Design, synthesis and docking-based 3D-QSAR study of novel 2-substituted 2-aminopropane-1,3-diols as potent and selective agonists of sphingosine-1-phosphate 1 (S1P1) receptor

Article abstract of DOI:10.1039/c3md00079f

Spingosine-1-phosphate receptor 1 (S1P₁) has been actively pursued as an important therapeutic target in immune regulation. A series of 2-substituted 2-aminopropane-1,3-diols were designed and synthesized as selective S1P₁ agonists. Most of the compounds with a biphenyl ether scaffold showed moderate to excellent S1P₁/S1P₃ selectivity. Compound 40c is identified as a potent S1P₁ agonist with 350-fold S1P₁/S1P₃ selectivity. 39c, the alcohol form of 40c exerted good lymphopenia activity in vivo but with weak influence on heart rate. To investigate the SARs of 2-substituted 2-aminopropane-1,3-diols in more details, COMFA (q² = 0.547, r² = 0.986) and COMSIA (q ² = 0.544, r² = 0.943) models were established based on molecular docking alignment, which were validated with high reliability in predicting activities of agonists. The 3D-QSAR models will be helpful in the design of novel, potent and selective S1P₁ agonists. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3md00079f

View Article Online
View Journal
MedChemComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Y. Tian, J. Jin, X. Wang, W. Han, G. Li,
W. Zhou, Q. Xiao, J. Qi, X. Chen and D. Yin, Med. Chem. Commun., 2013, DOI: 10.1039/C3MD00079F.
This is an Accepted Manuscript, which has been through the RSC Publishing peer
review process and has been accepted for publication.
Accepted Manuscripts are published online shortly after acceptance, which is prior
www.rsc.org/medchemcomm
Volume 1 | Number 2 | August 2010 | Pages 105–172
to technical editing, formatting and proof reading. This free service from RSC
Publishing allows authors to make their results available to the community, in
citable form, before publication of the edited article. This Accepted Manuscript will
be replaced by the edited and formatted Advance Article as soon as this is available.
To cite this manuscript please use its permanent Digital Object Identifier (DOI®),
which is identical for all formats of publication.
More information about Accepted Manuscripts can be found in the
Information for Authors.
ISSN 2040-2503
REVIEW
Elisabeth-Maria Bissinger, Ralf Heinke,
Wolfgang Sippl and Manfred Jung
Targeting epigenetic
modifiers: Inhibitors of histone
methyltransferases
Please note that technical editing may introduce minor changes to the text and/or
graphics contained in the manuscript submitted by the author(s) which may alter
content, and that the standard Terms & Conditions and the ethical guidelines
that apply to the journal are still applicable. In no event shall the RSC be held
responsible for any errors or omissions in these Accepted Manuscript manuscripts or
any consequences arising from the use of any information contained in them.
REVIEW
Kenneth K. Y. Wong and Xuelai Liu
Silver nanoparticles—the real
“silver bullet”in clinical medicine?
2040-2503(2010)1:2;1-
X
www.rsc.org/medchemcomm
Registered Charity Number 207890

Page 1 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Design, synthesis and docking-based 3D-QSAR study of novel
2-substituted 2-aminopropane-1, 3-diols as potent and selective
agonists of sphingosine-1-phosphate 1 (S1P1) receptor
Yulin Tiana, Jing Jina, Xiaojian Wangb, Weijuan Hanb, Gang Lib, Wanqi Zhoua, Qiong Xiaoa,
Jianguo Qia, Xiaoguang Chena, Dali Yina,*
A series of 2-substituted 2-aminopropane-1, 3-diols were designed and synthesized as selective
S1P1 agonists. COMFA and COMSIA models were established based on molecular docking
alignment, which will be helpful in the design of novel, potent and selective S1P1 agonists.

