Novel immunomodulators based on an oxazolin-2-one-4-carboxamide scaffold

Article abstract of DOI:10.1016/j.bmcl.2011.10.088

A series of oxazolidin-2-one-4-carboxylic amide compounds (1a-f) were designed and synthesized as the non-phosphate S1P₁ receptor agonists. The single crystal of 1e was prepared and solved to elucidate the structure of 1a-f. EC₅₀ of 1a-d were about 1.1-3.6 μM in S1P₁ Redistribution assay, and their cytotoxicity was 8-40-fold lower than FTY720. Though its S1P₁ agonist activities in vitro were about 1000-folds weaker than (S)-FTY720-P, at a dose of 10 mg/Kg, the immunosuppressive effects of 1a were comparable to FTY720. So oxazolidin-2-one-4-carboxylic amide derivatives were found as potential immunomodulator, compound 1a could be considered as a lead compound, rational modifications of 1a are anticipated using medicinal chemistry techniques and molecular modeling to obtain analogs with higher affinity and better clinical therapeutic properties.

Full text of DOI:10.1016/j.bmcl.2011.10.088

Bioorganic & Medicinal Chemistry Letters 22 (2012) 553–557
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel immunomodulators based on an oxazolin-2-one-4-carboxamide scaffold q
⇑
⇑
Xinhua He , Lili Wang, Zhibing Zheng, Junhai Xiao, Wu Zhong, Song Li
Beijing Institute of Pharmacology & Toxicology, No.27 Taiping Road, Beijing 100850, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 May 2011
Revised 22 September 2011
Accepted 26 October 2011
Available online 6 November 2011
A series of oxazolidin-2-one-4-carboxylic amide compounds (1a–f) were designed and synthesized as the
non-phosphate S1P1 receptor agonists. The single crystal of 1e was prepared and solved to elucidate the
structure of 1a–f. EC50 of 1a–d were about 1.1–3.6 l
M in S1P1 RedistributionÒ assay, and their cytotox-
icity was 8–40-fold lower than FTY720. Though its S1P1 agonist activities in vitro were about 1000-folds
weaker than (S)-FTY720-P, at a dose of 10 mg/Kg, the immunosuppressive effects of 1a were comparable
to FTY720. So oxazolidin-2-one-4-carboxylic amide derivatives were found as potential immunomodula-
tor, compound 1a could be considered as a lead compound, rational modifications of 1a are anticipated
using medicinal chemistry techniques and molecular modeling to obtain analogs with higher affinity and
better clinical therapeutic properties.
Keywords:
S1P1 receptor
Immunosuppressive agent
Oxazolidin-2-one-4-carboxylic amide
Ó 2011 Elsevier Ltd. All rights reserved.
The need to develop less toxic and more effective immunosup-
pressants is essential owing to the worldwide growth of the autoim-
mune disease population. The novel immunosuppressant FTY720,1,2
unlike conventional immunosuppressive agents, does not inhibit T
or B cell activation or proliferation, nor their effector function.3,4
FTY720 is a prodrug, phosphorylated in vivo and transformed to
its active metabolite FTY720-phosphate (FTY720-P) by sphingosine
kinase 2.5 FTY720-P acts as a high-affinity agonist at S1P1-R in lym-
phocytes, thereby inducing aberrant internalization of the receptor,
rendering the lymphocyte cells unresponsive to sphingosine 1-
phosphate (S1P, Fig. 1), and depriving them of an obligatory signal
to egress from lymphoid organs. Consequently, FTY720 causes the
rapid and dramatic depletion of circulating lymphocytes, which re-
duces and/or inhibits the infiltration of lymphocytes into graft sites
and inflamed tissues,6 thereby exerting powerful immunosuppres-
sive activity. This new mechanism of FTY720 opens a new way to
develop novel immunomodulator based on S1P1-R agonist.7–11 Re-
cently, FTY720 has been approved in United States and other coun-
tries for the treatment of patients with relapsing multiple sclerosis;
however, it could possibly cause several side effects including
potentially fatal infections, bradycardia, skin cancer, macular ede-
ma, breathing and liver problems.12 Therefore an S1P-R agonist with
fewer side effects and more powerful immunosuppressive effects
would be highly desirable to be developed. Here, we focused on
designing innovative S1P1 receptor agonists with more effective
immunosuppression but lower toxicity, and we found a series of
oxazolidin-2-one-4-carboxylic amide compounds as a novel scaf-
fold of S1P receptor agonists.
