
or F were well tolerated in position 6 of the ring (39, 40).
F
17
18
R2
R2
30
70
Cl
R4
R4
10600
3090
H O
23
24
O
O
H O
O
OH
25
R4
>20000
O
H2
N
26
27
28
30
R4
R4
R4
R4
752
152
32
In position R3, compounds 4 and 5 were racemic mixtures at
the carbon bearing the cyclopentyl unit. Variations in this
position demonstrated, that a 5-membered ring was optimal for
activity, although a cyclobutyl ring was also tolerated (41).
Analogues substituted with acyclic chains such as an isobutyl or
methyl substituent were almost inactive. Replacing the
cyclopentyl ring with a phenyl ring was not possible (44).
Moreover, attempts to introduce polarity into the preferred 4 and
5-membered rings, for example with ether, amide or sulfone
groups, were unsuccessful (45). The amide connecting R3 to R4
was found to be essential and could neither be alkylated (48) nor
reversed (49).
N
N
O
O
N
N
H
O
H O
H O
29
O
H O
H O
H O
31
32
33
R4
R4
R4
30
The importance of the chiral center within R3 was investigated
next. Separation of the enantiomers of compound 5 by chiral SFC
yielded a highly active (-) enantiomer 5a and a less active (+)
enantiomer 5b . The absolute configurations of each enantiomer
were elucidated by BioTools (BioTools, Inc., 17546 Bee Line
O
O
245
381
Highway (SR 710), Jupiter, FL 33458, USA) using
a
combination of IR and VCD spectroscopy. Here, it was found
that the more active enantiomer had the R-configuration.
F3C

highly aligned (Fig. 3A) whereas a comparable binding mode for
the S-isomer of 5a was not feasible without losing at least 20
kcal/mol of estimated MMGBSA binding energy and at least two
important H-bond interactions. As shown in Fig. 3, the preferred
R-configuration of the central chiral carbon of 5a (purple)
orientates the cyclopentyl substituent in a dent caused by Gly35
and in between the non-aromatic, but hydrophobic side chains of
Ile32 and Leu39. Furthermore, the SAR of the underlying series is
well explained based on H-bond interactions of the carboxylic
acid of 5a with Ser127, Arg52 and His53, of the amide carbonyl of
5a with glutathione (GSH) and of the meta-chloro-pyridyl
nitrogen with Gln134, as well as - stacking of 5a with Tyr130
(yellow and cyan dotted lines, Fig. 3B).
With knowledge of the absolute stereochemistry, we were
then interested in an efficient synthesis of target compounds
leading to the more active R enantiomers thus obviating the need
for chiral chromatographic separation. An advanced synthetic
route for compound 5a is depicted in Scheme 1. Compound 5a
can be synthesized in a convergent manner by 5 (50 to 5a), and 7
linear steps (53 to 5a) respectively. Starting from commercially
available phenyl acetic acid derivative 53, introduction of the
cyclopentyl unit was achieved under standard basic conditions
using cyclopentyl bromide giving racemic ester 54. Subsequent
ester saponification yielded racemic acid 55, which was subjected
to a diastereomeric salt separation using R-phenethylamine: The
desired R enantiomer 56 precipitated and could be isolated by
filtration and washing. Treatment of 56 with thionyl chloride
afforded acid chloride 57. Production of the second required
building block 52 was achieved by Heck olefination of nitro-
bromo-precursor 50 and subsequent simultaneous reduction of
the double bond and the nitro group. Formation of 58 was
conducted using 52 and 57 and standard basic conditions at room
temperature. Subsequent Suzuki coupling with meta-pyridyl
boronic acid and deprotection of the t-butyl ester with TFA
finally yielded compound 5a.
Figure 3. Compound 5a (purple) docked into the crystal structure of human
microsomal prostaglandin E synthase-1 with synthetically modified
glutathione analog PS-1 (green, 4AL1.pdb). Glide docking of compound 5a
(purple) resulted in a binding mode in which the biaryl substituents of PS-1
and 5a in the proximity of Ile32 are highly aligned. The preferred R-
configuration of the central chiral carbon of 5a (purple) orientates the
cyclopropyl substituent in a dent caused by Gly35 and in between the
hydrophobic side chains of Ile32 and Leu39. (B) Identical binding mode of
compound 5a (purple) as in panel A, further explaining the SAR of the
underlying series based on H-bond interactions (yellow dotted lines) of the
carboxylic acid of 5a with Ser127, Arg52 and His53, of the amide carbonyl of 5a
with glutathion (green) and of the meta-chloro-pyridyl nitrogen with Gln134
as well as pi-stacking (cyan dotted line) with Tyr130
,
.
