Discovery, synthesis and SAR analysis of novel selective small molecule S1P4-R agonists based on a (2Z,5Z)-5-((pyrrol-3-yl)methylene)-3- alkyl-2-(alkylimino)thiazolidin-4-one chemotype

Article abstract of DOI:10.1016/j.bmcl.2011.09.049

High affinity and selective S1P₄ receptor (S1P₄-R) small molecule agonists may be important proof-of-principle tools used to clarify the receptor biological function and effects to assess the therapeutic potential of the S1P₄-R in diverse disease areas including treatment of viral infections and thrombocytopenia. A high-throughput screening campaign of the Molecular Libraries-Small Molecule Repository was carried out by our laboratories and identified (2Z,5Z)-5-((1-(2-fluorophenyl)-2,5-dimethyl-1H- pyrrol-3-yl)methylene)-3-methyl-2-(methylimino) thiazolidin-4-one as a promising S1P₄-R agonist hit distinct from literature S1P₄-R modulators. Rational chemical modifications of the hit allowed the identification of a promising lead molecule with low nanomolar S1P₄-R agonist activity and exquisite selectivity over the other S1P _1-3,5-Rs family members. The lead molecule herein disclosed constitutes a valuable pharmacological tool to explore the effects of the S1P₄-R signaling cascade and elucidate the molecular basis of the receptor function.

Full text of DOI:10.1016/j.bmcl.2011.09.049

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6739–6745
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery, synthesis and SAR analysis of novel selective small molecule
S1P4-R agonists based on a (2Z,5Z)-5-((pyrrol-3-yl)methylene)-3-alkyl-
2-(alkylimino)thiazolidin-4-one chemotype
Mariangela Urbano a, Miguel Guerrero a, Subash Velaparthi a, Melissa Crisp b, Peter Chase b,
Peter Hodder b,c, Marie-Therese Schaeffer d,f, Steven Brown d,f, Hugh Rosen d,e,f, Edward Roberts a,d,
⇑
a Department of Chemistry, The Scripps Research Institute, 10550 N. Torrey Pines Rd., La Jolla, CA 92037, United States
b Scripps Research Institute Molecular Screening Center, Lead Identification Division, Translational Research Institute, 130 Scripps Way, Jupiter, FL 33458, United States
c Department of Molecular Therapeutics, Scripps Florida, 130 Scripps Way, Jupiter, FL 33458, United States
d Department of Chemical Physiology, The Scripps Research Institute, 10550 N. Torrey Pines Rd., La Jolla, CA 92037, United States
e Department of Immunology, The Scripps Research Institute, 10550 N. Torrey Pines Rd., La Jolla, CA 92037, United States
f The Scripps Research Institute Molecular Screening Center, 10550 N. Torrey Pines Rd., La Jolla, CA 92037, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
High affinity and selective S1P4 receptor (S1P4-R) small molecule agonists may be important proof-of-
principle tools used to clarify the receptor biological function and effects to assess the therapeutic poten-
tial of the S1P4-R in diverse disease areas including treatment of viral infections and thrombocytopenia. A
high-throughput screening campaign of the Molecular Libraries-Small Molecule Repository was carried
out by our laboratories and identified (2Z,5Z)-5-((1-(2-fluorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)meth-
ylene)-3-methyl-2-(methylimino) thiazolidin-4-one as a promising S1P4-R agonist hit distinct from liter-
ature S1P4-R modulators. Rational chemical modifications of the hit allowed the identification of a
promising lead molecule with low nanomolar S1P4-R agonist activity and exquisite selectivity over the
other S1P1–3,5-Rs family members. The lead molecule herein disclosed constitutes a valuable pharmaco-
logical tool to explore the effects of the S1P4-R signaling cascade and elucidate the molecular basis of the
receptor function.
Received 10 August 2011
Revised 12 September 2011
Accepted 14 September 2011
Available online 20 September 2011
Keywords:
S1P4 receptor
Selective small molecule S1P4-R agonists
Thrombocytopenia
Viral infections
Ó 2011 Elsevier Ltd. All rights reserved.
Sphingosine-1-phosphate (S1P) is a bioactive lysophospholipid
metabolite formed from ceramide and sphingosine in various cells
including mast cells, platelets, and macrophages in response to
diverse stimuli such as growth factors, cytokines, G-protein-cou-
pled receptors (GPCRs) agonists and antigens. S1P regulates a broad
variety of cellular signaling pathways resulting in calcium homeo-
stasis, actin polymerization, cell proliferation, motility and survival.