Medicinal Chemistry Communications
Page 2 of 15
View Article Online
Design, synthesis and docking-based 3D-QSAR study of nDoOvI:e10l.1039/C3MD00079F
2-substituted 2-aminopropane-1, 3-diols as potent and selective
agonists of sphingosine-1-phosphate 1 (S1P1) receptor
Yulin Tiana, Jing Jina, Xiaojian Wangb, Weijuan Hanb, Gang Lib, Wanqi Zhoua, Qiong
Xiaoa, Jianguo Qia, Xiaoguang Chena, Dali Yina,*
a
State Key Laboratory of Bioactive Substances and Functions of Natural Medicines, Institute of Materia Medica,
Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing 100050, PR China
b
Department of Medicinal Chemistry, Beijing Key Laboratory of Active Substances Discovery and Drugability
Evaluation, Institute of Materia Medica, Peking Union Medical College and Chinese Academy of Medical
Sciences, Beijing 100050, PR China
Abstract
Spingosine-1-phosphate receptor 1 (S1P1) has been actively pursued as an important therapeutic
target in immune regulation. A series of 2-substituted 2-aminopropane-1, 3-diols were designed
and synthesized as selective S1P1 agonists. Most of the compounds with a biphenyl ether scaffold
showed moderate to excellent S1P1/S1P3 selectivity. Compound 40c is identified as a potent S1P1
agonist with 350-fold S1P1/S1P3 selectivity. 39c, the alcohol form of 40c exerted good
lymphopenia activity in vivo but with weak influence on heart rate. To investigate the SARs of
2-substituted 2-aminopropane-1, 3-diols in more details, COMFA (q2=0.547, r2=0.986) and
COMSIA (q2=0.544, r2=0.943) models were established based on molecular docking alignment,
which were validated with high reliability in predicting activities of agonists. The 3D-QSAR
models will be helpful in the design of novel, potent and selective S1P1 agonists.
Key words
2-substituted 2-aminopropane-1, 3-diols, S1P1 agonists, selectivity, docking, 3D-QSAR
* Corresponding author. Tel.: + 86 10 63165248; fax: + 86 10 63165248.
E-mail address: yindali@imm.ac.cn

Page 3 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Introduction
Sphingosine-1-phosphate (1, S1P) is an endogenous lysophoshpolipid that could
modulate multiple downtream signaling pathways by interacting with five receptor
subtypes (sphingosine-1-phosphate receptor subtype 1-5, S1P1-5). It could regulate a
variety of cellular functions, including differentiation, survival, proliferation and
chemotaxis.1 In vivo, S1P affects many important physiological processes, such as
angiogenesis, endothelial barrier enhancement, airway and blood vessel constriction,
heart rate modulation and bone homeostasis.2
Recently, spingosine-1-phosphate receptor 1 (S1P1) has been actively pursued as an
important therapeutic target due to its essential role played in immune regulation.3
One of the S1P1 modulators, FTY720 (2), is approved for treating multiple sclerosis
in 2010.4 Besides S1P1, FTY720 is also found to be an agonist to three other S1P
receptors (S1P3, 4, 5) after being phosphorylated to its monophosphate ester (3,
FTY720-P) in vivo by sphingosine kinase 2 (SPHK2).5 Interacting with S1P1 receptor
by FTY720-P (3) results in lymphopenia through sequestering lymphocytes in
secondary lymphoid organs.6 However, a number of side effects of FTY720 such as
bradycardia and hypertension were discovered in preclinical development and clinical
trials.7 KRP203(4), an analog of FTY720, is phosphorylated in vivo to KRP203-P (5),
which is a more selective agonist of S1P1 receptor and is believed to be safer than
FTY720,8 therefore has great potential in the treatment of autoimmune diseases and in
the prevention of transplant rejection.
Our group has sought to discover new potent S1P1 agonists with low potency on
S1P3 in order to retain the positive therapeutic properties and attenuate the potential
risk of adverse effects. Based on literature precedent, the polar head group of
FTY720-P is a key part forming salt bridge with amino acid residues in S1P1
receptor.9 Therefore, a series of new S1P1 agonists keeping the polar head group with
different lipophilic tail has been designed and synthesized (Scheme 1-4). We first
inserted phenyl rings into different positions along the aliphatic chain of FTY720 and
obtained compounds 14a-i, among which 14i with a biphenyl ether scaffold showed
moderate S1P1/S1P3 selectivity. This led to the design and synthesis of a series of
compounds bearing this biphenyl ether scaffold (21a-e, 25a-c, 30a-c and 33), as well
as compounds with substituents on one of the phenyl ring (40a-c and 47). Meanwhile,
we also investigated compounds with more rigid biphenyl scaffold (54a-b).
To investigate SARs of 2-substituted 2-aminopropane-1, 3-diols in more details, all
these compounds synthesized were docked into the binding site of S1P1 receptor
based on a crystal structure reported recently.10 A docking-based 3D-QSAR model
was successfully constructed using approaches of comparative molecular field
analysis (COMFA)11 and comparative molecular similarity indices analysis
(COMSIA)12. The 3D-QSAR models will be helpful in the design of novel, potent and
selective S1P1 agonists.