FTY720 is converted to (S)-FTY720-P by sphingosine kinase. (S)-
FTY720-P acts as an agonist on four S1P receptors (S1P1,3,4,5) with
IC50s 2.1, 5.9, 23 and 2.2 nM, respectively,13,14 while the (R)-isomer
binds with 5–10-fold lower affinity. FTY720 is only phosphorylated
partially in vivo, the nonphosphorylated FTY720 has no immuno-
suppressive effects. In this study, we engaged to develop novel
non-phosphate S1P1 receptor agonists which were independent
of sphingosine kinase 2 in vivo.
The structures of S1P and FTY720-P consist of a phosphate
moiety as ‘hydrophilic head’ and a linear alkyl group as ‘hydro-
phobic tail’. As previously reported, the ‘hydrophilic head’ forms
ion-pairing interaction with ARG120 and ARG292 of S1P1-R.15,16
This ion-pairing interaction might be important for proper confor-
mational change of S1P1-R in response to agonist binding.17 How-
ever, this ion-pairing interaction can be replaceable,18 therefore,
we choose amide group as the hydrophilic head to replace the
phosphate moiety of FTY720-P. And, we engaged the hydroxyl
and amino group of FTY720-P in a carbamatic moiety to form
an oxazolinone ring in order to increase the rigidity of the mole-
cule. This oxazolinone ring, which contains two oxygen atoms
and a nitrogen atom, has potential ability to form hydrogen-bond
network with S1P1-R. In addition, various alkyl chains bearing an
oxygen atom were introduced as the ‘hydrophobic tail’ based on
bioisosteric rule to investigate structure–activity relationship of
S1P1-R agonists (Fig. 2).
Before we synthesized the molecules, we evaluated the
compound’s potential to activate S1P1 receptor by computational
modeling. We carried out molecular docking of these compounds
into the binding site of S1P1-R by means of the affinity module in In-
sight II program package (Fig. 3). Since the crystal structure of S1P1
receptor has not been published, we constructed the S1P1 receptor
model using homology modeling module in InsightII/Homology
program. The S1P1 receptor sequence was aligned against the
q
This paper was supported by NSFC, No. 30973615.
⇑
Corresponding authors.
E-mail addresses: hexinhua2007@gmail.com (X. He), lisong_amms@yahoo.cn
(S. Li).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.10.088

554
X. He et al. / Bioorg. Med. Chem. Lett. 22 (2012) 553–557
OH
NH2
OH
FTY720
"Head group"
NH+3
O
O
P
S1P
O
H
O
OH
OH
NH+3
O
O
P
FTY720-P
O
O
H
Figure 1. Structure and regions of FTY720, S1P and FTY720-P.