The biological responses of S1P are mediated by the activation of cell
membrane GPCRs as well as by the modulation of incompletely
characterized intracellular targets. The GPCRs responsive to S1P
include GPR3,6,12 receptor family, and five receptor subtypes of the
endothelial differentiation genes (Edg) family termed S1P1–5 (for-
mally Edg-1, -5, -3, -6, and -8).1–5
Recently, S1P1–5 receptor (S1P1–5-Rs) family has gained growing
attention due to its physiological and pathophysiological role in im-
mune function, cardiovascular system and cancer invasion.2 S1P
receptor agonists have been demonstrated to alter immunological
responses and inhibit lymphocyte recirculation.3–8 Noteworthy,
numerous high affinity S1P1-R molecule agonists are currently in
preclinical and clinical trials as immunosuppressive drug candi-
dates for the treatment of autoimmune diseases. Amongst them,
S1P1,3–5-Rs pan-agonist Fingolimod (FTY720) has been recently ap-
proved by the FDA as orally active prodrugfor the treatment of mul-
tiple sclerosis.9,10
Understanding S1P-modulated pathways in the endothelial and
vascular smooth muscle cells is a critical ongoing area of investiga-
tion. Importantly, S1P-dependent activation of vascular endothelial
S1P1,3-Rs promotes vasorelaxation responses and antagonizes
vasoconstriction by activation of the endothelial isoform of nitric
oxide synthase and subsequent production of nitric oxide via the
small GTP-binding cytoskeleton signaling protein Rac 1. In con-
trast, S1P2,3-Rs in smooth muscle cells can elicit vasoconstriction
responses through the activation of the small G-protein RhoA sig-
naling cascade, particularly at high concentrations of S1P.4,11
Interestingly, molecules targeting S1P-metabolizing enzymes
have been recently proposed as innovative potential therapeutics
for viral infections.12
Additionally, sphingosine kinase
cascade has been recently linked to the transcription factor hypox-
ia-inducible factor 1 (H1F-1 ) in distinct tumor cell lines. Given
the key role of H1F-1 in the adaptive changes to hypoxia
1 (SphK1)/S1P signaling
a
a
⇑
Corresponding author. Tel.: +1 858 784 7770; fax: +1 858 784 7745.
E-mail address: eroberts@scripps.edu (E. Roberts).
a
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.049

6740
M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
including angiogenesis and metastasis, SphK1/S1P pathway has
been proposed as an innovative target for therapeutic intervention
in oncology, thus demanding further studies to fully elucidate the
function of S1P and its target receptors upon Sphk1 activation.13
our laboratories and identified the hit (2Z,5Z)-5-((1-(2-fluo
rophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)-3-methyl-2-
(methylimino)thiazolidin-4-one 4 endowed with moderate S1P4-R
agonist activity, modest selectivity against S1P1,5-Rs and no activ-
Amongst S1P1–5-Rs, the S1P4-R couples to Gai,
G
ao and G
ity over S1P2,3-Rs at concentrations up to 25 lM (Fig. 1).
a
12/13
proteins leading to the stimulation of MAPK/ERK signaling path-
The structural integrity and biological activity of the original hit
4 were confirmed by re-synthesizing the title compound (Scheme
1). Reaction of commercially available dimethylthiourea 7I with
ethylchloroacetate 8 provided 3-methyl-2-(methylimino)-thiazoli-
din-4-one 9I which was then reacted with pyrrole-3-carbaldehyde
14. The (2Z,5Z)-configuration of 2-methylimino and the olefinic
bond of 4 was verified by 1H,1H-NOESY experiment.27
ways, as well as PLC and Rho-Cdc42 activation.14,15
S1P4-R is predominantly expressed in lymphocyte-containing
tissues including the thymus, spleen, bone marrow, appendix,
peripheral leukocytes, lung, and shows approximately 100–600-
fold reduced binding affinity for S1P compared to the other S1P-
R family members.16
Although poorly characterized, the contribution of S1P4-R to the
immune response is becoming increasingly evident. It has been
demonstrated that S1P induces migratory response of murine
T-cell lines expressing both S1P1-R and S1P4-R mRNA; in
D10.G4.1 and EL-4.IL-2 murine cells, S1P-induced migration was
significantly inhibited by treatment with (S)-FTY720-phosphate, a
potent agonist at S1P1-R and S1P4-R. Additionally, S1P-induced
T-cell migration involved the activation of Rho family small
GTPase, Cdc 42 and Rac, in murine CHO cells co-expressing
S1P4-R and S1P1-R on the cell surface. These results have suggested
that the association of S1P4-R and S1P1-R may play an important
role in the migratory and recirculation response of T-cells toward
S1P.1Additional studies demonstrated that the intratracheal deliv-
ery of the synthetic sphingosine analogs with mixed activity over
S1P1,3–5-Rs efficiently inhibited the T-cell response to influenza
virus infection by impeding the accumulation of dendritic cells
(DCs) in draining lymph nodes. The inhibitory effects were not ob-
served upon specific chemical activation of S1P1-R, and persisted in
S1P3-R null mice. Based on these findings and on the evidence that
S1P4-R is highly expressed in DCs in contrast to S1P5-R expression,
it has been hypothesized that the S1P4-R modulation in the lung
may be effective at controlling the immunopathological response
to viral infections.17,18
Moreover, while the S1P effects on the systemic vasoregulation
are predominantly mediated by S1P1,2,3-Rs subtypes, the S1P4-R
has been demonstrated to have a key role in S1P-induced vasocon-
striction in the rat pulmonary circulation by activating Rho kinase.19
Both in vitro and in vivo experiments in animal models have
recently indicated an additional potential therapeutic application
of S1P4-R molecule modulators in the terminal differentiation of
megakaryocytes. Notably, the application of S1P4-R antagonist
might be exploited for inhibiting potentially detrimental reactive
thrombocytosis, whereas S1P4-R agonists represent a potential
therapeutic approach for stimulating platelet repopulation after
thrombocytopenia.20
To start our SAR studies on 4, we prepared a series of derivatives
varying the polar thiazolidin-4-one head while maintaining the
2-fluorophenyl coil and the 2,5-dimethylpyrrol-3-yl spacer as
constant moieties. Representative examples of the explored modi-
fications are represented by compounds 15a–15n (Scheme 1, Table
1). 15a–15e, 15g, 15j–15n were synthesized using the appropriate
commercially available thioureas (7II–7VII, 7IX–7XIII) and 8 to
give the corresponding thizolidin-4-ones (9II–9VII, 9IX–9XIII).