Medicinal Chemistry Communications
Page 4 of 15
View Article Online
DOI: 10.1039/C3MD00079F
Figure 1 Structures of S1P, FTY720, FTY720-P, KRP203 and KRP203-P.
Results and discussions
Chemistry
The syntheses of compounds 14a-i were described in Scheme 1. Friedel-Crafts
acylation of the aromatic ring with chloroacetyl chloride afforded 7a-i. Subsequent
replacement of chloro with diethyl acetamidomalonate gave 8a-i. 9a-i were obtained
by triethylsilane / titanium tetrachloride reduction of the carbonyl.13 Reduction of 9a-i
in the presence of NaBH4/K2HPO4 buffer gave diols 10a-i. Hydrolysis of amide 10a-i
with NaOH afforded compounds 11a-i. Then the primary amino was protected with
CbzCl to give compounds 12a-i. One of hydroxyl groups was phosphorylated by
tetrabenzyl pyrophosphate with silver(I) oxide and tetra-n-hexylammonium iodide to
give compounds 13a-i. Desired phosphate 14a-i were obtained by removing all
protective groups through hydrogenolysis.14
Scheme 2 illustrates the syntheses of compounds 21a-e, 25a-c, 30a-c and 33. 17
was prepared after Friedel-Crafts acylation of 15 and condensation with diethyl
acetamidomalonate. Acylation of 17 with benzoyl chloride gave 18a-e. The benzyl
carbonyl groups of 18a-e were reduced to methylene followed by removal of
protecting groups and phosphorylation to afford compounds 21a-e. 22 was achieved
by carbonyl reduction of 17. Compounds 25a-c were synthesized from 22 through
Suzuki coupling, followed by removal of protecting groups and phosphorylation.
Acylation of 22 with chloroacetyl chloride afforded 26. Replacement of chloro with
carboxylic acid gave esters 27a-c. The ring closing of 27a-c with acetamide gave
28a-c or 31. Finally compounds 30a-c and 33 were obtained by removing protecting
groups and phosphorylation.
Compounds 40a-c and 47 were prepared as outlined in Scheme 3. Acylation of 34
with n-butyryl chloride or chloroacetyl chloride afforded 35 or 41, respectively. 42
was achieved after esterification and ring closure. Substituted phenols were coupling
with 35 or 42 to afford phenyl ether 36a-c or 43. Acylation, condensation, carbonyl
reduction, removal of protecting groups and phosphorylation were proceeding
successively to give compounds 40a-c and 47.
The syntheses of compounds 54a-b were carried out with similar procedures to
those of compounds 30a-c, as shown in Scheme 4.

Page 5 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Scheme 1. Reagents and conditions: (a)chloroacetylchloride, AlCl3, CH2Cl2, 0C to r.t., 2h; (b) diethyl
acetamidomalonate, NaH, THF, 70C, 12h; (c) Et3SiH, TiCl4, CH2Cl2, r.t., 12h; (d) NaBH4, K2HPO4 buffer, EtOH,
rt, 12h; (e)NaOH, MeOH, 80C, 8h; (f)CbzCl, NaHCO3, ethyl acetate, rt, 4h; (g)TBPP, Ag2O, Hex4NI, CH2Cl2, rt,
20h; (h) H2, Pd/C, MeOH, rt, 8h.
Scheme 2. Reagents and conditions: (a) bromoacetyl bromide, AlCl3, CH2Cl2, 0C to r.t., 2h; (b) diethyl
acetamidomalonate, NaH, THF, 70C, 12h; (c)benzoyl chloride, AlCl3, CH2Cl2, 0C to r.t., 4h. (d) Et3SiH, TiCl4,
CH2Cl2, r.t., 12h; (e) NaBH4, K2HPO4 buffer, EtOH, rt, 12h; (f)NaOH, MeOH, 80C, 8h; (g)CbzCl, NaHCO3, ethyl
acetate, rt, 4h; (h)TBPP, Ag2O, Hex4NI, CH2Cl2, rt, 20h; (i) H2, Pd/C, MeOH, rt, 8h; (j) phenylboronic acid,
Pd(PPh)4, 2M Na2CO3, toluene/EtOH, reflux, 3h; (k)chloroacetylchloride, AlCl3, CH2Cl2, 0C to r.t., 5h; (l)
carboxylic acid, Et3N, CH3CN, reflux, 2h; (m) acetamide, BF3·Et2O, xylene, reflux, 40h; (n) acetamide, xylene,
reflux, 40h;