transmembrane helices (THs) of a rhodopsin crystal model (Protein
Data Bank code: 1boj).17 The preliminary model was initially mini-
mized for 5000 iterations of steepest descent, followed by conjugate
gradient and Truncated Newton minimizations to a RMSG (root
mean square gradient) of 0.01 kcal/mol Å. The active site includes
a long and narrow hydrophobic pocket formed by residue Met124,
Leu135, Leu136, Leu128, Phe264, Leu271, Phe295, Ala299, as well
as the two key residues, Arg120 and Arg292, which are important
to interactions of sphingosine 1-phosphate (S1P) with S1P1-R. The
molecular docking result indicated that our designed compounds
bind to S1P1-R via H-bonds instead of ion-paring interaction be-
tween S1P and S1P1-R. The strong hydrogen-bond interactions were
found between amino acid Arg120, Phe295 and oxazolidin-2-one
ring. Furthermore, the hydrophobic tail of the molecule affected
its binding mode. If the R group in Figure 2 was n-CnH2n+1
(3<n<12) or p-substituted benzyl, the conformation was linear and
fit in the S1P1 receptor active site in a pattern similar to S1P. Based
on the binding energy and geometry match, we chose to synthesize
the four candidate molecules 1a–d (Table 1) with lower binding en-
ergy of À50 to À69 kcal/mol. However, if the R groups were
o-substituted benzyls, the molecules fitted in the S1P1 receptor ac-
tive site with a different orientation in which the molecules were re-
versed as showed in Figure 3b. Their binding energy were À20 to
À40 kcal/mol, higher than the molecule with linear alkyl group or
p-substituted benzyl. For comparison, we also synthesized two mol-
ecules with o-substituted benzyl, 1e and 1f. In summary, after eval-
uation by molecular docking, six molecules were chosen to be
synthesized to investigate their S1P1-R agonist activity.
was treated with lithium iodide to give iodide 4 in 96% yield. On
the other hands, cyano-acetic acid ethyl ester reacted with sodium
nitrite and hydrogen chloride to give cyano-hydroxyimino-acetic
acid ethyl ester 5 which was then reduced by Zn/AcOH to give ami-
no-cyano-acetic acid ethyl ester 6 in 60% yields. Subsequently, 6
was reacted with (BOC)2O to produce C intermediate 7 in good
yield. Condensation reaction of 4 and 7 d by NaH in DMF provided
8 in 90% yield. 8 was then reduced by LiAlH4 to get 9. Finally, target
product 1a–e was obtained by hydrolyzing and lactonizing 9 by
‘one pot process’ under the condition of CF3COOH in anhydrous
CH2Cl2. This is the first-time report on a ‘one pot process’ for the
synthesis of oxazolin-2-one-4-carboxylic amide derivatives.19 In
order to confirm the structure of the products, single crystals of
1e prepared by slow evaporation of a solution of 1e in ethyl ace-
tate. The molecular structure was solved by using SHELXL97 pro-
gram. The crystal structure has been deposited at the Cambridge
Crystallographic Data Centre and allocated the deposition number
is CCDC 844034.
Compounds 1a–f have been evaluated for their in vitro S1P1
receptor (formerly called EDG1) agonist activity using the S1P1
RedistributionÒ assay (a commercial model with good reproduc-
ibility, Fisher BioImage ApS). This assay was designed to screen
for S1P1 agonists including internalization by monitoring the trans-
location of a membrane-localized S1P1-EGFP fusion protein to
endosomes.20,21
The results are shown in Table 2. EC50 of compounds 1a–d were
about 1–4
lM. Compound 1e showed weak S1P1 agonist activity
with an EC50 (>100
lM). Compound 1f did not show any S1P1 ago-
The syntheses of compounds 1a–f are shown in Scheme 1. Com-
mercially available 4-(2-hydroxyl-ethyl)-phenol was transformed
to ether 2 by reaction with the corresponding halide. Ether 2
reacted with methanesulfonyl chloride to give intermediate 3. 3
nistic activity in this test. The cytotoxicity was measured by the
MTT method with HLF cells (Table 2). It was found that the cyto-
toxicity of 1a–d was 8–40-fold lower than FTY720. In particular,
TC50 of 1b–d are at same level, while TC50 of 1a (121 lM) was
about four-fold of the one of 1b–d. The structure–activity relation-
ship of 1a–f can be summarized as follow, 1) Introducing a phenyl
ring to the lipophilic tail did not affect the compounds S1P1 recep-
tor affinity, but improved its cytotoxicity. 2) Replacing the C8 car-
bon-chain in tail group with an o-substituted benzyl would lead to
significant loss in agonistic activity.