The synthesis of compound 15h involved the preparation of the
thiourea precursor (7VIII) from methyl isothiocyanate (5) and
2-ethanolamine (6), followed by the synthesis of the corresponding
thiazolidin-4-one(9VIII). 15i was synthesized from oxazolidin-
4-one (12) obtained by reaction of dimethyl urea (10) with chloro
acetylchloride (11), while compound 15f was easily obtained from
commercially available 3-methylthiazolidine-2,4-dione (13). Final
condensation of the aforementioned five-membered ring interme-
diates (9II–9XIII, 12, 13) with 14 furnished the desired products
15a–15n.
The biological results of the obtained molecules are listed in
Table 1.28
When the methyl group was removed from either position 3
(15a) or both positions 2 and 3 (15b), the potency decreased sub-
stantially. Moreover, the 3-unsubstituted 2-phenylimino analog
(15d) was devoid of activity, and significantly reduced potency
was found for the corresponding 2-(pyridin-2-yl) derivative (15e)
as well as for the 3-methyl-2-phenylimino analog (15c). Com-
pounds containing bulky alkylic groups at both positions 2 and 3
were devoid of potency (15l–15n). By contrast, similar activity to
the hit was found in the presence of ethyl groups at the same posi-
tions (15g). Noteworthy, the analog containing the polar 2-
hydroxyethyl substituent at position 2 (15h) was only 1.5-less
potent than the hit. Taken all together, these data suggest that
bulky alkylic groups with different polarity are tolerated at posi-
tion 2, while position 3 may be involved in a lipophilic interaction
within a smaller binding pocket. Replacement of the 2-(methylimi-
no)thiazolidin-4-one head with thiazolidine-2,4-dione (15f) or
oxazolidin-4-one (15i) led to complete loss of activity. Interest-
ingly, conformationally restricted analogs (15j, 15k) were also
inactive. To note, the 2-alkylimino functionality of 15j and15k is
locked into the E-geometry suggesting that the Z-configuration at
position 2 may be a binding requirement.
Successively, we synthesized compounds 19a–19x (Scheme 2,
Table 2)in order to optimize the lipophilic aryl coil while keeping
the thiazolidin-4-one polar head and the 2,5-dimethylpyrrol-3-yl
spacer as in the hit and 15g. The synthesis of these derivatives is
depicted in Scheme 2. Condensation of hexadione 16 with the
appropriate amine 17 furnished the corresponding N-substituted-
2,5-dimethylpyrrole, which yielded the pyrrol-3-carbaldehyde
derivative 18 via Vilsmeier–Haack reaction. Condensation of 18
with the opportune thiazolidin-4-one (9I or 9VII) afforded the de-
sired products 19a–19x. The biological responses of the obtained
compounds are listed in Table 2.28
To date, despite the S1P4-R therapeutic potential, the in vivo
function of the target receptor remains largely unknown. Indeed,
the limited number of known selective S1P4-R small molecule
antagonists is restricted to the molecules recently disclosed by
our research group,21,22 whereas promiscuous selectivity profile
is found for the S1P4-R agonists reported in the literature. Amongst
them, benzymidazole derivative 1 was reported as a potent S1P4-R
agonist with low nanomolar partial agonist (pa) activity on S1P1,5
-
Rs subtypes (Fig. 1).23 The constrained azacyclic analog of FTY720 2
was described as a potent non selective S1P4,5-Rs agonist (Fig. 1).24
Remarkably, the indole-alanine derivative 3 has been reported as
apotent S1P4-R agonist with good selectivity against the other
S1P family receptors (Fig. 1).25 However, 3 and its structurally re-
lated compounds have been also studied as high affinity modula-
tors of glycine recognition site on the N-methyl-D-aspartate
receptor complex.26
In an effort to discover novel and selective S1P4-R agonists, a
high-throughput screening (HTS) campaign of the Molecular
Libraries-Small Molecule Repository (MLSMR) was carried out by
When the ortho-fluorine of the hit was replaced with either
chlorine, bromine or methyl group (19a, 19b, 19c), no activity

M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
6741
Figure 1. S1P4-R agonist modulators.
