A knowledge-based, structural-aided discovery of a novel class of 2-phenylimidazo[1,2-a]pyridine-6-carboxamide H-PGDS inhibitors

Article abstract of DOI:10.1016/j.bmcl.2021.128113

Through an internal virtual screen at GlaxoSmithKline a distinct class of 2-phenylimidazo[1,2-a]pyridine-6-carboxamide H-PGDS inhibitors were discovered. Careful evaluation of crystal structures and SAR led to a novel, potent, and orally active imidazopyridine inhibitor of H-PGDS, 20b. Herein, describes the identification of 2 classes of inhibitors, their syntheses, and their challenges.

Full text of DOI:10.1016/j.bmcl.2021.128113

Journal Pre-proofs
A knowledge-based, structural-aided discovery of a novel class of 2-phenyli‐
midazo[1,2-a]pyridine-6-carboxamide H-PGDS inhibitors
David N. Deaton, Elsie Diaz, Young Do, Robert T. Gampe, Jeffrey H. Guss,
Ashley P. Hancock, Heather Hobbs, Simon T. Hodgson, Jason Holt, Michael
R. Jeune, Kirsten M. Kahler, H. Fritz Kramer, Joelle Le, Paul N. Mortenson,
Caterina Musetti, Robert T. Nolte, Lisa A. Orband-Miller, Gregory E.
Peckham, Kim G. Petrov, Beth L. Pietrak, Chuck Poole, Daniel J. Price,
Gordon Saxty, Christie A. Schulte, Anthony Shillings, Terrence L. Smalley
Jr., Don O. Somers, Eugene L. Stewart, J. Darren Stuart, Stephen A. Thomson
PII:
S0960-894X(21)00339-5
DOI:
Reference:
https://doi.org/10.1016/j.bmcl.2021.128113
BMCL 128113
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
14 February 2021
4 May 2021
8 May 2021
Please cite this article as: Deaton, D.N., Diaz, E., Do, Y., Gampe, R.T., Guss, J.H., Hancock, A.P., Hobbs, H.,
Hodgson, S.T., Holt, J., Jeune, M.R., Kahler, K.M., Kramer, H.F., Le, J., Mortenson, P.N., Musetti, C., Nolte,
R.T., Orband-Miller, L.A., Peckham, G.E., Petrov, K.G., Pietrak, B.L., Poole, C., Price, D.J., Saxty, G., Schulte,
C.A., Shillings, A., Smalley, T.L. Jr., Somers, D.O., Stewart, E.L., Stuart, J.D., Thomson, S.A., A knowledge-
based, structural-aided discovery of a novel class of 2-phenylimidazo[1,2-a]pyridine-6-carboxamide H-PGDS
inhibitors, Bioorganic & Medicinal Chemistry Letters (2021), doi: https://doi.org/10.1016/j.bmcl.2021.128113
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. 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.
© 2021 Published by Elsevier Ltd.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com
A Knowledge-based, structural-aided discovery of a novel class of 2-phenylimidazo[1,2-a]pyridine-6-carboxamide H-
PGDS inhibitors
David N. Deaton,a Elsie Diaz, Young Do,a Robert T. Gampe,a Jeffrey H. Guss,b Ashley P. Hancock,c Heather Hobbs,c Simon T.
Hodgson,c Jason Holt,a Michael R. Jeune,a Kirsten M. Kahler,a H. Fritz Kramer,a Joelle Le,c Paul N. Mortenson,d Caterina
Musetti,b Robert T. Nolte,b Lisa A. Orband-Miller,a Gregory E. Peckham,a Kim G. Petrov,a Beth L. Pietrak,b Chuck Poole,a
Daniel J. Price,a Gordon Saxty,d Christie A. Schulte,a Anthony Shillings,c Terrence L. Smalley Jr.,a Don O. Somers,c Eugene L.
Stewart,a J. Darren Stuart,a Stephen A. Thomson,a
aGlaxoSmithKline, 5 Moore Drive, P.O. Box 13398, Research Triangle Park, NC 27709, USA
bGlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
cGlaxoSmithKline, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
dAstex Pharmaceuticals, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0QA, UK
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Through an internal virtual screen at GlaxoSmithKline a distinct class of 2-
phenylimidazo[1,2-a]pyridine-6-carboxamide H-PGDS inhibitors were
discovered. Careful evaluation of crystal structures and SAR led to a novel,
potent, and orally active imidazopyridine inhibitor of H-PGDS, 20b. Herein,
describes the identification of 2 classes of inhibitors, their syntheses, and their
challenges.
