Identification and characterization of small molecule modulators of the epstein-barr virus-induced gene 2 (EBI2) receptor

Article abstract of DOI:10.1021/jm4019355

Oxysterols have recently been identified as natural ligands for a G protein-coupled receptor called EBI2 (aka GPR183) (Nature 2011, 475, 524; 519). EBI2 is highly expressed in immune cells (J. Biol. Chem. 2006, 281, 13199), and its activation has been sho

Full text of DOI:10.1021/jm4019355

Article
pubs.acs.org/jmc
Identification and Characterization of Small Molecule Modulators of
the Epstein−Barr Virus-Induced Gene 2 (EBI2) Receptor
Francois Gessier,† Inga Preuss,‡ Hong Yin,∥ Mette M. Rosenkilde,⊥ Stephane Laurent,‡ Ralf Endres,§
Yu A. Chen,∥,∇ Thomas H. Marsilje,∥ Klaus Seuwen,‡ Deborah G. Nguyen,∥,# and Andreas W. Sailer*,‡
‡
§
†Global Discovery Chemistry; Developmental & Molecular Pathways, Analytical Sciences, Novartis Institutes for BioMedical
Research, Basel 4002, Switzerland
∥Genomics Institute of the Novartis Research Foundation (GNF), San Diego, California 92121, United States
⊥Laboratory for Molecular Pharmacology, Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences,
University of Copenhagen, Copenhagen 2200, Denmark
S
* Supporting Information
ABSTRACT: Oxysterols have recently been identified as
natural ligands for a G protein-coupled receptor called EBI2
(aka GPR183) (Nature 2011, 475, 524; 519). EBI2 is highly
expressed in immune cells (J. Biol. Chem. 2006, 281, 13199),
and its activation has been shown to be critical for the adaptive
immune response and has been genetically linked to
autoimmune diseases such as type I diabetes (Nature 2010,
467, 460). Here we describe the isolation of a potent small
molecule antagonist for the EBI2 receptor. First, we identified
a small molecule agonist NIBR51 (1), which enabled
identification of inhibitors of receptor activation. One
antagonist called NIBR127 (2) was used as a starting point
for a medicinal chemistry campaign, which yielded NIBR189
(4m). This compound was extensively characterized in binding and various functional signaling assays. Furthermore, we have
used 4m to block migration of a monocyte cell line called U937, suggesting a functional role of the oxysterol/EBI2 pathway in
these immune cells.
regulators, modulation of signaling pathways governed by G
protein-coupled receptors (GPCR) have recently become
apparent. Activation of the hedgehog pathway through direct
allosteric interaction of 20-hydroxycholesterol with Smooth-
ened, a seven transmembrane receptor, regulates this critical
pathway in development and cancer.13−15 We and others1,2
have identified 7α,25-dihydroxycholesterol (7α,25-OHC) and
closely related compounds as natural ligands of the EBI2
receptor. In this pathway, oxysterol not only activates receptor
signaling but subsequently also leads to directed cell migration
of the receptor expressing immune cells toward the source of
oxysterol ligand production in vitro and in vivo. Especially for
activated B cells, this chemoattraction system was shown to be
critically involved in positioning cells at distinct locations within
the germinal center reaction.1,16
INTRODUCTION
■
Oxysterols are oxidized metabolites of cholesterol which have
been linked to many fundamental physiological processes. They
are key regulators in sterol and fat metabolism and are
metabolic intermediates in the synthesis of bile acids from
cholesterol and have been show to interact with a wide variety
of proteins. First and foremost, binding of oxysterol to the
insulin-induced gene 1 (INSIG1) with a subsequent block of
sterol regulatory element-binding protein (SREBP) mediates
gene transcription and coordinates cholesterol biosynthesis.5,6
Second, several nuclear hormone receptors have been shown to
interact with oxysterols and in part use them as natural ligands.