Medicinal Chemistry Communications
Page 6 of 15
View Article Online
DOI: 10.1039/C3MD00079F
Scheme 3. Reagents and conditions: (a) n-butyryl chloride, AlCl3, CH2Cl2, 0C to r.t., 2h; (b)phenol, Cs2CO3,
CuBr, DMF, 150C, 20h; (c) chloroacetylchloride, AlCl3, CH2Cl2, 0C to r.t., 2h. (d) diethyl acetamidomalonate,
NaH, THF, 70C, 12h; (e) Et3SiH, TiCl4, CH2Cl2, r.t., 12h; (f) NaBH4, K2HPO4 buffer, EtOH, rt, 12h; (g) NaOH,
MeOH, 80C, 8h; (h)CbzCl, NaHCO3, ethyl acetate, rt, 4h; (i)TBPP, Ag2O, Hex4NI, CH2Cl2, rt, 20h; (j) H2, Pd/C,
MeOH, rt, 8h; (k) propanoic acid, Et3N, CH3CN, reflux, 2h; (l) acetamide, BF3·Et2O, xylene, reflux, 40h;
Scheme 4. Reagents and conditions: (a) bromoacetyl bromide, AlCl3, CH2Cl2, 0C to r.t., 2h; (b) diethyl
acetamidomalonate, NaH, THF, 70C, 12h; (c) Et3SiH, TiCl4, CH2Cl2, r.t., 12h; (d)chloroacetylchloride, AlCl3,
CH2Cl2, 0C to r.t., 5h; (e) carboxylic acid, Et3N, CH3CN, reflux, 2h; (f) acetamide, BF3·Et2O, xylene, reflux, 40h;
(g) NaBH4, K2HPO4 buffer, EtOH, rt, 12h; (h)NaOH, MeOH, 80C, 8h; (i)CbzCl, NaHCO3, ethyl acetate, rt, 4h;
(j)TBPP, Ag2O, Hex4NI, CH2Cl2, rt, 20h; (k) H2, Pd/C, MeOH, rt, 8h;
Biological evaluation
S1P1 and S1P3 agonistic activities of these final compounds were evaluated by IP1
functional assay.15 As shown in table 1, biphenyl ether scaffold was considered as a
good moiety for S1P1/S1P3 selectivity enhancement (14i, 21a-e, 25a-c, 30a-c, 33). In
general, replacement of biphenyl ether with flexible scaffolds yielded less selectivity
(14a-b, 14d-e), with the exception of 14c which showed an unexpected good
S1P1/S1P3 selectivity (9-fold). Compounds with scaffold stronger than biphenyl ether,
such as naphthyl, chromone and biphenyl, often have lower potency (14f-14h, 54a-b).
These results indicated that proper structure rigidity is preferable for increasing S1P1
potency and selectivity. Introduction of chlorine and fluorine into biphenyl ether
scaffold was helpful to improve both the S1P1 potency and S1P1/S1P3 selectivity
(40a-c, 47). 40c was found to be the most potent S1P1 agonist with a 350-fold S1P1

Page 7 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
selectivity over S1P3.
Table 1 S1P1 and S1P3 agonistic activities of 2-substituted 2-aminopropane-1, 3-diols a
S1P1
EC50(nM)
14
S1P3
EC50(nM)
44
S1P1
EC50(nM)
254
S1P3
EC50(nM)
1711
383
compound
compound
3
21e
25a
25b
25c
30a
30b
30c
33
5
117
>5000
130
51
14a
14b
14c
14d
14e
14f
14g
14h
14i
21a
21b
21c
21d
91.5
37
529
>5000
2197
1273
1000
1130
113
21
329
72
674
40.9
89
203
3.72
184
341
19.8
48.1
66.3
52.9
14
886
2100
1000
508
932
40a
40b
40c
47
1790
204
800
66.3
85.8
172
302
>5000
349
391
27
225
54a
54b
876
1000
1000
157
1739
1141
632
35
aEC50 is the mean of three experimental determinations.
Based on these in vitro data, compounds 24a and 39c (the corresponding alcohol
form of compounds 25a and 40c) were evaluated in vivo for lymphopenia activities,
with FTY720 as the control. The number of lymphocytes in peripheral blood was
counted 24 h after oral administration of 1 mg/kg of each compound to Sprague
Dawley rats. As illustrated in Figure 2, both 24a and 39c showed excellent in vivo
activity equivalent to FTY720. Meanwhile, the influence on heart rate was also
determined in SD rats after oral administration of 10 mg/kg of the compounds. Both
24a and 39c showed much less effects on heart rate than FTY720 (Figure 3).
100
90
78.78
80
70
60
50
40
30
20
10
0
73.39
63.83
FTY720
24a
39c
Figure 2 The maximum lymphocyte-decreasing rate within 24h after 1 mg/kg oral administration
of FTY720, 24a, 39c to SD rats.