The R configuration of eutomer 5a was also consistent with
docking experiments for this series. For example, glide docking19
of compound 5a (purple) into the crystal structure of human
microsomal prostaglandin E synthase-1 with the synthetically
modified glutathione analog PS-1 (4AL1.pdb) resulted in a
binding mode in which the biaryl substituents of PS-1 and 5a are
.
O
O
Br
O
O
AND Enantiomer
b
a
+
AND Enantiomer
O
+
N
– O
O
NH2
52
N
– O
O
O
O
51
O
50
HN
O
g
h
HN
O
AND Enantiomer
Cl
O
H O
O
Cl
H O
O
f
e
Br
59
58
N
Br
Br
Br
56
O
57
55
i
AND Enantiomer
O
d
H O
O
O
O
HN
O
c
Cl
Br
Br
53
54
5a
N
Scheme 1. Synthesis of R-configurated compound 5a. Reagents and conditions: (a) tert-butyl acrylate, P(o-Tol)3, PdOAc2, TEA, DMF, 125°C, 107%; (b) H2
1 atm, Pd/C, EtOH, rt, 99%; (c) cyclopentyl bromide, KOtBu, DMF, 0°C, 100%; (d) NaOH, MeOH, H2O, rt, 94%; (e) (1R)-1-phenylethanamine, EtOH, H2O; (f)
SO2Cl2, 100°C, 99%; (g) TEA, DCM, rt, 100%; (h) 5-chloropyridin-3-yl)boronic acid, Pd(dppf)Cl2, K2CO3, THF, H2O, 80°C, 53%; (i) TFA, DCM, rt, 92%.

Highly active single stereoisomers where then selected for in
vitro pharmacokinetic profiling including tests for metabolic
stability in rat hepatocytes and permeability in Caco-2 cells
(Table 2).
potency frequently observed for various PTGES inhibitor classes
can be understood based on structural differences in the
proximity of the inhibitor binding site. For example, mutation
studies of Pawelzik et al. 21 revealed three key residues in human
PTGES which are substituted in rodents by considerably different
amino acids (Thr131 by Val, Leu135 by Phe and Ala138 by Phe)
and which were responsible for the observed species selectivity.
Table 2. PK properties for selected active (-) enantiomers
Caco-2
Papp A-
B Mari
[nm/s]
Delta
pKa
F
h
EH rat
hepato
cytes
Caco-2
efflux
ratio
Benchmark activities of our new reference compound 5a in
the rat and mouse PTGES assay were 1270 nM and 765 nM,
respectively. In general, we found that the compounds in our
series were significantly less active at mouse or rat enzymes with
factors typically ranging between 10- to 50-fold. However the
SAR trend across species were similar andincreasing human
potency typically also improved rodent potencies.
[%]
Cmpd
PTGESIC50
[nM]
calc.2 male
0
rat
4a
270
0.15
36
12
0.8
-
5a
8a
17a
28a
35
20
18
14
0.02
0.07
0.24
0.01
69
80
59
2.2
2.9
1.2
1.4
69
1.3
2.6
1.7
0.7
98
73
72
8
The diasteromeric mixture of compound 30 displayed high
human PTGES activities. Knowing that only the R-cyclopentyl
enantiomer was significantly active, and assuming that the
second stereocenter at C3 might influence human and rodent
potencies prompted us to perform a chiral separation of all 4
diastereomers and to assign the respective configuration using
VCD measurements. Indeed it turned out that the two cyclopentyl
isomers with the R-configuration 30a and 30b displayed the
highest human and rodent potencies (see Table 3). The stereo
centre in the carboxyclic acid sidechain had a relatively minor
influence on potency, however in the end 30a turned out to be the
most active compound in the mouse assay with an IC50 = 340
nM. On further examination of the available rodent data it
appeared that beyond increasing human potency there was
another structural handle available for improving rodent potency:
Racemic 3-methylpyridine-substituted compound 11, though not
very active at the human enzyme had an improved mouse:human
potency ratio of around 5-fold, indicating that alkylations in this
position may be beneficial. Further work was then directed
towards different combinations of alkyl groups in the 3-pyridyl
and C3 position of the acid sidechain (examples 60 – 62). Indeed,
we were able to obtain highly potent compounds at both human
and rodent enzymes, culminating in compound 62 with mouse
activity in the double digit nM range. However, all of our alkyl
analogues displayed inferior metabolic stability as compared to
our benchmark compound 5a as indicated by in vitro rat
hepatocyte data (see Table 4). Since 5a displayed excellent PK
properties in rats (low clearance (Clb 0.3 L/h/kg), a moderate
volume of distribution (Vss 0.9 L/kg), long half-lifeiv (t1/2 4.3 h)
and a high oral bioavailability (98%)), this compound was
selected for more detailed pharmacological in vivo
characterization in mechanistic and disease animal models.