Keywords:
Hematopoietic prostaglandin D2 synthase
H-PGDS
PGD2
2009 Elsevier Ltd. All rights reserved.
Virtual screen
Nonsteroidal anti-inflammatory drugs (NSAIDs), which function
as cyclooxygenase (COX) inhibitors, are known for their
analgesic, antipyretic, and anti-inflammatory actions; however,
chronic treatment causes severe GI disorders.1,2 Two enzymes,
COX-1 and COX-2, convert arachidonic acid to prostaglandin H2
(PGH2), with COX-1 generating prostaglandins (PGs) for
homeostatic regulation and COX-2 generating pro-inflammatory
stimuli. Thus, selective COX-2 inhibitors were developed to
maintain anti-inflammatory efficacy, while sparing mucosal
integrity. Although, these inhibitors alleviated GI distress, they
resulted in increased cardiovascular events.3,4
potential anti-allergic and anti-inflammatory drugs with fewer
side effects than COX inhibition.
Downstream of the COX pathway, PGH2 is converted into other
PGs by various synthases during the inflammatory response.5,6
One of these PGs, prostaglandin D2 (PGD2), is produced by
hematopoietic prostaglandin D synthase (H-PGDS) and/or
lipocalin prostaglandin D synthase (L-PGDS). H-PGDS derived
PGD2, signaling through two G-protein coupled receptors, d-type
prostanoid receptor DP1/DP2, has been implicated in allergic
diseases.7 Therefore, selective H-PGDS inhibitors could serve as
Figure 1. Examples of H-PGDS inhibitors
In recent years, inhibitors of H-PGDS have been disclosed by
several organizations,8 including Evotec,9 AstraZeneca,10 Sanofi-
———
 Corresponding author. e-mail: christie.a.schulte@gsk.com

Aventis,11 and Pfizer12 (see Figure 1); several of these
compounds demonstrated efficacy in animal models. Recently,
Taiho has entered the clinic with TAS-205.13 Although the
structure has not been disclosed, the inhibitor shown is a
representative structure.13,14
drug metabolism and pharmacokinetics (DMPK) of inhibitor 1b
were evaluated in mice. It’s modest oral exposure (DNAUC =
426 ng•hr/mL) may reflect both its moderate half-life (t1/2 = 2 hr)
and its poor FaS-SIF solubility (0.004 mg/mL).
An internal, knowledge-based virtual screening afforded hit
molecule 1a (H-PGDS IC50 = 2.8 μM), identified from a 5,6-
fused cluster, as a weak, competitive, reversible inhibitor of
human H-PGDS in a fluorescence polarization (FP) assay.15 This
compound possesses good drug-like properties, including
reasonable solubility (Fasted State-Simulated Intestinal Fluid
(FaS-SIF)16 = 0.552 mg/mL), high artificial membrane
permeability17,18 (PAPP = 650 nm/sec), and high ligand efficacy19
(LE = 0.41).
Figure 3. Elaboration from fragment 1a to inhibitor 1b
In order to maximize the likelihood of in vivo efficacy in future
molecules, we set goals to improve cellular inhibitory activity
(IC50 ≤ 100 nM), increase FaS-SIF solubility (≥ 0.05 mg/mL),
enhance oral exposure (DNAUC ≥ 1000 ng•hr/mL) and lengthen
the terminal half-life, t1/2 ≥ 2 hr). In parallel, we explored the
SAR of the phenyl region and the urea amine region.
The synthesis of the desired targets is outlined in Scheme 1.
First, commercially available protected piperidone 2 was
thermally condensed with Brederick’s reagent 322 neat at 110 oC
to afford dimethylaminomethylenopiperdinone 4. Subsequent
condensation of 4 with appropriate hydrazines 5a-5c, followed
by cyclization, and elimination resulted in 6,7-dihydro-1H-
pyrazolo[4,3-c]pyridines 6a-6c. Then, acid-catalyzed cleavage of
the protecting group gave the amines 7a-7c as their hydrochloride
salts. Ureas and carbamates were prepared from these
hydrochloride salts, using one of the following methods.
Activation via triphosgene, carbonyl diimidazole, or p-
nitrophenyl chloroformate, then coupling with amines 8a-8c or
alcohols 8d to form the desired ureas/carbamates 1b-1i.
Alternatively, Curtius rearrangement of carboxylic acids 8e-8f &
8h also afforded amines. Coupling with substrates 7a-7c
afforded ureas 1e-1f & 1h.
Figure 2. Hit molecule 1a (yellow) bound to H-PGDS and GSH (cyan)
illustrating two key protein/ligand interactions: 1) partial π- π stacking with
104Trp and 2) a dual water-mediated hydrogen bonding network among the N2
of the dihydropyrazolepyridine, the β-hydroxy group of 122Thr and the
backbone carbonyl of 13Glu.
