Discovery of novel S1P2 antagonists. Part 1: Discovery of 1,3-bis(aryloxy)benzene derivatives

Article abstract of DOI:10.1016/j.bmcl.2015.02.029

The structure-activity relationships of a novel series of sphingosine-1-phosphate receptor antagonists have been examined in detail. The initial hit compound 1 was modified through synthesis to improve its S1P2 activity. The synthesis of a series of analogs revealed that 1,3-bis(aryloxy)benzene derivatives, as represented by 22, are potent and selective S1P2 antagonists.

Full text of DOI:10.1016/j.bmcl.2015.02.029

Bioorganic & Medicinal Chemistry Letters 25 (2015) 1479–1482
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel S1P2 antagonists. Part 1: Discovery
of 1,3-bis(aryloxy)benzene derivatives
b
a
a
a
Kensuke Kusumi a, , Koji Shinozaki , Toshiya Kanaji , Haruto Kurata , Atsushi Naganawa ,
⇑
Kazuhiro Otsuki a, Takeshi Matsushita a, Tetsuya Sekiguchi a, Akito Kakuuchi a, Takuya Seko c
a Minase Research Institute, Ono Pharmaceutical Co., Ltd, 3-1-1 Sakurai, Shimamoto, Mishima, Osaka 618-8585, Japan
b Ono Pharma UK Ltd, MidCity Place, 71 High Holborn, London WC1V 6EA, United Kingdom
c Head Office, Ono Pharmaceutical Co., Ltd, 8-2, Kyutaromachi 1-chome, Chuo-ku, Osaka 541-8564, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The structure–activity relationships of a novel series of sphingosine-1-phosphate receptor antagonists
have been examined in detail. The initial hit compound 1 was modified through synthesis to improve
its S1P2 activity. The synthesis of a series of analogs revealed that 1,3-bis(aryloxy)benzene derivatives,
as represented by 22, are potent and selective S1P2 antagonists.
Received 13 January 2015
Revised 5 February 2015
Accepted 12 February 2015
Available online 23 February 2015
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
Sphingosine-1-phosphate receptor 2
Antagonist
G protein-coupled receptors
Sphingosine-1-phosphate (S1P) is
a
membrane-derived
antagonists could become useful drugs for the treatment of various
fibrotic diseases. Disappointingly, however, there have been very
few reports pertaining to the development of S1P2 antagonists.
We recently initiated a new discovery program with the aim of
identifying novel S1P2 antagonists. S1P2 was initially screened
against our in-house high throughput screening (HTS) collection
of 90,000 compounds, which resulted in the identification of sever-
al hit compounds. Following a period of hit exploration, we identi-
fied biphenyl ether 1 (shown in Fig. 1) as the initial lead compound
because it had the best antagonist activity of all of the hit com-
lysophospholipid that plays an important role in homeostasis.1,2
S1P exerts its biological effects through five different members of
the endothelial differentiation gene (EDG)/S1P family of
G
protein-coupled receptors (GPCRs), including S1P1 (EDG-1), S1P2
(EDG-5), S1P3 (EDG-3), S1P4 (EDG-6) and S1P5 (EDG-8).3 S1P1,
S1P2 and S1P3 are widely expressed in a variety of tissues and cell
types, whereas S1P4 and S1P5 are only expressed in a limited num-
ber of tissues, where they are independently responsible for per-
forming a diverse range of functions.4
Although a large number of synthetic compounds have been
reported as ligands for the S1P receptors, most of these compounds
are S1P1 agonists. FTY720 (Fingolimod/Gilenya; Novartis), which is
one of the most well-known S1P1 agonists,5 was recently approved
by the US Food and Drug Administration as a therapeutic agent for
multiple sclerosis. FTY720 is a potent agonist of S1P1, S1P3, S1P4
and S1P5, but shows no affinity for S1P2.6
S1P2 is a receptor at which S1P acts at a subnanomolar level.