All selected examples showed favourable metabolic stability
in rat hepatocytes with in vitro hepatic extraction ratios (EH)
between 0.02 and 0.24. Starting point 4a displayed a pronounced
efflux transport in Caco-2 cells demonstrating low permeability
in the absorptive direction and an efflux ratio (secretory versus
absorptive permeation) of 12. This efflux could be significantly
reduced by introduction of small electron-withdrawing
substituents into position 5 of the pyridyl ring. Efflux ratios of
1.2 -2.9 were observed for examples 5a, 8a and 17a. Here,
basicity of the pyridyl nitrogen is reduced, through introduction
of a chlorine, trifluoromethyl or fluoro substituent respectively.
Surprisingly,(-) enantiomer 28a, although having a chlorine
substituent as in 5a, had a very high efflux ratio of 69. 28a,
however, has an oxygen group in place of the beta-CH2-group in
the linker to the carboxylic acid which increases the acidity of
this group compared to the phenylpropanoic acid examples.
Inspection of Caco-2 permeability and efflux data (data for 5
compounds shown in Table 2) for 30 compounds in the series
suggested that a reduced acidity of the carboxylic acid and
basicity of the pyridyl group in R4 might be beneficial and delta
pKa (ac – bas) values above 1 appeared desirable in terms of
reducing active efflux. The observed in vitro pharmacokinetic
properties (high metabolic stability and Caco2-permeability) for
examples 5a, 8a, 17a translated into good oral bioavailability in
rats with values ranging from 72 - >95% making them suitable
candidates for further in vivo profiling. Similarly the high in vitro
efflux transport observed for example 28a translated into a very
low oral bioavailability (F<10%) in rats.
As mentioned earlier, reasonable rodent potency was desirable
in advanced compounds to enable meaningful in vivo
characterization in relevant animal models. The lower rodent
Table 3. Selected rodent-active compounds (R changes towards reference compound 5)
h PTGES IC50
r PTGES IC50
[nM]
m PTGES IC50
[nM]
Cmpd
R1
R3
R4
[nM]
O
O
O
O
HN
O
O
O
O
Cl
Cl
Cl
Cl
H O
H O
H O
H O
5a
35
1270
895
765
339
N
HN
HN
HN
30a
30b
30c
10
22
N
N
N
1360
8460
584
208
3080

O
O
HN
HN
O
O
Cl
H O
H O
30d
10
961
161
17700
2590
12600
818
N
N
Table 4. In vitro potency in different species and in vitro metabolic stability of selected rodent-active compounds (R1 and R4:
changes towards reference compound 5)
h PTGES IC50
[nM]
r PTGES IC50
[nM]
m PTGES IC50
[nM]
Cmpd
R1
R4
EH rat hepatocytes
O
H O
H O
60
21
15
10
190
-
203
191
84
0.44
N
O
61
62
0.45
0.46
N
O
H O
162
N
The effect of 5a on PGE2 formation and measures of pain
after CFA-induced inflammation in mice is depicted in Figure 4.