A striking feature of FTY720 is the induction of a marked de-
crease in the number of peripheral blood lymphocytes (lymphope-
nia) at doses that display an immunosuppressive effect in
O
O
N
H
O
R
NH2
O
Figure 2. General structure of oxazoline-2-one-4-carboxylic amide derivatives.

X. He et al. / Bioorg. Med. Chem. Lett. 22 (2012) 553–557
555
Figure 3. (A) Compound 1a–c docked in the active site of S1P receptor site: amino acid residues in hydrophobic pocket and H-bond interactions are presented. (B) A close part
view of H-bond interactions (green) of 1a with S1P receptor, the distance between donor Atom and acceptor Atom was presented. (C) Compound 1e and 1f docked in the
active site of S1P receptor site.

556
X. He et al. / Bioorg. Med. Chem. Lett. 22 (2012) 553–557
Table 1
Table 2
Structure of compound 1a–f
S1P1 receptor agonist activity and cytotoxicity results for compounds 1a–d
a
a
Compound
Structure
Compound
EC50
(l
M)
TC50
(
lM)
T. I.b
FTY720
1a
(S)-FTY720 P
1b
3.00–10.0c
2.84 ( 0.37)
2.1nM⁄
1.10 ( 0.89)
3.63 ( 4.55)
2.72 ( 2.70)
>100
3.15 ( 0.06)
121.10 ( 29.77)
O
42.64
O
O
1a
N
27.90 ( 4.52)
26.42 ( 2.14)
25.44 ( 2.20)
—
25.36
7.28
9.35
H
1c
O
NH2
1d
1e
1f
O
NO
—
O
O
a
On the basis of triplicate experiments with given standard deviation in
parentheses.
1b
1c
1d
N
H
O
b
In vitro therapeutic index (TC50/EC50).
The EC50 of FTY720 cannot be obtained due to its cytotoxicity, and its percent
Cl
NH2
C
O
activity (PCTACT) relative to the S1P control at 3.0 lM is 37 3.1% and 10.0 lM is
O
54 5.6%; —, not test; NO, inactive. ⁄, value of literature.13,14
O
N
H
O
NH2
Table 3
Compound 1a sequesters lymphocytes form circulation
O
O
Dose (mg/kg)
Percent of CD3+ cell decrease (%)
O
N
FTY720
1a
H
O
0.1
0.3
1
3
10
9.7 26.6
47.2 13.7
79.0 7.6
80.7 7.8
86.6 15.8
—
NH2
1.7 5.7
20.0 6.1
53.0 6.4
84.3 4.0
O
O
O
N
1e
1f
H
O
NH2
Cl
induces a rapid, potent, and dose-dependent decrease of peripheral
CD3+ T cell in BALB/C mouse (Table 3). At doses form 0.1 to 3 mg/
kg, the efficacy to decrease peripheral blood CD3+ cells was lower
for 1a when compared to FTY720. At a dose of 10 mg/kg, 1a shows
in vivo efficacy comparable to FTY720(at a dose of 10 mg/kg).
In conclusion, a series of oxazolidin-2-one-4-carboxylic amide
compounds were designed and synthesized as novel non-phos-
phate S1P1-R agonists. Though their S1P1 agonist activities
in vitro were about 1000-folds weaker than (S)-FTY720-P, the
immunosuppressive effect of 1a in vivo was comparable to
FTY720 at a dose of 10 mg/Kg. This result suggested that 1a, a
non-phosphate S1P1 receptor agonist, has its own merit. The T.I.
of 1a was 43 in vitro and oral administration of 1a induces a rapid,
O
O
O
N
H
O
NH2
CH3
experimental allograft models and autoimmune disease models.