S1P2 is mainly expressed in the heart, thyroid gland, lung, trachea
and liver, and it has been suggested that this receptor is involved in
a broad range of biological functions. JTE-013 is well known as an
S1P2 selective antagonist with an IC50 value in the range of 10–
20 nM.2 Numerous studies have been conducted to evaluate the
biological properties of JTE-013,7–10 and it is envisaged that S1P2
pounds (IC50: 1.2 l
M).11 Further structure activity relationship
(SAR) studies were then conducted based on compound 1 to obtain
more potent S1P2 antagonists.
Our initial SAR work focused on the diphenyl ether moiety of 1
to identify suitable positions for the introduction of further chemi-
cal modification. The tolerance of the different positions on the
diphenyl ether moiety towards modification was evaluated by
the introduction of a methyl group to the different positions of
the A- and B-rings. The results are shown in Table 1. The introduc-
tion of a methyl group to the different positions of the A-ring had
very little impact on the activity, with all of the methyl derivatives
(i.e., 3, 4 and 5) showing similar levels of activity. This result indi-
cated that the introduction of substituents to the terminal A-ring
could be well tolerated. In contrast, the introduction of a methyl
group at the 2- or 4-position of the B ring (i.e., 6 and 7) led to a sig-
nificant reduction in the antagonist activity, whereas the 5- and
6-methyl analogs 8 and 9 showed slightly improved activities
⇑
Corresponding author. Tel.: +81 75 961 1151; fax: +81 75 962 9314.
E-mail address: kusumi@ono.co.jp (K. Kusumi).
http://dx.doi.org/10.1016/j.bmcl.2015.02.029
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

1480
K. Kusumi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1479–1482
Table 2
H3C
H3C
OH
Effect of different substituents at the 5-position
H3C
H3C
H
N
O
N
HO
O
H
N
F
N
O
1
B
X
A
O
F
S1P2
IC50 = 1.2 µM
Compd
X
Binding assay IC50
(lM)
Ca2+ assay IC50
(lM)
Figure 1. Structure of hit compound 1.
1
H
Me
Cl
0.31
0.13
1.0
1.2
10
11
12
0.50
0.86
1.4
Table 1
OEt
1.7
Effect of introducing a methyl group at the different positions of the phenyl rings
*
H3C
HO
H3C
O
13
0.029
0.98
2
2'
A
H
N
3'
4'
F
N
O
B
5
O
4
6
The replacement of the fluorine group with a methylsulfonyl
group gave compound 15, which was 17-fold more potent (IC50
0.058 M) than 13. Furthermore, the corresponding sulfonamide
derivative 16 showed potent binding affinity as well as potent
antagonist activity, with IC50 values of 0.0041 and 0.021 M,
respectively. Carbamide 17 also showed good potency (IC50
Ca2+ assay IC50
(lM)
:
Compd
Substituent
l
2
3
4
5
6
7
8
9
None
40-Me
30-Me
20-Me
2-Me
4-Me
5-Me
6-Me
2.2
1.2
1.3
1.2
21
>25
0.82
1.5
l
:
0.0037 and 0.039 l
M). When the 400-carbamide moiety of 17 was
moved to the 300-position to yield 18, there was a significant
decrease in both the binding affinity and the antagonist activity
(IC50
: 0.013 and 0.15 lM, respectively). The corresponding
200-carbamide 19 showed similar levels of activity to 17 (IC50
:
0.049 and 0.088 lM, respectively). These results therefore suggest
compared with compound 1. We supposed that the 2- and
4-methyl group would twist the compound through steric repul-
sion against the urea group, which would cause the compounds
to adopt an unfavorable conformation and lose their activity.
Further investigation of the introduction of different substituents
at the 20-, 30-, 40-, 5- and 6-positions revealed that the 5-position
was the most tolerant of these positions towards the introduction
of different substituents (data not shown). Based on this result, we
focused our remaining SAR studies on investigating the effect of
introducing different substituents at the 5-position on the B-ring.
All of the compounds in the current study were synthesized as
the 40-F derivative because the initial lead compound 1 showed
slightly better activity than the corresponding 40-H derivative 2.