Female c57bl/6 wild type (WT, Taconic) mice were used for
evaluation of anti-nociceptive efficacy in the complete Freund´s
adjuvant (CFA) model of inflammatory pain. Animals received,
under isoflurane anesthesia, intraplantar injection of CFA (30 µl,
1 mg/ml, Sigma) into the left hind paw. Spontaneous pain-related
behavior in freely moving animals was assessed using the
automated dynamic weight bearing device (DWB, Bioseb,
22, 23
France) according to published and validated protocols
For
behavioral testing, the animal was placed inside a Plexiglas
chamber and allowed to move freely within the apparatus for a 5
min period, subsequently pain behavior was recorded for a test
period of another 5 min. 2 days after CFA injection, compound
5a was administrated as a single oral dose. 1h later, animal
behaviour was assessed. At termination of the experiments, hind
paw tissue is collected for analysis of PGE2 levels (ELISA) and
blood is collected for measurement of 5a concentration in
plasma. Single oral dosing of 5a (0, 30, 100 mg/kg, p.o.;
n=10/group; vehicle: EtOH:Solutol:water (10:40:50)) inhibited
PGE2 formation in inflamed hind paw (data not shown) and
reduced pain behavior (dynamic weight bearing) significantly in
this model (1-way ANOVA followed by Dunnett´s post hoc test,
GraphPad Prism, USA). The anti-nociceptive effect size was
similar to the technical control Ibuprofen (100 mg/kg, p.o.) with
residual levels of PGE2 retained after PTGES inhibition (Fig. 4).
Plasma analysis revealed that application of 100mg/kg of 5a led
to unbound exposures of 5a at the end of experiment covering
~3-fold IC50 concentrations in mice (IC50=765 nM).
Figure 4. Efficacy of 5a in CFA-induced inflammatory pain model in
rats.*p<0.05 1-way ANOVA comparing the CFA treated groups followed by
Dunnett’s multiple comparison test vs. vehicle group.
Further evaluation was then conducted in a rat endometriosis
model with the primary endpoint of vaginal hyperalgesia (vaginal
distension) modifying a method described in detail earlier24.
Female SD rats underwent endometriosis surgery in the estrus
phase. Uterine fragments were sewn onto the mesenteric artery
arc and other fragments were sewn onto the colonic wall. In
parallel, electrodes for later EMG recordings were implanted in
the abdominal wall to allow recording of the viscero-motor
response (VMR). Vaginal sensitivity was measured 5 and 6
weeks post-surgery in the proestrus phase. For vaginal distension,
a balloon was inserted into the vaginal canal and inflated with
increasing volumes of water using an infusion pump. In parallel,

EMG recordings were collected to measure the VMR during the
distension. Compound 5a (0, 75 mg/kg, p.o.) was administered
twice daily (bid) during three weeks (week 3 to week 5 post-
surgery). Repeated oral dosing of 5a (0, 75 mg/kg, bid; n=13-
14/group; vehicle: PEG400:DMSO (90:10)) significantly
decreased vaginal hyperalgesia versus vehicle-treated animals.
The effect was maintained in the treatment free period of week 6
(Fig. 5). Treatment with 5a also reduced proliferation in lesions
as measured by decreased gene expression of the proliferation
marker Ki67 vs. housekeeping control in Taqman (data not
shown). Thus, the data in the dyspareunia model indicate a
disease modifying component beyond analgesic effects following
sustained PTGES inhibition. In this chronic endometriosis model,
at 75 mg/kg bid unbound Caverage ~ IC60 of the rat PTGES enzyme
were achieved during the study. All experiments were performed
in strict compliance with company, regional, and federal
guidelines for the use of laboratory animals. The experiments
were approved and executed in accordance with policies and
directives of LAGESO (Landesamt fur Gesundheit und Soziales,
Berlin, Germany) and the Hamburg Animal Care and Use
Committee (Hamburg, Germany). All efforts were made to
minimize suffering.
The authors thank Steve Courtney and Markus Koch for
fruitful discussions and continued support of the project and
Henrik Dahlloef for contributions in the early phase of the
project.
Supplementary Material
Supplementary material associated with this article can be
found, in the online version.
References
1. Giudice LC. N Engl J Med 2010;;362:2389.
2. Bloski T, Pierson R. Nurs Womens Health 2008;12:382.
3. Zondervan KT, Becker CM, Koga K, MissmerSA, Taylor RN,
Vigano P. Nat Rev Dis Primers 2018;4:9.
4. Koeberle A, Werz O. Curr Med Chem 2009;16:4274.
5. Arosh J A, Lee J, Balasubbramanian D, Stanley J A, Long C R,
Meagher M W, Osteen K G, Bruner-Tran K. L, Burghardt R C,
Starzinski-Powitz A Banu, S. K. Proc Natl Acad Sci U S A 2015,
112, 9716
6. Brenneis C, Coste O, Altenrath K, Angioni C, Schmidt H, Schuh
CD, Zhang DD, Henke M, Weigert A, Brune B, Rubin B, Nusing
R, Scholich K, Geisslinger G. J Biol Chem 2011;286:2331.