Compound 1a as the most potential agonist was chosen to be
tested for its activity of sequester circulating lymphocytes into sec-
ondary lymphoid tissues in vivo.22 Oral administration of 1a
HO
O
O
(I)
(IV)
O
O
R
(II)
R
R
(III)
O
S
OH
O
OH
I
3
O
4
2
(VII)
()
(VI)
O
O
O
NC
O
NC
O
O
NC
(V)
O
NC
O
HN
O
N
NH2
OH
O
7
5
6
O
O
R
O
R
O
OH
O
O
£¨VIII£©
R
O
O
O
(IX)
N
O
NC
N
H
NC
9
H
N
O
H
NH2
1a-d
8
Scheme 1. Reagents and conditions: (I) alkyl halide, K2CO3, DMF, 70% (average); (II) methanesulfonyl chloride, Et3N, CH2Cl2, 98%; (III) LiI, anhydrous tetrahydrofuran, 96%;
(IV) NaNO2, HCl, 81%; (V)Zn, HAc, 61%; (VI) (BOC)2O, 83%; (VII) NaH, DMF, 90%; (VIII) LiAlH4,THF,62%; (IX) CF3COOH, CH2Cl2, 60%.

X. He et al. / Bioorg. Med. Chem. Lett. 22 (2012) 553–557
13. Forrest, M.; Sun, S.; Hajdu, R., et al J. Pharmacol. Exp. Ther. 2004, 309, 758.
557
potent, and dose-dependent decrease of peripheral CD3+ T cell in
14. Chem. Biol. 2005, 12, 703.
BALB/C mouse, So compound 1a could be considered as a candidate
for further development of novel immunosuppressive agents, or as
a lead compound, rational modifications of 1a are anticipated using
medicinal chemistry techniques and molecular modeling to obtain
analogs with higher affinity and better clinical therapeutic
properties.
15. Ann. N.Y. Acad. Sci. 2000, 905, 330.
16. Parrill, A. L.; Wang, D.; Bautista, D. L. J. Biol. Chem. 2000, 75, 39379.
17. Gonzalez-Cabrera, P. J.; Jo, E.; Rosen, H. Mol. Pharmacol. 2008, 74, 1308–1318.
18. (a) Neri, C.; Williams, J. M. J. Tetrahedron: Asymmetry 2002, 13, 2197; (b) Green,
R.; Taylor, P. J. M.; Bull, S. D., et al Tetrahedron: Asymmetry 2003, 14, 2619;
(c) Wybrow, R. A. J.; Edwards, A. S.; Harrity, J. P. A., et al Tetrahedron 2004, 60,
8869.
19. Rosen, H.; Goetzl, E. J. Nature Rev. Immunol. 2005, 5, 560.
20. Olivera, A.; Spiegel, S. Nature 1993, 365, 557.
21. Chiba, K. Pharmacol. Ther. 2005, 108, 308.
Supplementary data
22. Selected data for compound 1a: 1H NMR (400 MHz, CDCl3) MS (EI) [M+]: m/
z = 362; 1H NMR (400 MHz, DMSO-d6) d (ppm): 8.24(s,1H), 7.47(s,2H), 7.06–
7.08(d, 2H), 6.81–6.83(d, 2H), 4.33–4.35(d, 1H), 4.17–4.19(d, 1H), 2.54–2.60(m,
1H), 2.30–2.38(m, 1H), 1.98–2.06(m, 1H), 1.82–1.90 (m, 1H), 1.64–1.71(m, 2H),
1.25–1.40 (m, 10H), 0.84–0.87 (t, 3H). 1b: MS (EI) [M+]: m/z = 374; 1H NMR
(400 MHz, DMSO-d6) d (ppm): 8.14(s, 1H), 7.41–7.45(m, 6H), 7.08–7.10 (d, 2H),
