Synthesis and evaluation of alkoxy-phenylamides and alkoxy-phenylimidazoles as potent sphingosine-1-phosphate receptor subtype-1 agonists

Article abstract of DOI:10.1016/j.bmcl.2008.11.072

In the design of potent and selective sphingosine-1-phosphate receptor agonists, we were able to identify two series of molecules based on phenylamide and phenylimidazole analogs of FTY-720. Several designed molecules in these scaffolds have demonstrated selectivity for S1P receptor subtype 1 versus 3 and excellent in vivo activity in mouse. Two molecules PPI-4621 (4b) and PPI-4691 (10a), demonstrated dose responsive lymphopenia, when administered orally.

Full text of DOI:10.1016/j.bmcl.2008.11.072

Bioorganic & Medicinal Chemistry Letters 19 (2009) 369–372
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of alkoxy-phenylamides and alkoxy-phenylimidazoles
as potent sphingosine-1-phosphate receptor subtype-1 agonists
b
a
b
b
Ghotas Evindar a, , Sylvie G. Bernier , Malcolm J. Kavarana , Elisabeth Doyle , Jeanine Lorusso ,
Michael S. Kelley a, Keith Halley b, Amy Hutchings b, Albion D. Wright b, Ashis K. Saha a,
Gerhard Hannig b, Barry A. Morgan a, William F. Westlin b
*
a Department of Medicinal Chemistry, Praecis Pharmaceuticals Incorporated, 830 Winter Street, Waltham, MA 02451, USA
b Department of Preclinical Research, Praecis Pharmaceuticals Incorporated, 830 Winter Street, Waltham, MA 02451, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
In the design of potent and selective sphingosine-1-phosphate receptor agonists, we were able to identify
two series of molecules based on phenylamide and phenylimidazole analogs of FTY-720. Several designed
molecules in these scaffolds have demonstrated selectivity for S1P receptor subtype 1 versus 3 and excel-
lent in vivo activity in mouse. Two molecules PPI-4621 (4b) and PPI-4691 (10a), demonstrated dose
responsive lymphopenia, when administered orally.
Received 5 September 2008
Revised 18 November 2008
Accepted 20 November 2008
Available online 24 November 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
S1P
S1P agonist
EDG1
FTY-720
Sphingosine-1-phosphate
Alkoxy-phenylamides
Alkoxy-phenylimidazoles
The discovery of the sphingosine-1-phosphate (S1P) receptor
class of G-protein coupled receptors (GPCRs) has opened up excit-
ing and highly competitive areas of research and development. Sig-
therapeutic properties and reduce the side effect profile of non-
selective S1P receptor agonists like FTY-720.
Our effort to explore new chemical entities in the S1P agonist
area started with investigations of S1P and FTY-720 based analogs.
A preliminary exploration of S1P related structures has demon-
strated that an amide insertion in the S1P structure is well toler-
ated (Fig. 1a).6 Most recently, Clemens and co-workers have
demonstrated this by replacing the ethylene component of FTY-
720 with an amide bond to allow for active molecules across S1P
receptor subtypes 1, 3, 4, and 5 (Fig. 1b).7 We chose the amide in-
serted analog of FTY-720 as an entry point for medicinal chemistry
and determination of a structure–activity relationship (SAR). With
a simple synthetic strategy in hand (Scheme 1), one could envision
the use of an aniline moiety to take advantage of the amino acid
chiral pool as a potential agonist head-piece (Fig. 2) to establish
a rapid SAR of the region. This allowed for utility of a mono alcohol
species to avoid issues with mono phosphate synthesis from pro-
chiral diols. We envisioned introduction of a phenolic oxygen at
the lipophilic tail section to facilitate convergent synthesis and
expedited tail modifications and thus it was important to gain a
thorough understanding of the effects this modification would
exert on the in vivo biological activity of these analogs. The ether
insertion would allow for the utility of a commercial library of
naling through members of this cluster of five receptors (S1P1–5
)
with the endogenous natural ligand S1P can induce multiple
effects on cardiovascular and immune system function and other
yet poorly defined effects on additional physiological systems.1
FTY-720 is a new molecular entity undergoing clinical evalua-
tion. This prodrug is currently in phase III clinical studies for multi-
ple sclerosis and has completed clinical trials in solid organ
transplant rejection prevention.2 The prodrug FTY-720, upon
in vivo phosphorylation, converts to a potent non-selective S1P
receptor agonist (S1P1,3–5) that has profound immunomodulatory
activity through direct modulation of in lymphocyte trafficking.3
This immunomodulatory activity is due to triggering of S1P recep-
tor subtype 1 (S1P1) signaling cascades.4 In contrast, it is postu-
lated that agonists of S1P receptor subtype 3 (S1P3) have been
associated with deleterious side effects including bradycardia.5
We therefore sought to discover potent S1P1 receptor agonists with
low potency on S1P3 (S1P3-sparing) in order to retain the positive
* Corresponding author. Tel.: +1 781 795 4423.