As shown in Figure 2, a soaked crystal structure of molecule 1a
bound to the H-PGDS active site showed two key interactions.
First, a partial face to face π-π stacking interaction occurs
between the core 5,6-ring system and the indole of 104Trp.
Second, a hydrogen bonding interaction was observed between
N2 of the core ring to the hydroxy side chain of 159Thr through a
network of waters to protein backbone. These key interactions
are consistent with those observed for other H-PGDS inhibitors
described in the literature,20,21 however, the minimalist hit
molecule is missing interactions characteristic of more advanced
compounds as outlined in Figure 2. In particular, molecule 1a
does not make a hydrogen bond to the glutathione co-factor, nor
does it exploit a hydrophobic surface en route to the bulk solvent
region beyond the ethyl substituent. Achieving these interactions
provides an opportunity for improved potency.
Scheme 1. Synthesis of 1,4,6,7-tetrahydro-5H-pyrazolo[4,3-
c]pyridine-5- carboxamide analogs
Employing SAR from other series,21 we appended a cyclohexyl
amine derivative to yield urea 1b (see Figure 3). All compounds
herein were screened in a functional RapidFireTM mass
spectrometry assay, measuring the inhibition of the conversion of
PGH2, generated in situ by COX-2 from arachidonic acid, to
PGD2 by H-PGDS. Compound 1b exhibits good inhibitory
potency (IC50 = 45 nM), as shown in Table 1. It is also active in a
rat basophilic leukemia (RBL) cell PGD2 production assay (IC50
= 510 nM). Furthermore, compound 1b is stable in S9 liver
fractions (t1/2 > 180 min). Given the promising in vitro results, the
a) 3, 110 °C, 70-80%; b) 5a-5c, Na2CO3, HOAc, MeOH:H2O (2:1), 45-89%;
c) 4M HCl in dioxane, CH2Cl2, 80-100%; d) triphosgene, NaHCO3, K2CO3,
8a,8c (X = NH), CH2Cl2; 28-41%; e) CDI, iPr2NEt, 8g (X = NH), CH2Cl2,
57%; f) p-NO2-C6H4-COCl, NEt3, 8d,8i (X = O, NH), CH2Cl2 or MeCN, 43-
60%; g) 8e-8f & 8h (X = CO2), NEt3, DPPA, MePh, 100 °C, 51-72%.
Screening data for the substituted 1,4,6,7-tetrahydro-5H-
pyrazolo[4,3-c]pyridine-5-carboxamide analogues is summarized

in Table 1. As an initial starting point for the examination of the
SAR, fluorine substituents were explored at various positions on
the phenyl ring. As shown above by inhibitor 1b, the meta
position tolerates fluoro substitution. In contrast, the ortho and
para fluoro analogs 1c and 1d result in a significant loss in
enzymatic activity. Simultaneous investigation of the urea was
performed in the meta-fluoro analog. Despite this region
extending toward bulk water, the SAR is relatively tight. We
focused on cyclic aliphatic amine targets, rather than aromatic
ones, to maintain more sp3 character, in an effort to potentially
disrupt crystal packing and improve intestinal dissolution.
Utilizing internal SAR21, a small array of aliphatic analogs, with
hydrogen bond donor/acceptor moieties, were prepared, with the
aim of improving inhibitory activity and/or solubility. No
significant improvement in enzymatic potency was observed for
this set of analogs. Analogs 1g and 1h display comparable
potency to compound 1b, yet exhibit poor metabolic stability
(e.g. 1g microsomes t1/2= 15 min).
1h
1i
H
H
F
F
H
H
37
± 18
0.010
N/A
46
± 19
aIC50 = concentration that inhibits 50% of the conversion of in situ generated
PGH2 to PGD2 by human H-PGDS and SD = standard deviation with
replicates of N ≥ 4 unless otherwise noted.
bFaS-SIF = fasted state simulated intestinal fluid solubility with N=1
cN = 2.
Desiring to improve the π-π stacking interactions between the
central core and the indole of 104Trp, the core was fully
aromatized. The synthesis of indazole 12 is outlined in Scheme 2.
N-Arylation of commercially available 5-carboxyindazole methyl
ester 9 with iodide 10 under Hartwig-Buchwald copper-mediated
coupling conditions gave phenylindazole 1123, after ester
hydrolysis. Amide coupling of the acid 11 with amine 8a
provided the desired product 12.