Medicinal Chemistry Communications
Page 8 of 15
View Article Online
DOI: 10.1039/C3MD00079F
470
450
430
410
390
370
350
330
310
control
FTY720
24a
39c
0h
1h
4h
8h
12h
24h
Figure 3 Effect on heart rate in 0, 1, 4, 8, 12 and 24 h after 10 mg/kg oral administration of
vehicle and FTY720, 24a, 39c to SD rats.
Molecular docking
The crystal structure of S1P1 receptor complexing with inhibitor ML056 was
obtained from the RCSB protein data bank (PDB entry code: 3V2Y).10, 16 29 S1P1
agonists were docked into the active site of S1P1 receptor using Surflex-Dock in
SYBYL-X-2.0.17 All the compounds were grouped into training set and test set
containing 21 and 8 compounds, respectively. Both the training and test sets were
divided according to a representative range of biological activities and structural
variations. The reliability of the docking protocol was checked by comparing the best
docking pose obtained for the crystallized inhibitor (ML056) with its bound
conformation(Figure S1).18 As a result, a root mean square deviation (RMSD) of
0.897 Å was found suggesting that the docking procedure could be relied on to predict
the binding mode of the compounds.
Compound 3(FTY720-P) was taken as the representative to elucidate the binding
mode of agonists to S1P receptor (Figure 4). It was found that FTY720-P formed salt
bridge with residue Tyr29, Lys34, Asn101, Ser105, Arg120 and Glu121. According to
the site-mutagenesis results, the salt bridge interaction in the binding site of S1P1
receptor plays an important role in determining the level of agonists’ agonism.10 It is
believed that Arg120 (blue) and Glu121 (magenta) are the key residues for ligand
binding. The guanidyl of Arg120 and the carboxyl of Glu121 form salt bridge with
phosphate group and amino-group of FTY720-P, respectively.9 Moreover, the
aliphatic chain and phenyl ring of FTY720-P accommodate to the hydrophobic cavity
of S1P1 receptor showing Van der Waals and π-π stacking interactions with side chains
of Tyr98, Met124, Phe125, Leu128, Leu195, Thr207, Phe210, Trp269, Leu272,
Phe273, Leu276 and Leu297. These interactions have crucial contribution to ligand
binding and selectivity. The docking results were consistent with the site-mutagenesis
results, providing a reliable binding mode.

Page 9 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Figure 4 The interactions between 3(FTY720-P) and the active site of S1P1 receptor. Arg120 is
shown in bule and Glu121 is shown in magenta. The green dashed line represents the salt bridge
interaction.
3D-QSAR study
To explore the specific contributions of steric, electrostatic, hydrophobic,
hydrogen bond acceptor and donor in binding modes for the S1P1 agonists with active
site of S1P1 receptor, the alignment of 21 compounds in training set was obtained
based on binding conformations in the active site of the receptor (Figure 5) and used
for COMFA and COMSIA model generation.
Figure 5 Alignment of agonists in the binding site of S1P1 receptor (shown in green)
COMFA model. The COMFA steric and electrostatic interactions were calculated
using SYBYL-X-2.0. PLS analysis19 results are listed in Table 2. The value of
cross-validated q2 is 0.547. The noncross-validated PLS analysis with the optimum
components of 5 revealed a conventional r2 value of 0.986, F-value of 215.436 and an
estimated standard error of 0.083. The steric field descriptors explain 41.6% of the
variance, while the electrostatic descriptors explain 58.4%. The values indicated a
good conventional statistical correlation, and the COMFA model had a fair predictive
ability. Figure 6(A) shows correlation between actual values and predicted values of
21 compounds and the calculated results are also listed in Table S1.
COMSIA model. COMSIA method adds three other descriptors to the COMFA
method. These are: hydrogen bond donor, hydrogen bond acceptor and hydrophobic
field. The COMSIA results are also summarized in Table 2. The model showed
leave-one-out cross-validation q2 and conventional r2 values of 0.544 and 0.943 with
the optimum components of 4. The F-value and standard error of estimation are