Scheme 1. Synthesis of analogs 15a–15n. Reagents and conditions: (i) 5 (1 equiv), 6 (1.05 equiv), CH2Cl2, 0 °C to rt, (ii) 7I–7XIII (1 equiv), 8 (2 equiv), NaOAc (5 equiv), EtOH,
60 °C, overnight; (iii) 10 (1 equiv), 11 (1.1 equiv), tolune, 100 °C, 5 h; (iv) 9I–9XIII or 12 or 13 (1 equiv), 14 (1 equiv), piperidine (1.5 equiv), EtOH, 60 °C, 4 h.
was observed. Interestingly, the 3-fluorophenyl isomer (19d) was
eight-fold less potent, while the 4-fluorophenyl derivative (19e)
showed only two-fold lower potency than the hit. Compounds
containing variously disubstituted phenyl rings including but not
limited to the examples herein reported (19f–h, 19j) were found
to be inactive or have significantly reduced activity, thus suggest-
ing that polar and bulky groups in this region are detrimental for
the activity. Consistent with this hypothesis, naphthalenyl deriva-
tives (19n, 19o) were inactive. Interestingly, amongst the disubsti-
tuted phenyl rings, the 4-chloro-2-fluorophenyl analog (19k) was
only two-fold less potent and the 2,4-difluorophenyl analog (19i)
was equipotent to the hit, while the unsubstituted phenyl deriva-
tive (19l) was only 2–3-fold less potent than 15g. The fluorine
was therefore identified as suitable bioisoster of the hydrogen

6742
M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
Table 1
Table 2
S1P4-R agonist activity of compounds 15a–15n (EC50 nM)
S1P4-R agonist activity of compounds 19a–19x (EC50 nM)
a
a
Compd
R2
R1
X
EC50 (nm)
Compd
Ar
R
R1
EC50 (nm)
15a
15b
15c
15d
15e
15f
15g
15h
15i
15j
15k
15l
15m
15n
NMe
NH
NPh
NPh
N(pyridin-2-yl)
O
NEt
NCH2CH2OH
NMe
NCH2CH2CH2
NCH2CH2
NiPr
H
H
Me
H
H
Me
Et
Me
Me
S
S
S
S
S
S
S
S
O
S
S
S
S
S
3600
NA
2800
NA
19a
19b
19c
19d
19e
19f
19g
19h
19i
2-Chlorophenyl
2-Bromophenyl
2-Methylphenyl
3-Fluorophenyl
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
Me
Me
Et
Me
Et
Me
Me
Me
Et
Me
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
Me
Me
Et
Me
Et
Me
Me
Me
Et
Me
Me
Et
NA
NA
NA
1600 (60%)b
2100 (70%)b
NA
4-Fluorophenyl
433
3-Cyano-4-fluoro-phenyl
4-Methoxy-3-(trifluoromethyl) phenyl
2-Methoxy-5-methyl phenyl
2,4-Difluorophenyl
2,4-Dimethylphenyl
4-Chloro-2-fluorophenyl
Phenyl
Pyridin-2-yl
Naphthalen-1-yl
Naphthalen-1-yl
3200 (50%)b
250
306
NA
NA
NA
NA
NA
NA
NA
NA
210
NA
440
602
1800
NA
19j
19k
19l
iPr
nPr
nBu
NnPr
NnBu
19m
19n
19o
19p
19q
19r
19s
19t
19u
19v
19w
19x
NA
a
3-Fluoro-pyridin-2-yl
3-Fluoro-pyridin-2-yl
Benzyl
3100 (70%)b
Data are reported as mean of n = 3 determinations.
Percentage of response. NA = not active at concentrations up to 25 lM.
b
5400 (70%)b
262
72
92
314
2-Flurobenzyl
2,6-Difluorobenzyl
2,6-Difluorobenzyl
2,4-Diflurobenzyl
1-Phenyleth-1-yl
1-Phenyleth-1-yl
atom at positions 2,4 of the phenyl ring. By contrast, replacement
of 2-fluophenyl with the basic pyridinyl ring (19m) resulted in
nine-fold decreased potency than the hit; reduced activity com-
pared to the hit and 15g, was also found for 3-fluoropyridinyl
derivatives (19p–19q). Elongation of the hydrophobic coil by inser-
tion of a methylene at the pyrrole nitrogen was tolerated as
observed for the benzyl (19r), 2-flurobenzyl (19s), 2,6-difluoroben-
zyl (19t–19u) and 2,4-difluorobenzyl (19v) analogs. However, the
introduction of a methyl group at the benzylic carbon negatively
affected the potency (19w, 19x) probably as a result of a steric
clash within the binding site or due to conformational changes in
the molecule.