Based on the environment of the glutathione cysteine in the
crystal structure, we suspected that it can exist as either the thiol
or the thiolate, thus acting as either a donor or acceptor. To probe
the necessity of the urea NH for potential hydrogen bonding to
the glutathione cofactor, it was replaced with an oxygen, as in
carbamate 1i. Surprisingly, the carbamate analog retains similar
potency to the parent urea. While consistent with our hypothesis
of the thiol/thiolate equilibrium, it is equally possible that the N-
H of the urea does not participate in hydrogen bonding or the
interaction is weak and the carbamate pays a lower desolvation
penalty versus the urea.
Scheme 2. Synthesis of indazole and indole
Table 1. Screening data for substituted 1,4,6,7-tetrahydro-5H-
pyrazolo[4,3-c]pyridine-5- carboxamide analogs
a) 9 or 14, 10, CuI, K3PO4, (1R,2R)-N1, N2-dimethylcyclohexane-1,2-
diamine, PhMe, 110 °C (9-99%) b) LiOH, MeOH, THF, H2O, (95-99%) c) 11
or 13, HATU, 8a, iPr2NEt, DMF (13-100%).
Gratifyingly, compound 12 achieves the desired improvement in
potency with a ~7-fold increase in enzymatic potency (Figure 4).
The preference for a fully aromatic bicyclic core encouraged us
to consider alternative heterocycles. In particular, we wished to
challenge the necessity of the putative hydrogen bond interaction
between the pyrazole N2 and a structural water, identified in the
crystal structure of the original fragment 1a. The indole analog
15 was therefore synthesized, as outlined in Scheme 2.
Commercially available 5-carboxyindole acid 13 was coupled
with amine 8a using a common amide coupling protocol to give
amide 14. Subsequent Hartwig-Buchwald copper-mediated N-
arylation with iodide 10 gave compound 15.23 Interestingly, the
indole 15 had similar potency relative to the indazole 12,
suggesting that the aforementioned hydrogen bond to water is no
longer relevant for enzymatic potency. A possible explanation is
the lower desolvation penalty upon indole binding compensates
for the loss of the hydrogen bond.
a
#
R1
R2 R3 R4 IC50
FaS-
SIFb
± SD
nM mg/mL
5a 5b 5c
1b
1c
H
F
H
45
± 21
380
±
220
1700
±
1100
160
±
0.004
0.19
F
H
H
1d
1e
H
H
H
F
F
N/A
H
0.054
100
1f
H
H
F
F
H
H
230
±
130
20c
0.90
1g
0.036
± 10

Figure 4. Comparison of 1,4,6,7-tetrahydro-5H-pyrazolo[4,3-
c]pyridine 1b, indazole 12, and indole 15
a) 16, 17a or 17b, EtOH, 80 °C (77%); b) iPr2NEt, CO, Pd(dppf)Cl2, EtOH,
90 °C (16-92%); c) LiOH, EtOH, THF, H2O (78-99%); d) 8a, 8g, 8h, or 8i,
HATU, iPr2NEt, DMF (32-99%); e) 22a, 22b, 22c, or 22d, 2M Na2CO3,
Pd(dppf)Cl2•CH2Cl2, 1,4-dioxane. 80 °C (11-51%).
The discrepant trends in the SAR between this series and
literature8 suggests some uniqueness to 1b binding mode.
Specifically, the fact that the potency of 15 is equivalent to 12,
which coordinates the structural water, is compelling. As a result,
we endeavored and ultimately obtained a cocrystal structure of
hH-PGDS (human H-PGDS) bound to 1b, which revealed a
distinct rearrangement of the water network relative to
compounds such as 20a (see Figure 5). The modifications that
evolved compound 1b to 15 raise concerns that potency in this
binding mode is predominantly hydrophobic in nature, and the
hazards of developing molecules with entropy-driven potency are
well-documented.24,25 We therefore considered a strategy to
modify the core to specifically target the previously reported
through-water hydrogen-bonding network, and thereby provide
greater enthalpic balance to the affinity. As an initial test, a series
of substituted 2-phenylimidazopyridines were prepared. We
rationalized that shifting the phenyl ring of inhibitor 15 one atom,
and providing a basic nitrogen capable of hydrogen bonding to
water, would together more closely mimic both the shape and
polar interactions of previously reported inhibitors.11,12As shown
in Scheme 3, two routes were employed for the synthesis of the
imidazopyridines. In route one, condensation of an appropriate
aryl α-bromo-ketone with commercially available 2-amino-5-
bromo-pyridine provided the desired 2-phenyl-5-
Inhibition data for the substituted imidazopyridine analogs are
summarized in Table 2. Gratifyingly, the phenyl imidazopyridine
20a improves enzymatic inhibitory potency with a corresponding
increase in cellular activity relative to pyrazole 1b. Moreover,
this inhibitor 20a (LE = 0.40) also has a similar ligand efficiency
to starting fragment 1a. To gain greater insight into the binding
mode of this analog, a co-crystal structure of 20a with hH-PGDS
and glutathione was obtained, as shown in Figure 5. Similar to
the previously reported quinoline class of inhibitors, the
imidazopyridine core participates in a π- π stacking interaction
with the indole side chain of 104Trp. Also, the imidazopyridine
nitrogen adjacent to the phenyl ring forms a hydrogen bond with
the conserved water molecule held in place by hydrogen bonds to
159Thr and backbone carbonyl of the N-terminal 122Leu and an
additional water molecule. Furthermore, the amide nitrogen
donates its hydrogen to the sulfur of the glutathione cofactor
retaining this key interaction. This structure provided validation
that the design principles had been achieved.