Medicinal Chemistry Communications
Page 10 of 15
View Article Online
DOI: 10.1039/C3MD00079F
65.659 and 0.165, respectively. The steric, electrostatic, hydrophobic, hydrogen-bond
donor, and hydrogen-bond acceptor field distributions are 9.8%, 39.9%, 17.5%,
20.6%, 12.2%, respectively. These data indicated that a reliable COMSIA model with
satisfactory predictive ability was successfully constructed. Figure 6(B) shows
correlation between actual values and predicted values of 21 compounds and the
calculated results are also listed in Table S1.
Validation of the 3D-QSAR models. Molecular docking alignment of eight
compounds in test set was performed using the same method as that of the training set
to validate the QSAR models. The results are simultaneously shown in Figure 6, Table
S1. The predicted activities are in good agreement with the observed activities in a
statistically tolerable error range, r2= 0.924 and 0.851 for COMFA and COMSIA
model, respectively. These results show that both COMFA and COMSIA model are
reliable and can be useful in designing new potent S1P1 agonists.
Table 2 Regression summary of COMFA and COMSIA models
Field contribution in %g
d
Descriptors Na
q2b
r2c
r2
SEEe F-valuef
pred
S
E
H
--
D
--
A
--
COMFA
5
4
0.547 0.986 0.924 0.083 215.436 41.6 58.4
0.544 0.943 0.851 0.165 65.659 9.8
COMSIA
39.9 17.5 20.6 12.2
a
Optimum number of components obtained from cross-validated PLS analysis and same used in final
non-cross-validated analysis.
b Cross-validated correlation coefficient.
c Non-cross-validated correlation coefficient.
d Predictive r2.
e Standard error of estimation
f F -test value.
g Field contributions: Steric (S) and electrostatic (E) fields from CoMFA. Steric (S), electrostatic (E), hydrophobic
( H ), hydrogen bond donor (D), and hydrogen bond acceptor (A) from CoMSIA.
2
1.5
1
COMFA model
training set
test set
0.5
0
0
0.5
1
1.5
2
-0.5
pEC50(experimental)
(A)

Page 11 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
COMSIA model
2
1.5
1
training set
test set
0.5
0
0
0.5
1
1.5
2
-0.5
pEC50(experimental)
(B)
Figure 6. Correlation between predicted activities by COMFA (A) and COMSIA (B) models and
the experimental activities of 29 S1P1 agonists
Interpretation of 3D-QSAR contour maps. Based on COMFA and COMSIA
model, the STDEV*COEFF contour maps were constructed. The coefficient contour
maps and the binding pocket of S1P1 receptor were superimposed to check whether
the variance in the field values corresponding to contours were known to be important
for agonist binding and obtain more explicit understanding of the key structural
features required for biological activity. To aid in visualization, the most active
compound 40c is displayed in the map (Figure 7).
The COMFA contour plots of steric and electrostatic interactions are shown in
Figure 7a-b. In the steric fields, green areas indicate regions where bulky groups
increase the agonistic activity whereas yellow contours indicate regions where bulky
groups decrease the activity (Figure 7a). Green contours are seen near the phenyl ring
close to head group suggesting that bulky substituents in these positions will
significantly improve the biological activities. It seems that bulky substituents around
this region could form steric interactions with Tyr98, Met124, Leu128, Leu195,
Leu272, Trp269 and Leu297. Meanwhile, another green contour is close to the
lipophilic tail. That explains why 30c (R4=cyclopropyl) is more potent than 30a
(R4=methyl) and 30b (R4=ethyl). Actually, bulky group in this positon would bind to
Ile203, Thr207 and Leu276 to favor the activity. There is enough space left to allow
bulky group interact with S1P1 receptor residues around the hydrophobic cluster. The
yellow polyhedron near the other phenyl ring indicates that bulky substitutions around
this position are not favorable. The reason may be that the π-π stacking interactions
between agonists and Phe125, Phe273 will be destroyed by bulky group.
In the electrostatic fields, regions where increase positive charge is favorable for
agonistic activity are indicated in blue, while regions where increased negative charge
is favorable for activity are indicated in red (Figure 7b). The electrostatic contours are
mainly distributed around the head group of 40c. A large red contour around

