Potent and highly selective DP1 antagonists with 2,3,4,9-tetrahydro-1H- carbazole as pharmacophore

Article abstract of DOI:10.1016/j.bmcl.2010.10.018

We discovered that the introduction of a methyl group to the benzylic position of the N-benzyl group in lead compound 1a has a dramatic effect on improving the binding selectivity of this ligand for the prostanoid receptors DP1 (receptor for prostaglandin

Full text of DOI:10.1016/j.bmcl.2010.10.018

Bioorganic & Medicinal Chemistry Letters 20 (2010) 7462–7465
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Potent and highly selective DP1 antagonists
with 2,3,4,9-tetrahydro-1H-carbazole as pharmacophore q
⇑
,
Lianhai Li a, , Christian Beaulieu a, Marie-Claude Carriere b, Danielle Denis b, Gillian Greig b,
Daniel Guay a, Gary O’Neill b, Robert Zamboni a, Zhaoyin Wang a
a Department of Medicinal Chemistry, Merck Frosst Centre for Therapeutic Research, 16711 Trans Canada Hwy., Kirkland, Québec, Canada H9H 3L1
b Department of Biochemistry, Merck Frosst Centre for Therapeutic Research, 16711 Trans Canada Hwy., Kirkland, Québec, Canada H9H 3L1
a r t i c l e i n f o
a b s t r a c t
Article history:
We discovered that the introduction of a methyl group to the benzylic position of the N-benzyl group in
lead compound 1a has a dramatic effect on improving the binding selectivity of this ligand for the pro-
stanoid receptors DP1 (receptor for prostaglandin D2) as compared to TP (receptor for thromboxane A2).
Based on this discovery, we have synthesized a series of potent and highly selective DP1 antagonists.
Among them, compound 1h was identified as a highly selective DP1 antagonist with excellent overall
properties. It has a Ki of 0.43 nM to DP1 in binding assay and an IC50 of 2.5 nM in the DP1 functional assay.
Its selectivity for DP1 over TP (the most potent receptor after DP1) exceeds 750-fold based on both bind-
ing and functional assays. These properties make 1h a very potent and highly selective DP1 receptor
antagonist suitable for investigating the biological functions of DP1 in normal physiology and models
of disease
Received 19 August 2010
Revised 4 October 2010
Accepted 5 October 2010
Available online 12 October 2010
Keywords:
DP1 receptor
Antagonist
Selectivity
Ó 2010 Elsevier Ltd. All rights reserved.
It is found that, in response to allergic stimulus, the production
of prostaglandin D2 (PGD2) along with a variety of other lipid medi-
ators such as histamine, some leukotrienes (cysLTs), and throm-
boxane A2 (TxA2) in mast cell are increased.1 This leads to the
therapeutic potential of using the receptors of PGD2, namely, DP1
and CRTH2/DP2 as targets for mediating various allergic disorders.
Such potential has been supported by a series of discoveries that
PGD2 is involved in various allergic disorders and appears to be
an important mediator of such disorders like allergic asthma,2 ato-
pic dermatitis,3 allergic rhinitis4 and allergic conjunctivitis.5 The
clinical utility of DP1 and CRTH2/DP2 ligands in allergic inflamma-
tion are currently under investigation.6 Recently, DP1 antagonists
have found application for treatment of the side-effects of niacin-
induced flushing. Thus, the combination of a DP1 antagonist,
Laropiprant (Fig. 1), with niacin has the therapeutic potential to
reduce flushing without lessening the beneficial effects of niacin
on treating dyslipidemias.
O
F
F
O
OH
Cl
N
N
OH
Cl
MeSO2
MeSO2
Laropiprant
1a
Figure 1.
1a. As a racemate, 1a has a Ki of 0.89 nM and 2.51 nM to DP1
and TP, respectively. We hoped that the modification of the N-ben-
zyl substituent could provide compounds with similar potency but
with much improved selectivity for DP1 over TP. Previous SAR
studies in the 1,2,3,4-tetrahydrocyclopenta[b]indole series showed
that the following N-benzyl modifications, including (1) the
replacement of the phenyl with other aromatic rings or hetero-
aromatic rings, and (2) the introduction of various substituents
to various positions of the phenyl ring, could not help achieve
the goal.9 This prompted us to examine the possibilities of intro-
ducing alkyl substituents to the benzylic position. We hoped that,
by doing so, the orientation of the benzyl substituent could be
positioned to discriminate the ligand structural requirement
between DP1 and TP, and thus led to the improvement of DP1
potency and selectivity over TP. In this communication, a detailed
account of this study is provided.