E-mail address: ghotas.x.evindar@gsk.com (G. Evindar).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.11.072

370
G. Evindar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 369–372
Table 1
a
b
NH2
Percent lymphopenia obtained upon 10 mg/kg oral (PO) administration of the alcohol
O
at 6 h post-dosing in mice11
O
P
HO OH
OH
H2N
O
H
NH2
R1
Me
OH
H
O
O
N
N
P
C7H15
HO OH
O
X
Alcohol
X
R1
Percent lymphopenia
NH2
O
5a
4a
4b
4c
4d
–CH2–
–O–
–CH2O–
–CH2O–
–CH2O–
H
H
H
2-F
3-F
36
72
78
65
66
OH
O
P
HO OH
NH2
H
O
N
O
P
OH
HO
O
Figure 1. (a) S1P and its corresponding amide analog; (b) phospho-FTY-720 and its
corresponding amino acid based analog.
based on synthetic (Scheme 1). Alkylation of the hydroxyl group
of a substituted aminophenol 1 was achieved using alkyl bromide
and a catalytic amount of NaI in the presence of either Cs2CO3 in
DMF (60 °C) or KOtBu in acetone (50 °C). The amino group of the
desired intermediate was then acylated with the desired Boc-pro-
tected amino acid using either N-ethylcarbodiimide (EDC),
1-hydroxybenzotriazole (HOBt), and N,N-diisopropylethylamine
(DIPEA) or O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate (HATU) and DIPEA. The final compound was
obtained in good yields from Boc deprotection of the intermediate
with 30% trifluoroacetic acid (TFA).
R
R
NH2
NH2
i or ii
iii or iv
HO
Alkyl O
1
2
Boc
NH
R2
NH2
H
R3
Alkyl O
H
N
R3
Alkyl O
R2
R1
N
v
OH
n
R1
n
OH
Based on the short synthetic strategy (Scheme 1), commercial
availability of the starting material, and in vivo lymphopenia, a
preliminary SAR was obtained as illustrated in Figure 1. This SAR
demonstrated that the S-chirality of the amine was essential for
O
O
3
4
Scheme 1. Reagents: (i) alkyl bromide, Cs2CO3, NaI, DMF; (ii) alkyl bromide, KOtBu,
NaI, acetone; (iii) N-Boc-amino acid, EDC, HOBt, DIPEA, CH2Cl2; (iv) N-Boc-amino
acid, HATU, DIPEA, DMF; (v) TFA, CH2Cl2.
compound activity with
a-methyl-serine providing better in vivo
activity than serine, homo-serine or threonine.8 Incorporation of
an oxygen in the alkyl tail section of the molecule provided better
in vivo activity in comparison to the des-oxy analog (Table 1). Fur-
ther investigation of alkyl ether tail modifications (C5 through C10)
provided moderate to good in vivo activity with C8 (4b, PPI-4621)
demonstrating optimal in vivo activity. Fluorine substitution on
the aromatic linker was the only tolerated substitution.
head-piece
linker
NH2
R3
R1
H
tail
N
OH
p
R2
Our next approach was modification of the amide linker region
to apply a rigid imidazole ring to explore the SAR relative to the
azole series of molecules. The designed compounds were synthe-
sized as described in Scheme 2. Substituted phenols were alkylated
with the appropriate alkyl bromide using KOtBu in acetone and a
catalytic amount of NaI at 50 °C, or in a microwave at 80 °C using
KOtBu in THF. Friedel-Crafts acylation of the corresponding phenyl
ether provided the bromoacetophenone precursor. Reaction of the
bromoacetophenone with N-protected-amino acid provided the
O
X
n
R4
NH2
R1
R3
p
OH
R2
X
n
R4
HO
R1
O
i or ii
O
iii
Figure 2. Structure–activity relationship study based on in vivo efficacy: p = 0 or
1; R1 and R2 were selected from H or Me (chiral center S or R); R3 or R4 are
selected from H, F, Cl, Me, OMe, or CO2Me, X is selected from CH2 or O; and
n = 1–6.