A summary of the SAR work at the 5-position is shown in
Table 2. The 40-fluoro-5-methyl derivative 10 showed slightly bet-
ter binding affinity and antagonist activity than 1, with IC50 values
that the 200 and 400-positions appeared to be superior to the
300-position in terms of achieving both good binding affinity and
antagonist activity. Meanwhile, mono-methyl and di-methyl car-
bamide derivative of 17 (i.e., 20 and 21) showed similar levels of
antagonistic activity (IC50: 0.027 and 0.043 lM, respectively) and
binding affinity (IC50: 0.003 and 0.0088, respectively) to 17. These
results suggest that the amide protons of 17 are not necessary to
show higher levels of binding affinity and antagonist activity.
The replacement of ring C with a heterocyclic ring was also
investigated. The 4-pyridine derivative 22 showed good binding
and antagonistic activities, with IC50 values of 0.045 and
0.093
ing 3-pyridine derivative 23 and 2-pyridine derivative 24 showed
much lower antagonistic activities (IC50: 0.21 and 0.26 M, respec-
lM, respectively. Disappointingly, however, the correspond-
l
tively) and binding affinities (IC50: 0.94 and 0.39, respectively) than
22 did. Additionally, none of the derivatives in this series showed
activity against any of the other EDG receptors (i.e., EDG-2(LPA1),
3(S1P3) and 4(LPA2)).
of 0.13 and 0.50
derivative 11 shows slightly weaker binding affinity (IC50
1.0 M). The introduction of an alkyl ether substituent at the
5-position led to a reduction in the potency. For example, the ethyl
ether 12 gave IC50 values of 1.7 and 1.4 M for its binding affinity
l
M, respectively, whereas the 40-fluoro-5-chloro
:
l
Table 4 shows the physicochemical properties of compound 22,
which was found to be an S1P2 selective antagonist because it did
not exhibit any activity towards any of the other S1P receptors
tested in the current study (i.e., S1P1 and S1P3-5). Compound 22
l
and antagonist activity, respectively. Interestingly, however, the
introduction of a 4-fluorophenoxy group at the 5-position resulted
in the ‘symmetrical’ compound 13, which showed a 10-fold
increase in binding affinity compared with 1 and slightly better
also showed good Caco-2 permeability (A–B: 6.0 ꢀ 10ꢁ6 cm sꢁ1
,
B–A: 1.2 ꢀ 10ꢁ6 cm sꢁ1), very poor solubility (<5
lM in pH 6.5
phosphate buffer), and high plasma protein binding (human:
potent antagonistic activity (IC50: 0.029 and 0.98 lM, respectively).
99.8%, rat: 99.9%).
Although 13 showed excellent binding affinity, its antagonistic
activity was still poor, so we proceeded to conduct a series of addi-
tional modifications in an attempt to optimize the substitution pat-
tern on the ‘third’ phenyl moiety (ring C in Table 3).
Our SAR work at the ring C started by introducing diverse sub-
stituents to the 400 position on the ring C. Although introduction of
most substituents had only a little impact on the antagonistic
activity (data not shown), we found that the cyano-derivative of
All of the compounds tested in the current study were synthe-
sized in a similar manner. A representative example of the synthetic
route used is shown in Scheme 1 for the synthesis of the methylsul-
fone derivative 15. Briefly, commercially available N-protected
piperidone 25 was treated with cerium chloride and
2-ethylbutane-1-yl magnesium bromide to yield alcohol 26, which
was subjected to a palladium-catalyzed hydrogenation to afford
hydroxypiperidine 27 as common intermediate. The second inter-
mediate, compound 34, was generated from 3-bromo-5-hydroxy
benzoic acid (28). Protection of the phenol in 28 with MOMCl, fol-
lowed by the methylation of the carboxylic acid moiety with
13 (compound 14 in Table 3) was 10-fold more potent (IC50
:
0.084 M) than 13. Thus, we focused on the introduction of the
l
electro-deficient substituents. The results of the electro-deficient
substituted compounds are shown in Table 3.