Those include the liver x receptor (LXR)7,8 and the two
retinoid related orphan receptors RORα and RORγ.8−10
Oxysterol-binding proteins (OSBP) are another class of lipid-
binding proteins that are conserved from yeast to humans and
have been implicated in sterol transport and regulators of
protein complex assembly (for a review, see ref 11).
Furthermore, oxysterols can also function as molecular
chaperons as demonstrated by their interaction with the
Niemann−Pick type C1 (NPC1) protein, an endosomal
membrane protein required for cholesterol efflux.12 In addition
to the interaction with this diverse range of biological
To further delineate the physiological role of EBI2, there is a
need for pharmacological tools which can modulate receptor
function. As of today, no small molecule agonist has been
described for EBI2 and there are only two reports on an EBI2
antagonist called GSK682753A.17,18 This compound was
Received: December 17, 2013
Published: March 28, 2014
© 2014 American Chemical Society
3358
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
Figure 1. EBI2 small molecule agonist 1 and antagonists 2,4m. A small molecule EBI2 agonist 1 (NIBR51) was identified by screening a chemical
library. Subsequently, the agonist was used to find antagonists of receptor activation such as 2 (NIBR127). Systematic optimization of 2 yielded 4m
(NIBR189), a potent and selective EBI2 antagonist with improved pharmacokinetic properties. Half-maximal effective concentration (EC50) and
half-maximal inhibitory concentration (IC50) values are given in nanomolar concentrations with standard deviation of the mean calculated on the
logarithmic value from three or more independent experiments. Intrinsic clearance (CLint) values are based on the rate and extent of metabolism of
the compounds as determined in a mouse liver microsome assay by the disappearance of the parent compound from the reaction mixtures.
Representative dose−response curves for all three compounds can be found in the Supporting Information (Figure S1).
identified by its ability to antagonize EBI2 constitutive activity
in a recombinant system17 and was characterized by its ability
to block selected second messenger systems.18 In this report,
we describe identification and characterization of a small
molecule agonist and detail the chemical optimization from a
screening hit to 4m, a potent and selective EBI2 antagonist with
improved pharmacokinetic properties which will help to probe
the physiological function of EBI2 in health and disease.
(human EBI2 IC50, 83 nM (log IC50 −7.13 0.23, n = 12))
and good potency in a functional calcium release assays (human
EBI2 IC50, 196 nM (log IC50 −6.91 0.41, n = 23)) and is
active at the mouse receptor (mouse EBI2 FLIPR IC50, 423 nM
(log IC50 −6.69 0.58, n = 31)). Furthermore, testing of 2
against the human CysLTR2, one of the most closely related
GPCRs to EBI2, did not show any inhibition to the maximal
tested concentration of 10 μM (data not shown). Nevertheless,
the in vitro analysis of 2’s properties revealed a significant
metabolic stability issue in mouse liver microsomes (CLint
=
RESULTS
■
301 μL·min−1·mg−1), which limited the utility of this
compound to support our planned in vivo experiments. Our
concern was not only the high intrinsic microsomal clearance of
2, which negatively impacts the pharmacokinetic profile of the
compound, but also the potential formation of active
metabolites, which makes the interpretation of any biological
observations much more difficult and complex. Moreover, as
our tool compound should also be deployed to in vivo mouse
models, increased potency on the mouse EBI2 receptor was
desired. Thus, we initiated a medicinal chemistry effort aimed
to quickly optimize 2 and identify a compound with an
adequate PK profile after intravenous administration by
improving the metabolic stability while increasing the potency
especially on the mouse receptor and maintaining selectivity.
In the absence of structural information about EBI2 and
experimental determination of the metabolic weak spots of 2
that would have been useful to guide our optimization, we
opted for a systematic and broad evaluation of the left-hand
side (LFS) naphthyl ring, the glycolamide core and the right-
hand side (RHS) benzylic amide of 2. These efforts led to the
discovery of compound 4m as disclosed in this paper (Figure
1). During this study, all new derivatives were tested on the
human and the mouse EBI2 receptors (FLIPR-type calcium
release: IC50), and the most potent antagonists were assayed in
a mouse liver microsomes assay to determine their CYP
metabolic stability (clearance CLint).