R2
R2
Br
R1
R1
O
6
7
8
substituted 4-aminophenols or precursors to develop an extensive
SAR around both the linker and lipophilic tail sections. Further-
more, we envisioned replacement of the amide bond with a more
rigid imidazole ring (Fig. 2) as a potential alternative scaffold with
the goals of establishing SAR around the linker section and ulti-
mately achieving an active S1P1 agonist with in vivo potency and
a moderate degree of selectivity for S1P receptor subtype 1 over 3.
In pursuit of obtaining a preliminary SAR in phenylamide deriv-
atives of FTY-720 we have generated a number of compounds
O
O
R2
vi R2
Boc
N
iv, v
NH2
Me
OH
NH
Me
OH
N
R1
R1
NH
NH
10
9
Scheme 2. Reagents and conditions: (i) alkylbromide, KOtBu, NaI, acetone; (ii)
alkylbromide, KOtBu, THF, microwaved, 45 min, 80 °C; (iii) 1—bromoacetyl bro-
mide, AlCl3, CH2Cl2; (iv) N-Boc-amino acid, Cs2CO3, DMF; (v) AcONH4, toluene or
xylenes, reflux; (vi) TFA, CH2Cl2.

G. Evindar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 369–372
371
the change from amide to imidazole allowed for SAR information
transfer to obtain potent S1P agonists. Compounds 12a–12e dem-
onstrated weaker but similar binding activity at S1P3, allowing for
moderate to good selectivity of eight- to 42-fold. Overall, relatively
weaker binding activity was observed for the amide series at
Boc
R2
Boc
R2
H2N
R2
NH
R1
NH
R1
R1
O
OH
i
ii
O
OEt
O
O
OH
P
P
HO
EtO
11
12
receptor subtypes 4 and 5. The [35S]GTP
cS functional assay was
Scheme 3. Reagents: (i) diethyl phosphorochloridate, Et3N, CH2Cl2; (ii) TMSBr,
used to further investigate the receptor subtype selectivity of both
series of molecules. The functional assay binding activity was mea-
sured as described by Davis and co-workers.10 Agonist 12e was
observed to be three times more potent than S1P itself with a
relatively moderate selectivity of eightfold for S1P1 over S1P3
(Table 4). Fluoro-substitution on the phenyl ring appeared to have
little or no effect on the agonist activity or selectivity. A further
three- to fivefold improvement in S1P1 receptor activity was
observed when the amide group was replaced with an imidazole
moiety (compounds 12d and 12e). Even though the S1P3 activity
for the amide and imidazole analogs remained the same, the im-
proved activity at S1P1 for the corresponding imidazole analogs
led to three- to fivefold improvement in agonist selectivity for
S1P1 over S1P3. Therefore, based on the functional assay against
human S1P1 and S1P3 receptors, both amide and imidazole series
of the S1P agonist demonstrated excellent S1P1 receptor agonist
activity with moderate to good overall selectivity for S1P receptor
subtypes 1 over 3.
CH2Cl2.
amino acid ester intermediate which, upon intramolecular cycliza-
tion in the presence of excess ammonium acetate, produced the
desired phenylimidazole. The phenylimidazole was either depro-
tected to remove the Boc group using 30% TFA in CH2Cl2, or was
phosphorylated as illustrated in Scheme 3.
The imidazole analogs of FTY-720 showed excellent in vivo
activity analogous to the corresponding amide analogs (Table 2).
The C8 analog (10a, PPI-4691) showed slightly better in vivo activ-
ity than the C7 analogs. Similar to the amide counter-part, only flu-
oro substitution was tolerated on the phenyl group.
In order to determine the binding activity and compound selec-
tivity against human S1P1 and S1P3 receptors, the desired phos-
phates were synthesized based on the synthetic strategy
illustrated in Scheme 3. Reaction of Boc-protected amino-alcohol
with excess diethyl chlorophosphate in the presence of triethyl-
amine gave phospho-triester 11 which upon treatment with excess
bromotrimethylsilane afforded the desired final phosphate 12.