147
660 (20%)b
NA
a
Data are reported as mean of n = 3 determinations.
Percentage of response. NA = not active at concentrations up to 25 lM.
b
The study of the 2,5-dimethylpyrrol-3-yl spacer was carried out
conserving the head 3-methyl-2-(methylimino)thiazolidin-4-one
as in the hit, and selecting the easily accessible coil fragments from
the active molecules. The synthesis of these derivatives is outlined
in Scheme 3. The condensation of thiazolidin-4-one 9I with com-
mercially available carbaldehydes 20 under the conditions previ-
ously described furnished the desired products 21a–21g that
were submitted for biological activity (Table 3).28
The 5-methylpyrazol-4-yl derivative (21c) was 18-fold less po-
tent than the hit. Complete loss of activity was observed for the
3,5-dimethyl and the unsubstituted pyrazol-4-yl analogs(21a–b)
as well as for the 2,5-unsubstituted pyrrol-3-yl (21d–21f) and
the 2-methyl indol-3-yl (21g) analogs, thus indicating that the
2,5-dimethylpyrrol-3-yl moiety is an essential molecular feature
for the receptor binding.
The functional activity of the most active compounds was tested
against S1P1-3,5-Rs subtypes (Table 4).The phenyl 3-ethyl-2-(ethyli-
mino)thiazolidin-4-one derivative 19l showed increased S1P1,5-Rs
activity, while retaining selectivity towards S1P2-3-Rs compared to
the hit. Interestingly, the benzyl derivatives 19r–19v showed high
to low nanomolar activity for S1P1,5-Rs; some activity versus
S1P2-R and S1P3-R was also observed for compounds 19r–19t. The
Scheme 3. Synthesis of analogs 21a–21g. Reagents and conditions: (i) 9I (1 equiv),
20 (1 equiv), piperidine (1.5 equiv), EtOH, 60 °C, 4 h.
4-fluorophenyl derivative 19e displayed two-fold decreased
potency against S1P1-R, while keeping similar activity pattern for
S1P2,3,5-Rs compared to the hit. Significantly improved selectivity
profile was found for the 2-fluorophenyl 3-methyl-2-((2-hydroxy-
ethyl)imino) derivative 15h which showed only a weak activity
for S1P1-R and S1P5-R subtypes. Similarly, the 4-chloro-2-fluoro-
phenyl derivative 19k showed only modest activity against
S1P1,5-Rs. Interestingly, the introduction of ethyl groups at the
thiazolidin-4-one head of the hit conferred to 15g exquisite
selectivity against all S1P1-3,5-Rs subtypes. Additionally, a good po-
tency/selectivity profile was observed in the presence of 2,4-difluo-
rophenyl as lipophilic coil with 19i displaying high selectivity
against S1P1-R (21-fold), and no activity against S1P2-3,5-Rs up to
25
l
M.
The SAR at the polar head were further investigated by the
synthesis of derivatives 24a–24g (Scheme 4, Table 5) containing
Scheme 2. Synthesis of analogs 19a–19x. Reagents and conditions: (i) 16 (1 equiv) 17 (1 equiv), sulfamic acid (0.05 equiv), rt, 3 h; (ii) POCl3 (2 equiv), DMF, 0–60 °C, 3 h; (iii)
18 (1 equiv), 9I or 9VII (1 equiv) piperidine (1.5 equiv), EtOH, 60 °C, 4 h.