Continuing the on-going research efforts in developing a potent
and selective DP1 antagonist with an indole scaffold, we explored a
2,3,4,9-tetrahydro-1H-carbazole scaffold in stead of the 1,2,3,4-tet-
rahydrocyclopenta[b]indole scaffold used in our previous stud-
ies.7,8 The first compound we synthesized in this new series is
q
In memory of legacy Merck Frosst Centre for Therapeutic Research.
⇑
Corresponding author. Tel.: +1 514 4283837.
E-mail addresses: lianhai_li@merck.com, lianhai.li@gmail.com (L. Li).
Our initial synthetic pathway to benzylic alkylated analogs was
outlined in Scheme 1.10 By following a standard procedure of
Present address: 4201 Hugo, Pierrefonds, Québec, Canada H9H 2V6.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.018

L. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7462–7465
7463
O
F
b,c
OEt
a
O
5
F
F
F
O
N
H
a, b
4
O
OH
NH2
HCl
MeSO2
NH2
N
H
c
9a
Br
Br
Br
2
3
Cl
O
(R)-10a
O
d
F
F
O
F
N
H
O
Br
OH
Cl
1
OH
N
N
+
5
MeSO2
d, e
MeSO2
Cl
Cl
6
11b
1b
F
F
O
O
e, f
O
N
N
OH
Cl
O
N
H
O
MeSO2
Br
MeSO2
Cl
8
7
(R)-9b
Scheme 1. Reagents and conditions: (a) NaNO2, HCl, À4 °C; (b) SnCl2, HCl, 0–5 °C;
quantitative for two-steps; (c) Compound 4, HOAC, reflux, 32%; (d) 6 (4 equiv),
Cs2CO3 (8 equiv), MeCN, reflux, 5 h, 99%; (e) MeSO2Na (5 equiv), CuI (5 equiv), NMP/
150 °C, 8 h, 73%; (f) NaOH (5 equiv), MeOH, THF, rt, 3 h, quantitative.
Scheme 2. Reagents and conditions: (a) MeSO2Na (5 equiv), CuI (5 equiv), NMP/
150 °C, 8 h, 63%; (b) tBuO2CN@NCO2tBu, Ph3P, (R)-10a, rt, quantative; (c) KOH
(5 equiv), MeOH, THF, reflux, 3 h, quantitative; (d) H2,Pd/C, MeOH, rt, 16 h;
(e) CH2N2, THF, 95% yield for two-steps.
hydrazine preparation with 2 as starting material, compound 3
was obtained in quantitative yield. The Fisher indole reaction of
3 with ketoester 4 in refluxed acetic acid led to the formation of
bromoindole 5. Bromoindole 5 was alkylated at the indole nitrogen
with 4 equivalents of benzyl bromide 6 in the presence of 8 equiv-
alents of Cs2CO3 as a base in reflux acetonitrile. The alkylated prod-
uct 7 was treated with 5 equiv of MeSO2Na and 5 equiv of CuI in
NMP at 150 °C, followed by the hydrolysis of ester functional to af-
ford the desired acid 8 as a mixture of four stereoisomers. The two
pairs of diastereomers could be easily separated by flash chroma-
tography on silica gel. In DP1 binding assay, the slow-eluting pair
exhibited much higher potency with a Ki of 1.3 nM, as compared
to the fast-eluting pair with a Ki of 12 nM. We were gratified to find
that the more potent pair has a Ki of 219 nM to TP in binding assay;
this represents a much improved selectivity for DP1 over TP com-
pared to 1a.
Encouraged by these results, a more efficient synthetic method
was developed to identify the most potent single diastereomer of
mixture 8 responsible for the DP1 affinity and selectivity (Scheme
2).10 Direct alkylation of compound 5 with chiral alcohol (R)-10a
under Mitsunobu conditions failed to afford desired products.