Binding activity of S1P and synthesized phosphate agonists
were measured using a modified version of the [33P]S1P binding
assay described by Davis and co-workers.9 S1P and compound
12a showed similar binding activity at S1P1 as reported in Table
3. Compounds 12b–12e showed similar binding activity at S1P1
with two- to threefold improvement over 12a. This demonstrated
The lead compounds 4b (PPI-4621) and 10a (PPI-4691) were
further investigated for in vivo dose response when orally
administered at doses between 0.3 and 10 mg/kg. Both compounds
showed excellent lymphopenia at 10 and 3 mg/kg (Fig. 3). At
0.3 mg/kg both compounds showed little or no response while at
1 mg/kg both compounds showed good to moderate response with
Table 4
[
35S]GTP
c
S binding activity on human S1P1,3–5 receptor subtypes
2
1
O
Table 2
3
C8H17
NH2
Me
O
Percent lymphopenia obtained upon 10 mg/kg oral (PO) administration of the alcohol
at 6 h post-dosing in mice9
Q
R1
O
P
O
R2
OH
HO
NH2
Me
N
Agonist
Q
R1
hS1P1
hS1P3
S1P3/S1P1
R1
OH
NH
EC50 (nM)
EC50 (nM)
S1P
12a
12b
12c
12d
12e
—
—
H
2-F
3-F
H
2-F
7.9
2.4
1.1
2.1
0.55
0.45
3
0.4
6.9
9.2
4.7
26.9
24
Alcohol
R1
R2
Percent lymphopenia
Amide
Amide
Amide
Imidazole
Imidazole
16.3
10.1
9.8
14.8
10.8
10a
10b
10c
10d
H
H
2-F
3-F
C8H17
C7H15
C8H17
C8H17
83
74
70
81
Table 3
[
33P]S1P binding activity on human S1P1,3–5 receptor subtypes
2
1
O
3
C8H17
NH2
Me
O
Q
R1
O
P
OH
HO
Agonist
Q
R1
hS1P1
hS1P3
hS1P4
hS1P5
S1P3/S1P1
IC50 (nM)
IC50 (nM)
IC50 (nM)
IC50 (nM)
S1P
12a
12b
12c
12d
12e
—
—
H
2-F
3-F
H
0.47
0.73
0.14
0.35
0.26
0.35
0.66
5.9
5.9
3.6
5.3
5.8
2.4
4.0
3.5
1.8
1.1
1.9
1.1
8.8
6.9
3.2
1.4
1.1
—
8.1
42
10.2
20.3
16.5
Amide
Amide
Amide
Imidazole
Imidazole
2-F

372
G. Evindar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 369–372
Kappos, L.; Antel, J.; Comi, G.; Montalban, X.; O’Connor, P.; Polman, C. H.; Haas,
14
12
10
8
T.; Korn, A. A.; Karlsson, G.; Radue, E. W.FTY720 D2201 Study Group N. Eng. J.
Med. 2006, 355, 1124; (c) Khoury, S. J. Watch Neurol. 2006, 1, 1.
3. (a) Sugiyama, A.; Yatomi, Y.; Ozaki, Y.; Hashimoto, K. Cardiovasc. Res. 2000, 46,
119; (b) 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; (c) Mandala, S.; Hajdu, R.;
Bergstrom, J.; Quackenbush, E.; Xie, J.; Milligan, J.; Thornton, R.; Shei, G.-J.;
Card, D.; Keohane, C. A.; Rosenbach, M.; Hale, J.; Lynch, C. L.; Rupprecht, K.;
Parsons, W.; Rosen, H. Science 2002, 296, 346.
6
4
2
4. (a) Matloubian, M.; Lo, C. G.; Cinamon, G.; Lesneski, M. J.; Xu, Y.; Brinkmann, V.;
Allende, M. L.; Proia, R. L.; Cyster, J. G. Nature 2004, 427, 355; (b) Goetzl, E. J.;
Graler, M. H. J. Leukocyte Biol. 2004, 76, 30.
0
5. (a) Forrest, M.; Sun, S.-Y.; Hajdu, R.; Bergstrom, J.; Card, D.; Hale, J.; Keohane, C.