M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
6743
Table 3
the opportune thiazolidin-4-ones (9II, 9VII–9VIII, Scheme 1) and
the pyrrol-3-carbaldehydes (18, Scheme 2). Analogously, the syn-
thesis of 24d–24g was accomplished starting from the appropriate
isothiocyanate (5, 22) and 2-methoxyethanamine (23) to provide
the corresponding thiazolidin-4-ones 9XIV–9XVI which were suc-
cessively reacted with appropriate pyrrol-3-carbaldehyde (18) to
give the final products (Scheme 4). The S1P4-R functional activity
of the monomethyl 24a and the diethyl 24b 2,4-difluorophenyl
derivatives paralleled those of the 2-fluorophenyl series 15a and
15g. The benzyl 2-(2-hydroxyethyl)iminoderivative 24c had simi-
lar potency compared to the corresponding 2-fluorophenyl analog
15h. Interestingly, the 2-(2-methoxyethyl)imino analog 24d
showed six-fold increased activity compared to 24c, thus prompt-
ing us to synthesize the 2,4-difluorophenyl analog 24f
(CYM50308).29 Notably, 24f was respectively 4- and 30-fold more
potent than 19i and the regioisomer 24g.30 These findings support
our working hypothesis that larger polar groups are better
tolerated at position 2 of the thizolidin-4-one head, as further
corroborated by the lack of activity shown by the di-substituted
3-(2-methoxyethyl)-2-(2-methoxyethyl)imino derivative 24e. To
test for selectivity, 24b, 24d and 24f were assayed on S1P1-3, 5-Rs
subtypes (Table 6). The 2,4-difluorophenyl 3-ethyl-2-(ethylimino)
derivative 24b showed decreased selectivity against S1P1,5-Rs
compared to the dimethyl (19i) and the 2-fluorophenyl (15g)
counter parts. The benzyl 3-methyl-2-(2-methoxyethyl)imino
derivative 24d was selective against S1P1-3-Rs but displayed only
three-fold selectivity versus S1P5-R. Remarkably, the 2,4-difluoro-
phenyl 2-(2-methoxyethyl)imino derivative 24f elicited exquisite
selectivity profile displaying 37-fold selectivity against S1P5-R
and no appreciable activity over S1P1-3-Rs subtypes at concentra-
S1P4-R agonist activity of compounds 21a–21g (EC50 nM)
a
Compd
Ar
EC50 (nm)
N
N
21a
NA
F
N
N
F
21b
21c
NA
N
N
3800
F
F
N
21d
21e
NA
NA
N
N
Ph
F
21f
NA
NA
tions up to 25 lM.
In summary, we have reported the discovery, synthesis and
SAR analysis of novel selective small moleculeS1P4-R functional
agonists based on a (2Z,5Z)-5-((pyrrol-3-yl)methylene)-3-alkyl-
2-(alkylimino)thiazolidin-4-one chemotype structurally unrelated
to the known S1P4-R modulators. Systematic SAR studies of the
MLSMR hit 4, endowed with moderate S1P4-R potency, high
selectivity over S1P2,3-Rs but submicromolar activity towards
both S1P1,5-Rs, led to the identification of a full spectrum selec-
tive S1P4-R agonist compound 24f (CYM50308). Indeed, the dis-
closed lead molecule 24f displayed low nanomolar S1P4-R
agonist activity and exquisite selectivity over the other S1P-Rs
subtypes. Noteworthy, 24f provides a novel valuable pharmaco-
logical tool to explore the effects of the S1P4-R signaling cascade
and elucidate the molecular basis of the in vivo receptor function.
Details of further research efforts will be communicated in due
course.
F
N
21g
a
Data are reported as mean of n = 3 determinations. NA = not active at concen-
trations up to 25 M.
l
the 2,5-dimethylpyrrol-3-yl moiety and either the benzyl or 2,
4-difluorophenyl nucleus, the latter selected as the most suitable
lipophilic coil in terms of receptor biological profile. Compounds
24a–24c were synthesized as previously described starting from
Table 4
S1P1–5-Rs selectivity counter screen of selected compounds
a
Compd
EC50 (nM)
S1P4
S1P1
S1P2
S1P3
S1P5
4
205
250
306
433
210
440
602
262
72
819
NA
20% (8.3
NA
NA
NA
NA
NA
NA
NA
3650
60% (25
NA
NA
NA
NA
NA
NA
NA
NA
721 (70%)b
NA
15g
15h
19e
19i
19k
19l
19r
19s
19t
19u
19v
l
M)c
60% (8.3 l
M)c
1800 (70%)b
4470
3010
68
652 (50%)b
NA
3400
237
144
51
34
22
37
720
250
34
18
l
l
M)c
M)c
25% (25
70%(25
NA
l
l
M)c
M)c
92
314
147
120% (25
NA
NA
287
NA
a
b
c
Data are reported as mean of n = 3 determinations.
Percentage of response.
Concentration producing the reported percentage of response. NA = not active at concentrations up to 25 lM.

6744
M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
Scheme 4. Synthesis of analogs 24a–24g. Reagents and conditions: (i) 5 or 22 (1 equiv), 23 (1.05 equiv), CH2Cl2, 0 °C to rt, 2 h; (ii) 8 (2 equiv), EtOH, 60 °C, overnight; (iii) 9II,
9VII–9VIII, 9XIV–9XVI (1 equiv), 18 (1 equiv), piperidine (1.5 equiv), EtOH, 60 °C, 4 h.
Table 5
2. Inagaki, Y.; Pham, T.; Fujiwara, Y.; Kohno, T.; Osborne, D. S.; Igarashi, Y.; Tigyi,
G.; Parrill, A. L. Biochem. J. 2005, 389, 187.