K. Kusumi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1479–1482
1481
Table 3
Optimization of the substituent on the third aryl ring
H3C
HO
H3C
H
N
N
O
B
O
A
O
F
C
4''
2''
3''
Compd
Substituent
Ca2+ assay IC50
EDG3(S1P3)
(
l
M)
Binding assay IC50
EDG5(S1P2)
(lM)
EDG2(LPA1)
EGD4(LPA2)
EDG5(S1P2)
13
14
15
16
17
18
19
20
21
22
23
24
400-F
>25
>25
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>25
>25
>25
>25
>25
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>25
>25
>25
>25
>26
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>0.8
>25
>25
>25
0.98
0.029
N.T.
400-CN
0.084
0.058
0.021
0.039
0.15
0.088
0.027
0.043
0.093
0.21
400-SO2Me
400-SO2NH2
400-CONH2
300-CONH2
200-CONH2
400-CONMe
400-CONMe2
400-aza
0.045
0.0041
0.0037
0.013
0.049
0.003
0.0088
0.045
0.94
300-aza
200-aza
0.26
0.39
Table 4
Profile of compound 22
Binding assay IC50
(
lM)
S1P1
>10
S1P2
0.045
S1P3
>10
S1P4
>10
S1P5
>10
Solubility in phosphate buffer (pH 6.8)
<5
Caco-2 (10ꢁ6 cm/s) (A–B/B–A)
6.0/1.2
Protein binding (human/rat)
99.8%/99.9%
l
M
H3C
H3C
O
OH
H3C
H3C
OH
(a)
(b)
N
O
N
O
NH
O
O
25
Br
26
27
HO2C
MeO2C
Br
O
MeO2C
O
(c)
(d)
(e)
(f)
F
OH
28
O
29
O
30
O
CH3
CH3
HO
H3C
H3C
HO2C
O
OCN
(g)
O
O
H
N
N
O
(h)
(j)
F
F
O
O
O
F
O
CH3
CH3
O
O
33
CH3
31
32
H3C
H3C
OH
HO
H3C
H3C
H
N
N
O
H
N
(i)
N
O
O
F
O
F
O
OH
CH3
O
15
34
S
O
Scheme 1. Synthesis of compound 15. Reagents and conditions: (a) 1-bromo-2-ethyl butane, Mg, CeCl3, THF, 0 °C–rt, 8 h, (b) H2, Pd/C, AcOEt, rt, 1 h, (c) MOMCl, K2CO3, DMF,
70 °C, 15 h then, NaOHaq, MeOH, rt, o.n, (d) K2CO3, MeBr, tBuOMe, rt, 1 h, (e) 4-fluorophenol, K3PO4, Pd2(dba)3, tBuXPhos, toluene, 100 °C, 46 h, (f) NaOHaq, MeOH, rt, 15 h, (g)
DPPA, AcOEt–THF, rt, 1.5 h, then, toluene, 100 °C, 1 h, (h) 27, THF, rt, o.n, (i) NaI, TMSCl, CH3CN, ꢁ10 °C, 3.5 h, (j) 1-bromo-4-methylsulfonylbenzene, picolinic acid, CuI, K3PO4,
DMSO, 100 °C, 2 h.
MeBr under basic conditions gave compound 29, which was cou-
pled with 4-fluorophenol to afford biphenyl ether 30. Hydrolysis
of the methyl ester under basic conditions gave the corresponding
carboxylic acid 31, which was treated with diphenylphosphoryl
azide (DPPA) to give isocyanate 32. The subsequent reaction of 32
with 27 gave the corresponding urea 33, which was treated with

1482
K. Kusumi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1479–1482
NaI and TMSCl to allow for the removal of the methoxymethyl
protecting group to afford another common intermediate phenol
34. Methylsulfone derivative 15 was synthesized by the
Ullmann coupling reaction of compound 34 with 1-bromo-4-
methylsulfonylbenzene in the presence of picolinic acid. The other
derivatives 13–24 were also synthesized in the same way using the
corresponding aryl halide reagents. Derivatives 1–12 were
synthesized from the corresponding 3-bromobenzoic acid deriva-
tives and phenols instead of 3-bromo-5-hydroxy benzoic acid and
4-fluorophenol.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2015.02.029.