Medicinal Chemistry Communications
Page 12 of 15
View Article Online
DOI: 10.1039/C3MD00079F
phosphate group indicates that negatively charged substituents in this position should
increase the agonistic activity. In fact, the negatively charged substituents can form
salt bridge interactions between agonists and Arg120, Tyr29 and Lys34. Another red
contour near hydroxyl group reveals that the negative charge of oxygen atom avails
agonist activity. The reason may be that the oxygen atom acts as a hydrogen bond
acceptor to form a hydrogen bond with Asn101. Near the amino group, there are one
large blue polyhedron region indicating that positively charged substituents in this
position may increase the activity by strengthening the salt bridge interactions
between agonists and Glu121. The results show that the head group of agonists is
crucial for S1P1 agonism.
The COMSIA contour plots are shown in Figure 7c-g. The steric and electrostatic
contour maps from the COMSIA analysis are generally in accordance with the field
distributions of COMFA maps. Besides the structural features already mentioned in
the COMFA electrostatic field analysis, there is a blue polyhedron near Glu294 and
Tyr98 (Figure 7d), suggesting that a positive group in that position may form a salt
bridge with those two residues to increase the agonistic activity.
For the hydrophobic maps of COMSIA, yellow and white contours indicate that
hydrophobic and hydrophilic groups are favored, respectively (Figure 7e). The yellow
contour covers most regions of the molecule especially the phenyl ring and lipophilic
tail, indicating that these structural moieties interact with the side chains of residues at
the binding site of S1P1 receptor through hydrophobic interaction. Relevant studies
have elucidated that hydrophobic residues play crucial roles in determining the target
selectivity9. Aromatic rings presented in this region lead to hydrophobic interaction
increase, thus 21a-e, 25a-c (containing three aromatic rings) exhibit high activities.
Meanwhile two small hydrophobic disfavored white regions are embedded in the
yellow contours, which explain why 14i, 21d, 25a, 30a-c, 33, 40c and 47 with a
biphenyl ether scaffold exhibit high activities.
In the COMSIA hydrogen bond donor field, hydrogen bond donor favored regions
are represented by cyan contours and unfavorable regions by purple contours (Figure
7f), while in the hydrogen acceptor field, hydrogen bond receptor favored regions are
represented by magenta contours and unfavorable regions by red contours (Figure 7g).
Both cyan and magenta contours are close to the hydroxyl group of 40c, indicating
that hydrogen bond interactions may exist between hydroxyl group and residues such
as Asn101. Removal of hydroxyl group may decrease the activity.

Page 13 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Figure 7. 3D-QSAR contour map displayed with 40c and superimposition in the active site
residues of S1P1 receptor for (a) COMFA steric contour map; (b) COMFA electrostatic contour
map; (c) COMSIA steric contour map;(d) COMSIA electrostatic contour map;(e) COMSIA
hydrophobic contour map;(f) COMSIA hydrogen bond donor contour map;(g) COMSIA hydrogen
bond receptor contour map.
Conclusions
In summary, twenty-seven 2-substituted 2-aminopropane-1,3-diols as S1P1 agonists
were designed and synthesized. The initial SARs revealed that properly rigid
structures like biphenyl ethers were favorable for these compounds to possess
preferable S1P1/S1P3 selectivity and a chlorine or a fluorine on phenyl rings were also

Medicinal Chemistry Communications
Page 14 of 15
View Article Online
DOI: 10.1039/C3MD00079F
favorable to increase S1P1/S1P3 selectivity. Compound 40c showed the most potent
S1P1 agonistic activity and the highest S1P1/S1P3 selectivity. Compound 39c (alcohol
form of compound 40c) exerted excellent lymphopenia activity and reduced effect on
heart rate. High predictive 3D-QSAR modeling towards S1P1 agonists was performed
based on molecular docking alignment. Both COMFA (q2=0.547, r2=0.986) and
COMSIA (q2=0.544, r2=0.943) models gave good statistical results in terms of q2 and
r2 values and matched well with the 3D topology of the binding site of S1P1 receptor.
The results provided us significant insights in further understanding of the
structure–activity relationship of 2-aminopropane-1,3-diol derivatives and an
encouraging way in elucidation of protein–ligand interaction, thereby are useful in the
optimization of lead compounds and predicting the activities of new designed
agonists.
Acknowledgments
This work was financially supported by National Natural Science Foundation of
China (NSFC No.81102322 and No.81202545).
Notes and references
1
(a)Y. Sadahira, F. Ruan, S.-i. Hakomori and Y. Igarashi, Proc. Nati. Acad. Sci., 1992, 89, 9686-9690;
(b)H. Rosen, P. J. Gonzalez-Cabrera, M. G. Sanna and S. Brown, Annu. Rev. Biochem., 2009, 78,
743-768.
2
3
4
(a)I. Ishii, N. Fukushima, X. Ye and J. Chun, Annu. Rev. Biochem., 2004, 73, 321-354; (b)S. E.
Gardell, A. E. Dubin and J. Chun, Trends. Mol. Med., 2006, 12, 65-75.
M. Matloubian, C. G. Lo, G. Cinamon, M. J. Lesneski, Y. Xu, V. Brinkmann, M. L. Allende, R. L.
Proia and J. G. Cyster, Nature, 2004, 427, 355-360.
V. Brinkmann, M. D. Davis, C. E. Heise, R. Albert, S. Cottens, R. Hof, C. Bruns, E. Prieschl, T.
Baumruker, P. Hiestand, C. A. Foster, M. Zollinger and K. R. Lynch, J Biol Chem., 2002, 277,
21453-21457.
5
(a)R. Albert, K. Hinterding, V. Brinkmann, D. Guerini, C. Muller-Hartwieg, H. Knecht, C. Simeon,
M. Streiff, T. Wagner, K. Welzenbach, F. Zecri, M. Zollinger, N. Cooke and E. Francotte, J. Med.
Chem., 2005, 48, 5373-5377; (b)A. Billich, F. Bornancin, P. De´vay, D. Mechtcheriakova, N. Urtz
and T. Baumruker, J Biol Chem., 2003, 278, 47408-47415.
6
7
S. Mandala, R. Hajdu, J. Bergstrom, E. Quackenbush, J. Xie, J. Milligan, R. Thornton, G.-J. Shei,
D. Card, C. Keohane, M. Rosenbach, J. Hale, C. L. Lynch, K. Rupprecht, W. Parsons and H. Rosen,
Science, 2002, 296, 346-349.
(a)M. G. Sanna, J. Liao, E. Jo, C. Alfonso, M.-Y. Ahn, M. S. Peterson, B. Webb, S. Lefebvre, J.
Chun,N. Gray and H. Rosen, J Biol Chem., 2004, 279, 13839-13848; (b)R. M. Fryer, A. i.
Muthukumarana, P. C. Harrison, S. N. Mazurek, Rong, R. Chen, K. E. Harrington, R. M. Dinallo, J.
C. Horan, L. Patnaude, L. K. Modis and G. A. Reinhart, PLOS ONE, 2012, 7, e52985.
(a)H. Shimizu, M. Takahashi, T. Kaneko, T. Murakami, Y. Hakamata, S. Kudou, T. Kishi, K.
Fukuchi, S. Iwanami, K. Kuriyama, T. Yasue, S. Enosawa, K. Matsumoto, I. Takeyoshi, Y.
8