S1P4-R agonist activity of compounds 24a–24g (EC50 nM)
3. Marsolais, D.; Rosen, H. Nat. Rev. Drug Disc. 2009, 8, 297.
4. Jo, E.; Sanna, M. G.; Gonzalez-Cabrera, P. J.; Thangada, S.; Tigyi, G.; Osborne, D.
A.; Hla, T.; Parrill, A. L.; Rosen, H. Chem. Biol. 2005, 12, 703.
5. Sanna, G.; Liao, J.; Jo, E.; Alfonso, C.; Ahn, M.; Peterson, M. S.; Webb, B.;
Lefebvre, S.; Chun, J.; Gray, N.; Rosen, H. J. Biol. Chem. 2004, 279, 13839.
6. Marsolais, D.; Hahm, B.; Walsh, K. B.; Edelmann, K. H.; McGavern, D.; Hatta, Y.;
Kawaoka, Y.; Rosen, H.; Oldstone, M. B. Proc. Natl. Acad. Sci. U.S.A. 2009, 106(5),
1560.
R1
N
R
N
S
O
7. 7 Alfonso, C.; McHeyzer-Williams, M. G.; Rosen, H. Eur. J. Immunol. 2006, 36,
149.
8. Wei, S. H.; Rosen, H.; Matheu, M. P.; Sanna, M. G.; Wang, S.; Jo, E.; Wong, C.;
Parker, I.; Cahalan, M. Nat. Immunol. 2005, 6, 1228.
9. Cusack, K.; Stoffel, R. H. Curr. Opin. Drug Discov. Dev. 2010, 13, 481.
10. Swedin, H. FDA Advisory Committee Unanimously Recommends Approval of
Novartis Investigational Treatment FTY720 to Treat Relapsing Remitting MS.
Novartis Press Release, PR Newswire, June 10, 2010.
N
Ar
24a-d
Ar
a
Compd
R
R1
EC50 (nm)
24a
24b
24c
24d
24e
24f
Me
Et
H
Et
Me
Me
2,4-Difluorophenyl
2,4-Difluorophenyl
Benzyl
Benzyl
Benzyl
1000
104
396
68
11. Igarashi, J.; Michel, T. Cardiovasc. Res. 2009, 82, 212.
CH2CH2OH
CH2CH2OMe
CH2CH2OMe
CH2CH2OMe
Me
12. Seo, Y.-J.; Blake, C.; Alexander, S.; Hahm, B. J. Virol. 2010, 84, 8124.
13. Ader, I.; Malavaud, B.; Cuvillier, O. Cancer Res. 2009, 69, 3723.
14. Toman, R. E.; Spiegel, S. Neurochem. Res. 2002, 27, 619.
15. Kohno, T.; Matsuyuki, H.; Inagaki, Y.; Igarashi, Y. Genes Cells 2003, 8, 685.
16. Mandala, S.; Hajdu, R.; Bergstrom, J.; Quackenbush, E.; Xie, J.; Milligan, J.;
Thornton, R.; Shei, G.-J.; Card, D.; Keohane, C.; Rosenbach, M.; Hale, J.; Lynch, C.
L.; Rupprecht, K.; Persons, W.; Rosen, H. Science 2002, 296, 346.
17. Maeda, Y.; Matsuyuki, H.; Shimano, K.; Kataoka, H.; Sugahara, K.; Kenji Chiba,
K. J. Immunol. 2007, 178, 3437.
CH2CH2OMe
Me
CH2CH2OMe
NA
56
2,4-Difluorophenyl
2,4-Difluorophenyl
24g
1700
a
Data are reported as mean of n = 3 determinations. NA = not active at concen-
trations up to 25 lM.
18. Marsolais, D.; Hahm, B.; Edelmann, K. H.; Walsh, K. B.; Guerrero, M.; Hatta, Y.;
Kawaoka, Y.; Roberts, E.; Oldstone, M. B.; Rosen, H. Mol. Pharmacol. 2008, 74, 896.
19. Ota, H.; Ito, M.; Beutz, M. A.; Homma, N.; Holberg, G.; Abe, K.; Oka, M.;
McMurtry, I. F. Am. J. Respir. Crit. Care Med. 2009, 179, A6244.
20. Golfier, S.; Kondo, S.; Schulze, T.; Takeuchi, T.; Vassileva, G.; Achtman, A. H.;
Gräler, M. H.; Abbondanzo, S. J.; Wiekowski, M.; Kremmer, E.; Endo, Y.; Lira, S.
A.; Bacon, K. B.; Lipp, M. FASEB J. 2010, 24, 4701.
Table 6
S1P1–5-Rs selectivity counter screen of compounds 24b, 24d, 24f (EC50 nM).
a
Compd
EC50 (nm)
S1P2
S1P4
S1P1
S1P3
S1P5
21. Urbano, M.; Guerrero, M.; Zhao, J.; Velaparthi, S.; Schaeffer, M. T.; Brown, S.;
Rosen, H.; Roberts, E. Bioorg. Med. Chem. Lett. 2011, 21, 5470.