Table 2. Screening data for substituted imidazopyridine analogs
bromoimidazopyridines 18a-18b. Then, conversion of the
bromides 18a-18b to the corresponding esters, employing
palladium-catalyzed carbonylation conditions, followed by
subsequent base-promoted hydrolysis afforded the carboxylic
acids 19a-19b. Finally, amide formation using standard
conditions with various amines yielded the desired targets 20a-
20i. To overcome low yields in the carbonylation reaction, an
alternative synthetic route was also developed. Commercially
available 2-bromoimidazopyridine-5-methyl ester 21 was
saponified, and the resulting carboxylic acid 21 was coupled to
various amines, under standard conditions, yielding amides 21a-
21d. Finally, the aryl moieties were installed, employing standard
Suzuki conditions, resulting in the desired targets 20a-20i.
S9d
t1/2
min
#
R1
R
R
R
H-PGDS
IC50
± SD
nM
9
± 6
18
± 10
RBL
IC50
FaS-
SIFc
mg/
mL
2
3
4
a
b
± SD
nM
100
± 92
160
±
140
120
± 76
190
0
20a
20b
H
H
H
F
H
H
0.001 >180
0.002 >180
20c
20d
F
H
H
H
F
8
± 0.4
270
± 130
N/A
N/A
>180
N/A
Scheme 3. Synthesis of imidazopyridines
H
± 31
430
0
±
140
20e
20f
H
H
H
H
H
H
1100
± 240
1.6
N/A
120
67
84
0.004
± 24
± 48

20g
20h
20i
H
H
F
H
F
H
H
H
30c
± 15
460
± 30
0.058 >180
min
>180
>180
>180
>180
>180
ng/mL
350
120
1300
810
430
ng•hr/mL
430
hr
2.2
1.1
2.1
1.1
0.5
1b
20a
20b
20g
20i
49
2200
260
130
45
± 21
710e
N/A
>180
aS9 = S9 liver slice in vitro metabolism half-life in minutes
bCMax = maximum concentration of inhibitor after oral dose of 3 mg/kg p.o.
cDNAUC = oral dose normalized area under the curve in nghr/mL
dt1/2 = terminal half-life after oral dosing at 3 mg/kg
H
12
± 5
190
± 40
0.002 >180
In an effort to improve solubility, we focused our attention on
optimizing the amide linked solvent exposed region employing a
similar strategy explored earlier using various amines. Piperdine
analogues 20e-20f were incorporated to introduce solubilizing
groups in the form of a hydrogen bond acceptor with attenuated
basicity through the ring nitrogen. The more basic ethyl
aH-PGDS IC50 = concentration that inhibits 50% of the conversion of in situ
generated PGH2 to PGD2 by human H-PGDS and SD = standard deviation
with replicates of N ≥ 4 unless otherwise noted.
bRBL IC50 = concentration that inhibits 50% of the conversion of PGH2 to
PGD2 by rat basophilic leukemia cells and SD = standard deviation with
replicates of N ≥ 2 unless otherwise noted.
cFaS-SIF = fasted state simulated intestinal fluid solubility with N =1
dS9 = S9 liver slice in vitro metabolism half-life in minutes
eN = 1.
piperidine compound 20e unfortunately provides >100-fold loss
in enzymatic potency in comparison to compound 20a, whereas
incorporating the less basic trifluoroethyl piperidine reported by
Pfizer12 (compound 20f) resulted in an approximately 8-fold loss
in enzymatic potency. However, compound 20f was found to be
slightly less metabolically stable in vitro (S9 t1/2 = 120 min).