Prior to the isolation of oxysterols as natural ligands for EBI2,1,2
we searched for small molecule surrogate agonists which could
be used to functionally activate the EBI2 receptor. First, we
generated a Chinese hamster ovary (CHO) cell line expressing
recombinant human EBI2, a promiscuous Gα16 protein, as well
as apo-aequorin. We screened a small molecule library with
about 100000 compounds (100 K) using release of intracellular
calcium detected by aequorin bioluminescence as primary assay
readout. With this experiment, we identified 1 as an agonist for
the EBI2 receptor (Figure 1) with a half-maximal concentration
of activation (EC50) for the human EBI2 of 1148 nM (log EC50
−5.94 0.05, n = 2). The potency of the compound on mouse
EBI2, with a 88% amino acid sequence identity to the human
receptor, was confirmed in another calcium release assay
(FLIPR-type; human EBI2 EC50, 2236 nM (log EC50 −5.66
0.08, n = 11); mouse EBI2 EC50, 2237 nM (log EC50 −5.65
0.06, n = 11)). The amino acid identity between the human and
mouse EBI2 is 88%. Specificity of receptor activation was tested
in an orthogonal assay format using a β-arrestin interaction
based Pathhunter platform19,20 with albeit weaker activity
(EC50, 6050 nM (log EC50 −5.23 0.13, n = 2)). In this assay
format, a β-galactosidase enzyme is split into two parts. While
one part is fused C-terminally to the EBI2 receptor, the other
part is attached to the β-arrestin. Ligand activation of the
receptor is followed by phosphorylation of the C-terminal part
of EBI2, which subsequently leads to the attraction of β-
arrestin. Proximity of the receptor and β-arrestin complements
the β-galactosidase enzyme, which can be detected by
conversion of the appropriate substrate. Availability of this
surrogate ligand enabled the rescreen of the 100 K small
molecule library for compounds which can block receptor
activation by 1. One of the identified hits was 2, which
demonstrated high affinity in the radioligand binding assay
The optimization of 2 started with the RHS naphthyl ring,
known to be readily metabolized in vivo toward different
oxidative pathways and, therefore, considered to likely play a
major role in the high metabolic clearance of 2. The naphthyl
ring was replaced first by the corresponding quinolines in order
to block potential metabolic weak spots and reduce the
electronic density of the aromatic system, making it less prone
3359
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
Table 1. Impact of the LHS Optimization of 2 on the Inhibition of the Mouse and the Human EBI2 Receptors (FLIPR-Type
a
Assay) and on the Metabolic Stability (Mouse Liver Microsomes Assay)
b
b
c
compd
R1
R2
human EBI2 IC50 (nM) (log IC50 SD)
mouse EBI2 IC50 (nM) (log IC50 SD)
CLint (μL·min−1·mg−1
)
2
−benzene−
196 (−6.91 0.41)
330 (−6.51 0.18)
28 (−7.55)
2934 (−5.53)
21 (−7.68)
19 (−7.72)
29 (−7.54)
24 (−7.62)
17 (−7.77)
423 (−6.69 0.58)
979 (−6.01 0.06)
100 (−7.00)
4341 (−5.36)
29 (−7.54)
92 (−7.04)
207 (−6.68)
44 (−7.36)
46 (−7.34)
301
107
131
2a
H
Cl
2b*
2c*
3a*
3b*
3c*
3d*
3e*
H
Br
Me
H
Me
Br
107
447
113
H
Me
OMe
Cl
H
Cl
F
Me
a
Receptor binding data for all compounds using a [3H] 7α,25-OHC radioligand binding assay can be found in the Supporting Information (Table
S2). IC50 were determined in a calcium release assay (FLIPR-type) with duplicate values in 8-point dose titrations. Values are given with standard
b
deviation of the mean (SD) from at least two independent experiments. Compounds with * were determined in screening mode with duplicate
c
values in a single dose titration. Intrinsic clearance (CLint) values are based on the rate and extent of metabolism of the compounds as determined
in a mouse liver microsome assay by the disappearance of the parent compound from the reaction mixture.