However, when compound 5 was transformed into methyl sulfone
9 and then alkylated with (R)-10a under Mitsunobu conditions, the
desired alkylation products were obtained in quantitative yield.
The increased reactivity of 9 is likely due to the enhanced acidity
by the electron-withdrawing nature of the methyl sulfonyl substi-
tuent. The alkylated diastereomeric esters, which have an absolute
(S)-configuration at the N-benzylic position, were hydrolyzed to
their corresponding acids and separated by silica gel flash chroma-
tography. Compound 1b, the slow-eluting diastereomer exhibits a
Ki of 0.75 nM and 184 nM toward DP1 and TP, respectively. The
potency and selectivity of 1b on DP1 indicates that it is the most
potent single diastereomer among the mixture of four (compound
8) obtained by the pathway outlined in Scheme 1. Thus a (S)-con-
figuration at N-benzylic position was determined to be essential to
DP1 potency and selectivity over TP. The absolute stereochemistry
at the carbon center bearing the acetic acid group was established
to be (R)-configuration by X-ray crystallography of 9b, which is
obtained by the debenzylation of 1b under Pd-catalyzed
hydrogenation conditions following by treatment with diazometh-
ane. The confirmed stereochemistry is in consistence with same
stereochemistry required for DP1 activity observed for the
1,2,3,4-tetrahydrocyclopenta[b]indole series. Interestingly, it was
found later that the ester which leads to the desired slow-eluting
diastereomer could be selectively hydrolyzed to acid 1b by stirring
with 0.6 equiv of LiOH (1 M solution) in a solvent mixture (THF/
MeOH 3:1) at 0 °C, thus facilitating the isolation of 1b in prepara-
tive scale.
More analogs in the series were synthesized as outlined in
Scheme 210 by using suitable chiral alcohols with absolute stereo-
chemistry represented by 10. Some of the alkylation products thus
obtained, as a mixture of 12 and 13, which contained chemically
reactive functional groups on R or on Ar, were further exploited
as starting point for preparation of other analogs. For example,
when R = CH2OH, it can be transformed into R = CH2F or CHF2;
when Ar = 4-BrC6H4, it can be transformed into Ar = 4-NCC6H4 or
4-MeSO2C6H4.
The newly synthesized compounds were evaluated in the DP1
and TP binding assays, and the results are summarized in Table
1. As discussed above, the introduction of a methyl group to the
benzylic position of the N-benzyl group in the initial lead
F
F
a
O
O
O
O
9a
+
OH
Ar
N
N
MeSO2
F
MeSO2
Ar
N
Ar
R
R
R
12
10
13
b or c
O
MeSO2
OH
Ar
R
1
Scheme 3. Reagents and conditions: (a) tBuO2CN@NCO2tBu, Ph3P, 10, rt; (b) KOH
(5 equiv), MeOH, THF, reflux, 3 h, quantitative; (c) LiOH (0.6 equiv), MeOH, THF,
0 °C, 24 h.

7464
L. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7462–7465
Table 1
Table 2
DP1 and TP receptor binding affinities
Binding affinities of other receptors with 1h
F
F
O
O
N
N
OH
CF3
MeSO2
OH
MeSO2
Ar
R
1h
1
a
a
Compound
Ki (nM)
Entry
Compound
Ar
Ki (nM)
EP1
EP2
EP3
830
EP4
FP
IP
No.
R
DP1
0.89
TP
DP1/TP (fold)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1ab
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
H
4-ClC6H4
4-ClC6H4
2.51
184
<3
245
526
339
658
1005
941
786
303
278
182
89
1h
4590
2270
6610
4320
15,510
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
CFH2
CF2H
0.75
0.38
0.33
0.41
0.36
0.31
0.43
0.33
0.38
0.51
a
Average of three or more assays.
3-F-4-ClC6H4
3,4-Cl2C6H4
2-F-4-ClC6H4
C6H5
200
112
270
362
292
338
100
106
93
CFH2 (entry 14, 1n, Table 1) was also attempted. When compared
to 1b, both compounds 1m and 1n are more potent DP1 antago-
nists with slight improvement of selectivity for DP1 over TP.