We also focused efforts on the trans-cyclohexyl-tertiary alcohol
amine found in compound 20a where we envisioned placing a
solubilizing heteroatom in the cyclohexyl ring, achieved by
changing the cyclohexyl ring to a tetrahydropyran ring resulting
in analogues 20g-20i. Gratifyingly, compound 20g (H-PGDS
IC50= 30 nM) retained similar enzymatic potency compared to
compound 20a, but more importantly gave an increase in FaS-
SIF solubility (0.058 mg/mL). Compound 20g also exhibited
good in vitro metabolic stability (S9 t1/2 > 180 min) and was
therefore dosed orally in a mouse DMPK study. Compound 20g
showed a 3-fold improvement in oral absorption (DNAUC = 260
ng-hr/mL) relative to compound 20a, and interestingly, did so
with an unsubstituted phenyl that was previously inferred to be
metabolically labile. Based upon our earlier hypothesis that the
unsubstituted phenyl group could be a metabolic liability and the
results we observed with compounds 20a-20d, we progressed the
o-fluorophenyl analogue 20i into an oral mouse DMPK study.
Disappointingly, 20i has less than desirable exposure
Figure 5. Comparison of co-crystal structures of compounds 1b (left panel,
green) and 20a (right panel, magenta) bound to H-PGDS and GSH (cyan).
Compounds 1b and 20a both exhibit a π- π stacking with 104Trp and a
hydrogen bond to GSH. Compound 20a shows one additional ligand/protein
interaction- a through-water mediated hydrogen bond between the N1 of the
imidazopyrdine and the N-terminal backbone carbonyl of 199Leu.
Compound 20a exhibits poor exposure (DNAUC = 75 ng-hr/mL)
in an oral mouse pharmacokinetic study, despite having excellent
in vitro stability in both mouse liver microsomes (t1/2 > 90 min)
and S9 liver fraction (t1/2 > 180 min). We speculated that the
unsubstituted phenyl ring could be subject to metabolism,
therefore fluorophenyl analogs were made 20b-20d. Despite the
phenyl ring of the imidazopyridines being slightly displaced in
the binding pocket relative to the previous series, similar SAR
was observed nonetheless. The o-fluorophenylimidazopyridine
20c and the m-fluorophenyl imidazopyridine 20b has similar
enzyme potency to the des-fluorophenyl analogue 20a with a
slight drop in cellular potency. In contrast, the p-fluorophenyl
imidazopyridine 20d loses significant enzymatic potency.
Analogues 20b-20d were all found to be stable with similar
results in the S9 liver fraction metabolic stability assay (t1/2 > 180
min). Of those analogues compound 20b was selected for an oral
mouse DMPK study, the results are summarized in Table 3.
Compound 20b provides desirable oral exposure (DNAUC =
2200 ng/hr/mL) and reasonable terminal half-life (t1/2 = 2 hr). We
achieved improvement in both enzymatic potency and cellular
activity with the imidazopyridines, as well as demonstrated a
reasonable oral pharmacokinetic profile. However, poorly soluble
molecules have significant development risks, so the low FaS-
SIF solubility of 20b limited its progression and warranted
further optimization of this series
(DNAUC=125 ng-hr/mL), perhaps related to poor solubility
(FaS-SIF = 0.002 mg/mL) and resultant poor absorption, or
possibly related to metabolism occuring elsewhere.
Unfortunately, the alternatively m-substituted fluorophenyl (20h)
was not sufficiently potent to progress.
Conclusions
We have described the discovery of a novel class of H-PGDS
inhibitors originating from an internal virtual screen-based drug
discovery effort. Initial optimization led to potent compounds
with a unique binding characteristic, whereby the through-water
hydrogen-bonding network that is commonly seen in previously
disclosed ligands, is not required for potency. The discovery of
this binding mode may benefit design of future H-PGDS
inhibitors. The deliberate reintroduction of features that do
exploit the through-water interactions and a focus on physical
properties in subsequent design cycles led to the identification of
compound 20b. Compound 20b represents a quality tool
molecule for H-PGDS studies, characterized by good oral mouse
PK, but carrying a high risk for clinical development due to its
poor FaS-SIF solubility.
Table 3. Pharmacokinetic Results (p.o.) for H-PGDS Inhibitors
b
d
#
S9a
t1/2
CMax
DNAUCc
t1/2
Animal Welfare:

All studies were conducted in accordance with the GSK Policy
on the Care, Welfare and Treatment of Laboratory Animals and
were reviewed the Institutional Animal Care and Use Committee
either at GSK or by the ethical review process at the institution
where the work was performed.
14. Takeshita, E.; Komaki, H.; Shimizu-Motohashi, Y.;
Ishiyama, A.; Sasaki, M.; Takeda, S. A phase I study of
TAS-205 in patients with Duchenne muscular dystrophy.
Ann. Clin. Trans. Neurol. 2018, 5, 1338-1349.