to oxidation. We also speculated that well positioned nitrogens
could potentially induce additional interaction with EBI2
leading to a gain in potency. Unfortunately, all compounds
made along this line did not show improved activity at the
highest tested concentration of 10 μM, suggesting that the
naphthyl moiety binds into a hydrophobic region of EBI2 with
a preference of lipophilic residues. As an alternative, we decided
to mimic the naphthyl moiety by various meta- and para-
substituted phenyl rings. On the basis of our previous
observation, we performed a systematic screen of both aryl
positions focused on lipophilic residues such as halogens, alkyls
or methoxy. While all meta-substituted derivatives were found
completely inactive (IC50 >10 μM) on both human and mouse
EBI2, the para-substitution of the phenyl proved to be a
successful alternative to replace the naphthyl ring. Even there, a
very restricted number of substituents were tolerated and from
all compounds made, and only the 4-bromo analogue 2b led to
a gain of potency of about 7-fold compared to 2 (Table 1).
Even the 3,4-dimethyl analogue 2c, which was expected to
better mimic the naphthyl ring, led to a decreased EBI2
antagonism. Therefore, we concluded that the LHS region of 2
binds most likely in a hydrophobic pocket of EBI2 and
adequate hydrophobic contacts within this pocket are essential
for potency. Finally, the metabolic stability of compound 2a
was assessed and, as expected, revealed a significant reduction
of the metabolic clearance in our in vitro assay (CLint = 107 μL·
min−1·mg−1) but still remained in the medium range.
the observed SAR was very similar to the glycolamide series
with a pronounced preference for the para-substitution pattern,
it appeared that the range of tolerated substituents was not
limited to the unique bromine or chlorine anymore but could
be extended to several others, as illustrated with the 4-methyl
and 4-methoxy analogues 3b and 3c. Moreover, the 3,4-
disubstitution was well tolerated in the cinnamyl series, leading
to potent analogues on both human and mouse receptors such
as 3d and 3e. On the basis of these findings, we hypothesized
that the cinnamyl linker induced a slightly different positioning
of the LHS phenyl moiety into the EBI2 binding pocket,
consequently allowing the pocket to accommodate a broader
range of substituents with improved hydrophobic contacts,
leading to more potent antagonism. Finally, we investigated the
metabolic stability of some of the new potent analogues but did
not observe any improvement compared to 2a.
Because of structural constraints in the LHS region of 2, only
a small number of potent compounds could be identified with
only limited improvement in metabolic stability. Thus, in the
next step we focused on the optimization of the RHS benzyl
amide of compound 3c. Encouraged by a flexible chemical
space and easily accessible compounds for library production,
we systematically screened the phenyl ring of the benzyl amide
moiety. All positions of the aromatic ring of the benzyl amide
moiety were substituted by various small hydrophobic, H-bond
acceptor and H-bond donor residues. Disappointingly, the
obtained compounds led to a significant loss of activity toward
EBI2 and the parent compound 3c remained the most potent
analogue within the series. In a second round of optimization,
we moved from benzylic amides to aryl amides and
reinvestigated each positions of the phenyl ring. Ortho and
meta substitutions were found detrimental for potency, with the
exception of the 3-methyl analogue 4a, which was equipotent to