Among all the selective and potent antagonists we discovered
above, compound 1h was selected for further evaluations due to
its excellent overall properties. The binding affinities of 1h to other
prostanoid receptors are summarized in Table 2. Compound 1h has
a more than 1900-fold selectivity for DP1 over EP3 and even much
higher selectivities over other receptors. In cell based assays in
human platelet rich plasma, compound 1h inhibits the PGD2-in-
duced accumulation of cAMP with an IC50 of 2.5 nM and platelet
aggregation induced with the thromboxane A2 mimetic U46619
3-ClC6H4
4-CF3C6H4
4-FC6H4
4-BrC6H4
4-CNC6H4
4-MeSO2C6H4
4-ClC6H4
16.5
0.36
0.39
1481
130
138
361
353
4-ClC6H4
a
Average of three or more assays.
Racemate.
b
compound 1a led to compound 1b with a dramatic improvement
of the selectivity for DP1 over TP. In the TP binding assay, a 184-
nM Ki for 1b as compared to a 2.5-nM one for 1a was observed; this
represents an over 80-fold improvement of selectivity for DP1 over
TP by comparing to 1a.
Having identified 1b as a more selective lead, we tried to further
improve its DP1 potency and selectivity over TP by introducing
other substituents to the phenyl ring of its N-benzyl group. Intro-
ducing a 3-fluoro (compound 1c) improves the DP1 affinity by
two-fold while maintains the TP affinity, thus improves the selec-
tivity for DP1 over TP to more than 520-fold (entry 3, Table 1). In
comparison, the introduction of a 3-chloro, which leads to 1d, in-
creases potencies on both DP1 and TP and slightly improves the
selectivity to over 330-fold (entry 4, Table 1). The introduction of
a 2-fluoro to 1b produces a similar effect on potencies and selectiv-
ity (entry 5, 1e, Table 1) as of a 3-fluoro and improves the selectiv-
ity to more than 600-fold.
Further improvement of the selectivity for DP1 over TP can be
achieved by removing the 4-chloro substituent from the phenyl
ring of N-benzyl group in 1b or by moving it to the 3-position.
When the 4-chloro group was removed, compound 1f was ob-
tained. It is among the most potent DP1 antagonists discovered
in this series and is the only compound that has a 1000-fold selec-
tivity for DP1 over TP in binding assays (entry 6, Table 1). When the
4-chloro substituent in 1b was moved to the corresponding 3-po-
sition, DP1 antagonist (1g) with a two-fold higher potency was ob-
tained. It exhibits about 940-fold selectivity for DP1 over TP.
Our further SAR exploration was focused on replacing the 4-
chloro substituent in 16b with other substituents such as 4-CF3
(entry 8, 1h, Table 1), 4-F (entry 9, 1i, Table 1), 4-Br (entry 10, 1j,
Table 1), and 4-NC (entry 11, 1k, Table 1). Compared with 1b, these
replacements all help in increasing the DP1 potency (entries 8–11,
Table 1) by up to slightly more than two-fold; however, only the
replacement with 4-CF3 has an effect on improving the selectivity
to above 780-fold in binding assays. In contrast, the use of MeSO2
as a replacement of the 4-chloro led to much less potent DP1
receptor antagonist 1l (entry 12, Table 1).
(TP-PRP) with a mean IC50 of 19.4 lM (5.77–38.8 lM, n = 16). This
represents a 7760-fold selectivity for DP1 over TP. Besides the
binding profiles, another main factor that effects compound selec-
tion is the stabilities of these DP1 antagonists under acidic condi-
tions. It was found that when treated with acid at low pH, the
substituted benzyl groups of most of the selective antagonists pre-
sented in Table 1 could be cleaved. For example, only 79% of 1b
could be recovered after 2 h incubation with 0.01 N of HCl at
37 °C. In comparison, under same conditions compound 1h was
found to be stable (99.6% of 1h remained after 6 h).