6.90–6.92(d, 2H), 5.00(s, 2H), 4.33–4.35(d, 1H), 4.17–4.19(d, 1H), 2.54–2.60(m,
1H), 2.30–2.38 (m, 1H), 1.98–2.06(m, 1H), 1.82–1.90(m, 1H). 1c: MS(EI) [M+]:
m/z = 368; 1H NMR (400 MHz, DMSO-d6) d (ppm): 8.17(s, 1H), 7.44–7.47(d,
2H), 7.16–7.30(m, 5H), 7.07–7.9(d, 2H), 6.83–6.85(d, 2H), 4.33–4.35(d, 1H),
4.17–4.19 (d, 1H), 3.89–3.91(t, 2H), 2.71–2.73(t, 2H), 2.54–2.60(m, 1H), 2.30–
2.38(m, 1H), 1.98–2.06(m, 1H), 1.82–1.90(m, 1H). 1d: MS(EI) [M+]: m/z = 396;
1H NMR (400 MHz, DMSO-d6) d (ppm): 8.14(s, 1H), 7.33–7.44(m, 6H), 7.08–
7.10(d, 2H), 6.90–6.92(d, 2H), 5.00(s, 2H), 4.33–4.35(d, 1H), 4.17–4.19(d, 1H),
2.54–2.60(m, 1H), 2.30–2.38(m, 1H), 1.98–2.06 (m, 1H), 1.82–1.90 (m, 1H),
1.27 (s, 9H). 1e: EI-MS(m/e): [M+] = 374; 1H NMR (400 MHz, DMSO-d6) d
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.10.088.
References and notes
1. Brinkmann, V.; Davis, M. D.; Heise, C. E., et al J. Biol. chem. 2002, 277, 21453.
2. Kiuchi, M.; Adachi, K.; Kohara, T., et al J. Med. Chem. 2000, 43, 2946.
3. Kahan, B. D. Transplant. Proc. 2004, 36, 531S.
4. Pinschewer, D. D.; Ochsenbein, A. F.; Odermatt, B., et al J. Immunol. 2000, 164,
5761.
5. Yopp, A. C.; Ledgerwood, L. G.; Ochando, J. C., et al Clin. Transplant. 2006, 20,
788.
6. Budde, K.; Schmouder, R.; Brunkhorst, R., et al J. Am. Soc. Nephrol. 2002, 13,
1073.
7. Hale, J. J.; Neway, W.; Mills, S. G. Bioorg. Med. Chem. Lett. 2004, 14, 3351.
8. Lim, H. S.; Park, J. J.; Ko, K.; Lee, M. H.; Chung, S. K. Bioorg. Med. Chem. 2004, 14,
2499.
9. Jeremy, J. C.; Michael, D. D.; Kevin, R. L. Bioorg. Med. Chem. Lett. 2003, 13, 3401.
10. Jeremy, J. C.; Michael, D. D.; Kevin, R. L. Bioorg. Med. Chem. Lett. 2004, 14, 4903.
11. Cohen, J. A.; Barkhof, F.; Comi, G., et al N. Engl. J. Med. 2010, 362, 402.
12. Mandala, S.; Hajdu, R.; Bergstrom, J., et al Science 2002, 296, 346.
(ppm): 8.14(s,1H), 7.41–7.45(m,6H), 7.08–7.10 (d,
2 H), 6.90–6.92 (d,2H),
5.00(s,2H), 4.33–4.35(d,1H), 4.17–4.19(d, 1H), 2.54–2.60(m, 1H), 2.30–2.38
(m,1H), 1.98–2.06 (m,1H), 1.82–1.90(m, 1H). 1f: EI-MS(m/e): [M+] = 355; 1H
NMR (400 MHz, DMSO-d6)
d (ppm): 8.24 (s,1H), 7.44–7.47(d,2H), 7.30–
7.32(d,2H), 7.17–7.19(d,2H), 7.07–7.09(d,2H), 6.89–6.91 (d,2H), 5.00(s,2H),
4.33–4.35(d,1H), 4.17–4.19(d, 1H), 2.54–2.60(m, 1H), 2.30–2.38 (m,1H),
2.29(s,3H), 1.98–2.06(m,1H), 1.82–1.90(m, 1H).