22. Guerrero, M.; Urbano, M.; Velaparthi, S.; Zhao, J.; Schaeffer, M.-T.; Brown, S.;
Rosen, H.; Roberts, E. Bioorg. Med. Chem. Lett. 2011, 21, 3632.
23. Clements, J. J.; Lynch, K. R.; Davis, M. D.; Macdonald, T. L. Bioorg. Med. Chem.
Lett. 2004, 14, 4903.
24. Hanessian, S.; Charron, G.; Billich, A.; Guerini, D. Bioorg. Med. Chem. Lett. 2007,
17, 491.
25. Azzaoui, K.; Bouhelal, R.; Buehlmayer, P.; Guerini, D.; Koller, M.; U.S. Patent
7,754,896, July 13, 2010.
24b
24d
24f
104
68
56
816
85%c
NA
NA
NA
NA
NA
NA
NA
6180
196 (80%)b
2100
a
b
c
Data are reported as mean for n = 3 determinations.
Percentage of response.
Percentage of inhibition at 25
lM.
lM. NA = not active at concentrations up to
25
Acknowledgments
26. Urwyler, S.; Floersheim, P.; Roy, B. L.; Koller, M. J. Med. Chem 2009, 52, 5093.
27. 1H,1H-NOESY spectrum was acquired using spectrometer Bruker DRX-600
equipped with a 5 mm DCH cryoprobe. NOE interaction was not observed
between the methyl groups located at position 2 and 3 of the thiazolidin-4-one
This work was supported by the National Institute of Health
Molecular Library Probe Production Center grant U54 MH084512
(E.R., H.R.) and AI074564 (Michael Oldstone, H.R.). We thank Mark
Southern for data management with Pub Chem, Pierre Baillargeon
and Lina DeLuca (Lead Identification Division, Scripps Florida) for
compound management.
nucleus, thus indicating
Z configuration of the 2-methylimino bond
(numeration as in Scheme 1). The H atom of the olefinic bond at position 5
(numeration as in Scheme 1) gave rise to a cross-peak with the methyl group at
position 2 but not with the H atom at position 4 of the 2,5-dimethylpyrrol-3-yl
nucleus; Z configuration was therefore assigned to the aforementioned bond.
28. The biological assays were performed using Tango S1P4-BLA U2OS cells
containing the human Endothelial Differentiation Gene
6 (EDG6; S1P4-R)
linked to a GAL4-VP16 transcription factor via a TEV protease site. The cells
also express a beta-arrestin/TEV protease fusion protein and a beta-lactamase
(BLA) reporter gene under the control of a UAS response element. Stimulation
of the S1P4-R by agonist causes migration of the fusion protein to the GPCR,
and through proteolysis liberates GAL4-VP16 from the receptor. The liberated
References and notes
1. Matsuyuki, H.; Maeda, Y.; Yano, K.; Sugahara, K.; Chiba, K.; Kohno, T.; Igarashi,
Y. Cell. Mol. Immunol. 2006, 3, 429.

M. Urbano et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6739–6745
6745
VP16-GAL4 migrates to the nucleus, where it induces transcription of the BLA
gene. BLA expression is monitored by measuring fluorescence resonance
((2-methoxyethyl)imino)thiazolidin-4-one scaffold. No NOE effects were
observed between the -protons of the 2-(2-methoxyethyl)imino and the 3-
a
energy transfer (FRET) of
substrate. As designed, test compounds that act as S1P4-R agonists will
activate S1P4-R and increase well FRET. Compounds were tested in triplicate at
a
cleavable, fluorogenic, cell-permeable BLA
methyl substituent of the 3-methyl-2-(2-methoxyethyl)imino)thiazolidin-4-
one core; the proton of the exocyclic olefinic bond at position 5 gave rise to a
cross-peak only with the 2-methyl group of the pyrrol-3-yl ring (numeration as
in Scheme 1).
a final nominal concentration of 25 lM.
29. The regiochemistry and (2Z,5Z)-configuration of 24f were verified respectively
by Heteronuclear Multiple Bond Coherence (HMBC) and 1H,1H-NOESY
experiments performed on spectrometer Bruker DRX-600 equipped with a
30. The regiochemistry of 24g was verified by 1H,13C-HMBC experiment performed
using spectrometer Bruker DRX-600 equipped with a 5 mm DCH cryoprobe:
the a-protons of the 3-(2-methoxyethyl) substituent coupled to both carbons
C2 and C4 of the 3-(2-methoxyethyl)-2-(methylimino)thiazolidin-4-one
nucleus (numeration as in Scheme 1).
5 mm DCH cryoprobe. In 1H,13C-HMBC spectrum the
a
-protons of the 2-(2-
methoxyethyl)imino substituent only coupled to carbon C2 of the 3-methyl-2-