7. Allen, C.; Hopewell, S.; Prentice, A.; Gregory, D. Cochrane
Database Syst Rev 2009, CD004753
WEEK 5
WEEK 6
No treatment 5a
8. Marjoribanks, J.; Proctor, M. L.; Farquhar, C. Cochrane Database
Syst Rev 2003
40
30
20
10
0
9. Labianca R, Sarzi-Puttini P, Zuccaro SM, Cherubino P, Vellucci
R, Fornasari D. Clin Drug Investig 2012;32 Suppl 1:53.
10. Scholich, K.; Geisslinger, G. Trends Pharmacol Sci 2006, 27, 399
11. Koeberle A, Werz O. Biochem Pharmacol 2015;98:1.
12. Schiffler MA, Antonysamy S, Bhattachar SN, Campanale KM,
Chandrasekhar S, Condon B, Desai PV, Fisher MJ, Groshong C,
Harvey A, Hickey MJ, Hughes NE, Jones SA, Kim EJ, Kuklish
SL, Luz JG, Norman BH. Rathmell RE, Rizzo JR, Seng TW,
Thibodeaux SJ, Woods TA, York JS, Yu XP. J Med Chem
2016;59:194.
13. Priebke H, Doods H, Kuelzer R et al. New compounds. PCT Intl.
Appl. WO/2012/022793, 2012.
14. Pfau R, Arndt K, Doods H, et al. 3H-Imidazo[4,5-c]pyridine-6-
carboxamides as anti-inflammatory agents. PCT Intl. Appl.
WO/2010/100249, 2010.
*
*
15. Gharat LA, Banerjee A, Khairatkar-Joshi N, Kattige VG. Bicyclic
Compounds As mPGES-1 Inhibitors. PCT Intl. Appl.
WO/2013/118071, 2013.
16. Gharat LA, Banerjee A, Khairatkar-Joshi N, Kattige VG.
Substituted Bicyclic Compounds As mPGES-1 Inhibitors. PCT
Intl. Appl. WO/2014/167444 2014.
Vehicle
*p<0.05 t-test vehicle vs 5a
5a
75 mg/kg bid p.o.
17. Gharat LA, Muthukaman N, Tambe MS, Pisal D, Khairatkar-Joshi
N, Kattige VG. Substituted Pyrimidine Compounds As mPGES-1
Inhibitors. PCT Intl. Appl. WO/2015/059618, 2015.
18. Dhuppas U, Chaudhari S, Rajurkar S, Jain N, Dhatrak C, Kasliwal
A. Nanoparticulate Formulation Comprising a mPGES-1 Inhibitor.
WO/2016/016861, 2016.
19. Friesner RA; Banks JL, Murphy RB, Halgren TA, Klicic JJ, Mainz
DT.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry, J. K.;
Shaw, D. E.; Francis, P.; Shenkin, PS. J Med Chem 2004;47:1739.
20. Robinson, I, Sargent B, Hatcher JP. Neurosci Lett 2012;524:107.
21. Pawelzik S-C, Uda NR, Spahiu L, Jegerschöld C, Stenberg P,
Hebert H, Morgenstern, R, Jakobsson P-J. J Biol Chem 2010
285(38):29254
22. Tetreault, P, Dansereau MA, Dore-Savard L, Beaudet N, Sarret P.
Physiol Behav 2011;104:495.
23. Berkley KJ, McAllister SL, Accius BE, Winnard KP. Pain
2007;132 Suppl 1:S150.
24. Fraczkiewicz R, Lobell M, Goller AH, Krenz U, Schoenneis R,
Clark RD, Hillisch A. J Chem Inf Model 2015;55:389.
Figure 5. Effect of 5a on vaginal hyperalgesia and proliferation in the
dyspareunia rat model.
In conclusion we have identified and optimized a novel series
of pyridyl-cycloalkyl-carboxylic acids as inhibitors of
microsomal prostaglandin
E synthase-1 for the potential
treatment of endometriosis. The series lacks the frequently found
arylmethylen pivaloylamide motif of other PTGES inhibitors and
is an example of an optimization programin which the screening
hit and optimized compounds are structurally very similar. The
potency of compound 5a, its efficacy in in vivo pain and
endometriosis models along with its excellent pharmacokinetic
profile prompted us to select it for further profiling.
Acknowledgments