Page 15 of 15
Medicinal Chemistry Communications
View Article Online
DOI: 10.1039/C3MD00079F
Morishita and E. Kobayashi, Circulation, 2005, 111, 222-229; (b)J. Fujishiro, S. Kudou, S. Iwai, M.
Takahashi, Y. Hakamata, M. Kinoshita, S. Iwanami, S. Izawa, T. Yasue, K. Hashizume, T.
Murakami and E. Kobayashi, Transplantation, 2006, 82, 804-812.
9
S. C. Schurer, S. J. Brown, P. J. Gonzalez-Cabrera, M.-T. Schaeffer, J. Chapman, E. Jo, P. Chase, T.
Spicer, P. Hodder and H. Rosen, ACS Chem. Biol., 2008, 3, 486-498.
10 M. A. Hanson, C. B. Roth, E. Jo, M. T. Griffith, F. L. Scott, G. Reinhart, H. Desale, B. Clemons, S.
M. Cahalan, S. C. Schuerer, M. G. Sanna, G. W. Han, P. Kuhn, H. Rosen and R. C. Stevens,
Science, 2012, 335, 851-855.
11 R. D. Cramer, D. E. Patterson and J. D. Bunce, J. Am. Chem. Soc., 1988, 110, 5959-5967.
12 G. Klebe, U. Abraham and T. Mietzner, J. Med. Chem., 1994, 37, 4130-4146.
13 P. Durand, P. Peralba, F. Sierra and P. Renaut, Synthesis, 2000, 2000, 505–506.
14 S. Takeda, M. Chino, M. Kiuchi and K. Adachi, Tetrahedron Lett., 2005, 46, 5169-5172.
15 (a)A. Shenker, P. Goldsmith, C. G. Unson and A. M. Spiegel, J. Biol. Chem., 1991, 266,
9309-9313; (b)C. Ballatore, J. H. Soper, F. Piscitelli, M. James, L. Huang, O. Atasoylu, D. M.
Huryn, J. Q. Trojanowski, V. M. Lee, K. R. Brunden and A. B. Smith, J. Med. Chem., 2011, 54,
6969-6983.
16 A. L. Parrill, S. Lima and S. Spiegel, Sci Signal, 2012, 5, pe23.
17 A. N. Jain, J. Med. Chem., 2003, 46, 499-511.
18 C. Tintori, M. Magnani, S. Schenone and M. Botta, Eur. J. Med. Chem., 2009, 44, 990-1000.
19 B. L. Bush and R. B. Nachbar, Jr., J. Comput. Aided Mol. Des., 1993, 7, 587-619.