References and notes
1. Victoria, A. Blaho; Timothy Hla Chem. Rev. 2011, 111, 6299.
2. Victoria, A. Blaho; Timothy Hla J. Lipid Res. 2014, 55, 1596.
3. Spiegel, S.; Milstien, S. Biochim. Biophys. Acta 2000, 1484, 107.
4. Chun, J.; Hla, T.; Lynch, K. R.; Spiegel, S.; Moolenaar, W. H. Pharmacol. Rev. 2010,
62, 579.
5. Brinkmann, V.; Billich, A.; Baumruker, T.; Heining, P.; Schmouder, R.; Francis,
G.; Aradhye, S.; Burtin, P. Nat. Rev. Drug Discov. 2010, 9, 883.
6. Brinkmann, V.; Davis, M. D.; Heise, C. E.; Albert, R.; Cottens, S.; Hof, R.; Bruns,
C.; Prieschl, E.; Baumruker, T.; Hiestand, P.; Foster, C. A.; Zollinger, M.; Lynch, K.
R. J. Biol. Chem. 2002, 277, 21453.
7. Osada, M.; Yatomi, Y.; Ohmori, T.; Ikeda, H.; Ozaki, Y. Biochem. Biophys. Res.
Commun. 2002, 299, 483.
8. Sanchez, T.; Skoura, A.; Wu, M. T.; Casserly, B.; Harrington, E. O.; Hla, T.
Arterioscler. Thromb. Vasc. Biol. 2007, 27, 1312.
9. Salomone, S.; Potts, E. M.; Tyndall, S.; Ip, P. C.; Chun, J.; Brinkmann, V.; Waeber,
C. Br. J. Pharmacol. 2008, 153, 140.
10. Roviezzo, Fiorentina.; Brancaleone, Vincenzo.; De Gruttola, Luana.; Vellecco,
Valentina.; Bucci, Mariarosaria.; D’Agostino, Bruno.; Cooper, Dianne.;
Sorrentino, Raffaella.; Perretti, Mauro.; Cirino, Giuseppe. J. Pharmacol. Exp.
Ther. 2011, 337, 830.
In summary, we have explored the modification of the diphenyl
ether moiety of the lead compound 1 and found that the introduc-
tion of substitutions at both the 40 and 5-positions of the phenyl
rings led to significant improvements in the S1P2 potency. The
introduction of
a phenyl-oxy or pyridyl-oxy group at the
50-position led to significant improvements in both the binding
affinity and antagonistic activity. The representative compound
22 also showed excellent S1P2 selectivity against other S1P recep-
tors, although this compound did show poor physicochemical
properties. Further optimization of this series is currently under-
way in our laboratory and will be reported in due course.
11. Spectra data of lead compound
1
are shown below: 1H NMR(300 MHz,
Acknowledgements
DMSO-d6) d ppm 0.79 (t, J = 7.50 Hz, 6H) 1.31 (m, 11H) 3.10 (m, 2H) 3.73 (m,
2H) 4.09 (s, 1H) 6.53 (m, 1H) 7.03 (dd, J = 9.00, 4.50 Hz, 2H) 7.21 (m, 5H) 8.47
(s, 1H); 13C NMR(125 MHz, DMSO-d6) d ppm 10.7, 26.8, 34.7, 36.8, 40.1, 46.1,
68.5, 108.9, 111.1, 114.2, 116.4, 120.6, 129.5, 142.5, 152.7, 154.6, 157.2,
159.1; MS (ESI, Pos. 20 V) 829 (2M+H)+, 415 (M+H)+; LC–MS purity (ELSD) 100
area%.
The authors thank Dr. T. Maruyama for useful discussion during
the preparation of this Letter. Also authors grateful to Mr. Y. Ochi,
Dr. R. Nishizawa and Dr. Inagaki for their useful comments for this
Letter.