A.; Meyers, C.; Milligan, J.; Mills, S.; Nomura, N.; Rosenbach, M.; Shei, G.-J.;
Singer, I. I.; Tian, M.; West, S.; White, V.; Xie, J.; Rosen, H.; Proia, R.; Doherty, G.;
Mandala, S. J. Pharmacol. Exp. Ther. 2004, 309, 758; (b) Sanna, M. G.; Liao, J.; Jo,
E.; Alfonso, C.; Ahn, M.-Y.; Peterson, M. S.; Webb, B.; Lefebvre, S.; Chun, J.; Gray,
N.; Rosen, H. J. Biol. Chem. 2004, 279, 13839.
Compounds 10a (left) and 4b(right) (mg/kg)
Figure 3. Dose responsive lymphopenia for lead compounds 10a (PPI-4691) and 4b
(PPI-4621) relative to the vehicle.
6. Im, D.-S.; Clemens, J.; Macdonald, T. L.; Lynch, K. R. Biochemistry 2001, 40,
14053.
7. Clemens, J. J.; Davis, M. D.; Lynch, R. K.; Macdonald, T. L. Bioorg. Med. Chem. Lett.
2003, 13, 3401.
8. In order to induce in vivo activity, that is, lymphopenia, mice were treated with
the indicated compounds described herein as follow: Male C57Bl/6 mice were
divided into groups of three. A control group received the 3% BSA vehicle only.
compound 10a showing better in vivo activity. Overall, both com-
pounds 4b and 10a demonstrated excellent dose responsiveness
when administered orally at doses between 0.3 and 10 mg/kg.
In summary, we have generated a robust SAR around two differ-
ent scaffolds of S1P1 receptor agonists. We have demonstrated
excellent S1P1 receptor binding potency with moderate to good
selectivity in both series of molecules, with the phenyl-imidazole
series of molecules giving further improvement in receptor selec-
tivity and in vivo potency. Our lead molecules from these series,
4b (PPI-4621) and 10a (PPI-4691), demonstrated that both alk-
oxy-phenylamide and alkoxy-phenylimidazole analogs have excel-
lent in vivo oral activity. These new chemical entities provide a
solid foundation for further structural modifications and explora-
tion to enhance agonist selectivity.
The other groups received
administered orally (PO). After 6 h, the mice were anesthetized with isoflurane
and approximately 250
l of blood was removed from the retroorbital sinus
a specified dose of test compound in vehicle
l
and collected in an EDTA microtainer, mixed with an anticoagulant and placed
on a tilt table until complete blood count (CBC) analysis.
9. FTY-720 was used as the control in these experiments. Upon 1.0 mg/kg
oral (PO) administration of FTY-720 generally 80% lymphopenia was
observed.
10. The IC50 of a compound was determined using a modified version of the
[
33P]sphingosine-1-phosphate binding assay described by Davis, M. D.;
Clemens, J. J.; Macdonald, T. L.; Lynch, R. K. J. Biol. Chem. 2005, 280, 9833’ as
follows: For the binding assay, [33P]sphingosine-1-phosphate (obtained from
American Radiolabeled Chemicals, Inc) was added to membranes in 200
96-well plates with assay concentrations of 2.5 pM
33P]sphingosine-1-
phosphate, 4 mg/ml BSA, 50 mM Hepes, pH 7.5, 100 mM NaCl, 5 mM MgCl2,
and 5 g of protein. Binding was performed for 60 min at room temperature
with gentle mixing and terminated by collecting the membranes onto GF/B
filter plates. After drying the filter plates for 10 min, 50 l of Microscint 40 was
ll in
[
l
References and notes
l
1. (a) Spiegel, S.; Milstein, M. Nat. Rev. Mol. Cell Biol. 2003, 4, 397; (b) Spiegel, S.;
Milstein, S. Biochem. Soc. Trans. 2003, 31, 1216.
2. (a) Salvadori, M.; Budde, K.; Charpentier, B.; Klempnauer, J.; Nashan, B.;
Pallardo, L. M.; Eris, J.; Schena, F. P.; Eisenberger, U.; Rostaing, L.; Hmissi, A.;
Aradhye, S.FTY720 0124 Study Group. Am. J. Transplant. 2006, 6, 2912; (b)
added to each well, and filter-bound radionuclide was measured on a Packard
Top Count. Non-specific binding was defined as the amount of radioactivity
remaining in the presence of excess of unlabeled S1P.
11. Davis, M. D.; Clemens, J. J.; Macdonald, T. L.; Lynch, R. K. J. Biol. Chem. 2005, 280,
9833.