15. FP assay was purchased as a H-PGDS FP-Based Inhibitor
Screening Assay Kit – Green from Cayman Chemical
References
https://www.caymanchem.com/product/600007
1. Vane, J. R. Inhibition of prostaglandin synthesis as a
mechanism of action for aspirin-like drugs. Nat. New Biol.
1971, 231, 232–235.
16. Sou, T.; Bergström, C. A. S. Automated assays for
thermodynamic (equilibrium) solubility determination.
Drug Disc. Today: Technologies, 2018, 27, 11-19.
2. Laine, L. Approaches to nonsteroidal anti-inflammatory
drug use in the high-risk patient. Gastroenterology 2001,
120, 594-606.
17. Veber, D.F.; Johnson, S.R.; Chen, H-Y.; Smith, B. R.;
Ward, K. W.; Kopple, D. Molecular properties that
influence the oral bioavailability of drug candidates. J.
Med. Chem. 2002, 45, 2615-2623.
3. Nussmeier, N. A.; Whelton, A. A.; Brown, M. T.;
Langford, R. M.; Hoeft, A.; Parlow, J. L.; Boyce, S. W.;
Verburg, K. M. Complications of the COX-2 inhibitors
parecoxib and valdecoxib after cardiac surgery. N. Engl. J.
Med. 2005, 352, 1081-1091.
18. Kansy, M.; Senner, F.; Gubernator, K. Physicochemical
high throughput screening: parallel artificial membrane
permeation assay in the description of passive absorption
processes. J. Med. Chem. 1998, 41, 1007–1010.
4. Bresalier, R. S.; Sandler, R. S.; Quan, H.; Bolognese, J. A.;
Oxenius, B.; Horgan, K.; Lines, C.; Riddell, R.; Morton,
D.; Lanas, A.; Konstam, M. A.; Baron, J. A. Cardiovascular
events associated with rofecoxib in a colorectal adenoma
chemoprevention trial. N. Engl. J. Med. 2005, 352, 1092-
1102.
19. Hopkins, A.L.; Groom, C.R.; Alex, A. Ligand efficiency: a
useful metric for lead election. Drug Disc. Today. 2004, 9,
430-431.
5. Smith, W. L.; Urade, Y.; Jakobsson P.-J. Enzymes of the
cyclooxygenase pathways of prostanoid biosynthesis.
Chem. Rev. 2011, 111, 5821-5865.
20. Saxty, G.; Norton, D.; Affleck, K.; Clapham, D.; Cleasby,
A.; Coyle, J.; Day, P.; Frederickson, M.; Hancock, A.;
Hobbs, H.; Hutchinson, J.; Le, J.; Leveridge, M.;
McMenamin, R.; Mortenson, P.; Page, L.; Richardson, C.;
Russell, L.; Sherriff, E.; Teague, S.; Uddin, S.; Hodgson, S.
Identification of orally bioavailable small-molecule
inhibitors of hematopoietic prostaglandin D2 synthase
using X-ray fragment based drug discovery.
MedChemComm. 2014, 5, 134-141.
6. Watanabe, K. Recent reports about enzymes related to the
synthesis of prostaglandin (PG) F2 (PGF2α and 9α, 11β-
PGF2). J. Biochem. 2011, 150, 593-596.
7. Kupczyk,
M.;
Kuna,
P.
Targeting
the
PGD2/CRTH2/DP1signaling pathway in asthma and
allergic disease: Current status and future perspectives.
Drugs 2017, 77, 1281–1294.
21. Deaton, D. N.; Do, Y.; Holt, J.; Jeune, M. R.; Kramer, H.
F.; Larkin, A. L.; Orband-Miller, L. A.; Peckham, G. E.;
Poole, C.; Price, D. J.; Schaller, L. T.; Shen, Y.;
Shewchuck, L. M.; Stewart, E. L.; Stuart, J. D.; Thomson,
S. A.; Ward, P.; Wilson, J. W.; Xu, T.; Guss, J. H.; Musetti,
C.; Rendina, A. R.; Affleck, K.; Anders, D.; Hancock, A.
P.; Hobbs, H.; Hodgson, S. T.; Hutchinson, J.; Leveridge,
M. V.; Nicholls, H.; Smith, I. E. D.; Somers, D. O.;
Sneddon, H. F.; Uddin, S.; Cleasby, A.; Mortenson, P. N.;
Richardson, C.; Saxty, G. The Discovery of quinoline-3-
carboxamides as hematopoietic prostaglandin D synthase
(H-PGDS) inhibitors. Bioorg. Med. Chem. 2019, 27, 1456-
1478.
8. Thurairatnam, S. Hematopoietic prostaglandin D synthase
inhibitors. Prog. Med. Chem. 2012, 51, 97-133.