To conclude, we discovered that the introduction of a methyl
group to the benzylic position of the N-benzyl group in lead com-
pound 1a has a dramatic effect on improving binding selectivity for
DP1 over TP. Based on this discovery, we have synthesized a series
of potent and highly selective DP1 antagonists. Among them, com-
pound 1h has a Ki of 0.43 nM to DP1 in binding assay and an IC50 of
2.5 nM in DP1 functional assay. Its selectivity for DP1 over TP (the
most potent receptor after DP1) is over 780- and 7760-fold in bind-
ing and functional assays, respectively. The excellent overall prop-
erties of 1h warrant it for further preclinical development.
References and notes
1. For a review and references therein, see: Holgate, S. T. Clin. Exp. Allergy 1991, 21,
11.
2. Matsuoka, T.; Hirata, M.; Tanaka, H.; Takahashi, Y.; Murata, T.; Kabashima, K.;
Sugimoto, Y.; Kobayashi, T.; Ushikubi, F.; Aze, Y.; Yoshida, N.; Honda, Y.; Nagai,
H.; Narumiya, S. Science 2000, 287, 2013.
3. (a) Flower, R. J.; Harvey, E. A.; Kingston, W. P. Br. J. Pharmacol. 1976, 56, 229; (b)
Soter, N. A.; Lewis, R. A.; Corey, E. J.; Austen, K. F. J. Invest. Dermatol. 1983, 80,
115.
4. (a) Doyle, W. J.; Boehm, S.; Skoner, D. P. J. Allergy Clin. Immunol. 1990, 86, 924;
(b) Alving, K.; Matran, R.; Lundberg, J. M. Acta Physiol. Scand. 1991, 143, 93; (c)
Johnston, S. L.; Smith, S.; Harrison, J.; Ritter, W.; Hawarth, P. H. J. Allergy Clin.
Immunol. 1993, 91, 903.
5. (a) Woodward, D. F.; Hawley, S. B.; Williams, L. S.; Ralston, T. R.; Protzman, C.
E.; Spada, C. S.; Nieves, A. L. Invest. Ophthalmol. Vis. Sci. 1990, 31, 138; (b)
Woodward, D. F.; Nieves, A. L.; Friedlaender, J. J. Pharmacol. Exp. Ther. 1996, 279,
137.
6. For a review and references therein, see Kostenis, E.; Ulven, T. Trends Mol. Med.
2006, 12, 148.
7. Sturino, C. F.; Lachance, N.; Boyd, M.; Berthelette, C.; Labelle, M.; Li, L.; Roy, B.;
Scheigetz, J.; Tsou, N.; Brideau, C.; Cauchon, E.; Carriere, M.-C.; Denis, D.; Greig,
G.; Kargman, S.; Lamontagne, S.; Mathieu, M.-C.; Sawyer, N.; Slipetz, D.; O’Neill,
The replacement of the methyl group substituted at the ben-
zylic position of the N-benzyl group in 1b (R = Me) with fluorine
contained methyl group such as CF2H (entry 13, 1m, Table 1) or

L. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7462–7465
7465
G.; Wang, Z.; Zamboni, R.; Metters, K. M.; Young, R. N. Bioorg. Med. Chem. Lett.
2006, 16, 3043.
8. Sturino, C. F.; O’Neill, G.; Lachance, N.; Boyd, M.; Berthelette, C.; Labelle, M.; Li,
L.; Scheigetz, B.; Roy, J.; Tsou, N.; Aubin, Y.; Bateman, K. P.; Chauret, N.; Day, S.
H.; Levesque, J.-F.; Seto, C.; Silva, J. H.; Trimble, L. A.; Carriere, M.-C.; Denis, D.;
Greig, G.; Kargman, S.; Lamontagne, S.; Mathieu, M.-C.; Sawyer, N.; Slipetz, D.;
Abraham, W. M.; Jones, T.; McAuliffe, M.; Piechuta, H.; Nicoll-Griffith, D. A.;
Wang, Z.; Zamboni, R.; Young, R. N.; Metters, K. M. J. Med. Chem. 2007, 50, 794.
9. Unpublished results.
10. Li, L.; Beaulieu, C.; Guay, D.; Sturino, C.; Wang, Z. WO2004103970 2004, pp 35.