9. Hesterkamp, T.; Barker, J.; Davenport, A.; Whittaker, M.
Fragment based drug discovery using fluorescence
correlation spectroscopy techniques: Challenges and
solutions. Curr. Top. Med. Chem. 2007, 7, 1582-1591.
10. Hohwy, M.; Spadola, L.; Lundquist, B.; Hawtin, P.;
Dahmén, J.; Groth-Clausen, I.; Nilsson, E.; Persdotter, S.;
von Wachenfeldt, K.; Folmer, R.H.A.; Edman; K. Novel
prostaglandin D synthase inhibitors generated by fragment-
based drug design. J. Med. Chem. 2008, 51, 2178-2186.
22. Dhar, T. G. M.; Wrobleski, S. T. Preparation of
heterobicyclic compounds as p38 kinase inhibitors.
US 20080275052 A1, 2008
11. Weiberth, F.J.; Yu, Y.; Subotkowski, W.; Pemberton, C.
Demonstration on pilot-plant scale of the utility of 1,5,7-
triazabicyclo[4.4.0]dec-5-ene (TBD) as a catalyst in the
efficient amidation of an unactivated methyl ester. Org.
Proc. Res. Devel. 2012, 16, 1967-1969.
23. Connolly, P. J.; Bian, H.; Li, X.; Liu, L.; Macielag, M. J.;
McDonnell, M. E. Azetidinylpiperidine derivatives as
monoacylglycerol lipase inhibitors and their preparation
and use for the treatment of inflammatory pain.
WO1205471, 2012
12. Carron, C.P.; Trujillo, J.I.; Olson, K. L. Discovery of an
oral potent selective inhibitor of hematopoietic
prostaglandin D synthase (HPGDS) ACS. Med. Chem. Lett.
2010, 1, 59-63.
24. Lipinski, C. A. Drug-like properties and the causes of poor
solubility and poor permeability. J. Pharmacology Toxicol.
Methods. 2000, 44, 235–249.
13. Komaki, H.; Maegaki, Y.; Matsumura, T. Early phase 2
trial of TAS-205 in patients with Duchenne muscular
dystrophy. Ann. Clin. Transl. Neurol. 2020, 7, 181-190.

25. Benet, L. Z.; Hosey, C. M.; Ursu, O.; Oprea, T. I. BDDCS,
the Rule of 5 and drugability. Adv. Drug Delivery Rev.
2016, 101, 89–98.

Scheme 3. Synthesis of imidazopyridines
a) 16, 17a or 17b, EtOH, 80 °C (77%); b) iPr2NEt, CO, Pd(dppf)Cl2, EtOH,
90 °C (16-92%); c) LiOH, EtOH, THF, H2O (78-99%); d) 8a, 8g, 8h, or 8i,
HATU, iPr2NEt, DMF (32-99%); e) 22a, 22b, 22c, or 22d, 2M Na2CO3,
Pd(dppf)Cl2•CH2Cl2, 1,4-dioxane. 80 °C (11-51%).

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.
A Knowledge-based, structural-aided discovery of a novel class of
2-phenylimidazo[1,2-a]pyridine-6-carboxamide H-PGDS inhibitors
Leave this area blank for abstract info.
David N. Deaton,a Elsie Diaz,b Young Do,a Robert T. Gampe,a Ashley P. Hancock,c Heather Hobbs,c Simon T. Hodgson,c
Jeffrey H. Guss,b Jason Holt,a Michael R. Jeune,a Kirsten M. Kahler,a H. Fritz Kramer,a Joelle Le,c Paul N. Mortenson,d
Caterina Musetti,b Robert T. Nolte,b Lisa A. Orband-Miller,a Gregory E. Peckham,a Kim G. Petrov,a Beth L. Pietrak,b
Chuck Poole,a Daniel J. Price,a Gordon Saxty,d Christie A. Schultea, Anthony Shillings,c Terrence L. Smalley Jr.,a Don O.
Somers,c Eugene L. Stewart,a J. Darren Stuart,a Stephen A. Thomson,a
Through an internal virtual screen performed at GlaxoSmithKline a distinct class of 2-phenylimidazo[1,2-a]pyridine-6-
carboxamide H-PGDS inhibitors were discovered. Careful evaluation of crystal structures and SAR led to a novel, potent,
and orally active imidazopyridine inhibitor of H-PGDS, 20b. Herein, describes the identification of 2 classes of inhibitors,
their syntheses, and their challenges.

Declaration of Interest
The authors are employees of GlaxoSmithKline or
Astex Pharmaceuticals. This research did not
receive any specific grant from funding agencies in
the public, commercial, or not-for-profit sectors.