3c against the human EBI2 receptor (IC50 = 31 nM (log IC50
−7.53 0.13, n = 2)) and slightly more active on the mouse
receptor (IC50 = 84 nM (log IC50 −7.15 0.36, n = 2)) (Table
2). In contrast, the para-position appeared to be much more
To further expand SAR in the core of the scaffold, we
investigated the glycolamide function of compound 2b and
replaced it by a cinnamyl function. The corresponding analogue
3a showed similar activity on the human EBI2 receptor (IC50
=
21 nM (log IC50 −7.68, n = 1)), whereas the potency at the
mouse receptor (IC50 = 29 nM (log IC50 −7.54, n = 1))
surprisingly increased (Table 1). To verify if the cinnamyl
function did not alter the LHS structure−activity relationship
(SAR), we rapidly reinvestigated the phenyl moiety and
prepared various monosubstituted phenyl analogues. While
3360
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
Table 2. Impact of the RHS Optimization of 3c on the Inhibition of the Human and the Mouse EBI2 Receptors (FLIPR-Type
a
Assay) and on the Metabolic Stability (Mouse Liver Microsomes Assay)
b
b
c
compd
R1
R2
R3
human EBI2 IC50 (nM) (log IC50 SD)
mouse EBI2 IC50 (nM) (log IC50 SD)
CLint (μL·min−1·mg−1
)
4a
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
Cl
Me
H
H
CF3
31 (−7.53 0.13)
98 (−7.01)
71 (−7.17 0.18)
40 (−7.40 0.09)
32 (−7.49)
19 (−7.72)
17 (−7.77)
18 (−7.74)
13 (−7.89)
44 (−7.36)
16 (−7.80)
5 (−8.30)
11 (−7.98 0.21)
84 (−7.15 0.36)
81 (−7.09)
92 (−7.10 0.34)
65 (−7.27 0.39)
55 (−7.26)
20 (−7.70)
18 (−7.74)
29 (−7.54)
21 (−7.68)
52 (−7.28)
23 (−7.64)
6 (−8.22)
15 (−7.86 0.23)
4b*
4c
295
H
Br
4d
H
Me
231
90
4e*
4f*
4g*
4h*
4i*
4j*
4k*
4l*
4m
H
OMe
NMe2
CO2Me
H
H
−CH2CH2O−
Me
87
Cl
239
Me
Me
−CH2CH2O−
−CH2CH2O−
H
60
73
46
Br
Br
OMe
a
Receptor binding data for all compounds using a [3H] 7α,25-OHC radioligand binding assay can be found in the Supporting Information (Table
b
S3). IC50 were determined in a calcium release assay (FLIPR-type) with duplicate values in 8-point dose titrations. Values are given with standard
deviation of the mean (SD) from at least two independent experiments. Compounds with * were determined in screening mode with duplicate
c
values in a single dose titration. Intrinsic clearance (CLint) values are based on the rate and extent of metabolism of the compounds as determined
in a mouse liver microsome assay by the disappearance of the parent compound from the reaction mixture.
a
Table 3. Intravenous (iv) and Oral (po) Pharmacokinetics Parameters of Compounds 4e and 4m in Mice
compd
dose (mg·kg−1
)
CL (μL·min−1·mg−1
)
t1/2 (h)
Vss (L·kg−1
)
AUC (nmol·h·L−1
)
Cmax (nmol·L−1
)
tmax (h)
F (%)
4e iv
1
1
3
52
16
0.2
1.1
0.8
1.4
841
2 435
3 608
4m iv
4m po
835
1
49
a
Pharmacokinetic parameters were calculated from blood levels after iv (1 mg·kg−1) and po (3 mg·kg−1) administration to conscious mice (cassette
dosing experiments with six compounds in maximum). Shown are composite parameters from sparse sampling. Species/strain: male mice (C57BL/
6). Formulation iv solution in NMP: Plasma (10:90), administration volume 5 mL·kg−1. Formulation po suspension in Tween 80: CMC05
(0.5:99.5), administration volume 10 mL·kg−1. CL, total blood clearance. t1/2, terminal blood half-life. Vss, apparent volume of distribution at steady
state. AUC, area under the concentration−time curve (extrapolated to infinity). Cmax, maximal blood concentration after po administration. Tmax
time of peak blood concentration after po administration. F (%), oral bioavailability.
,
attractive and flexible for substitution. Different types of
substituents resulted in improved potency compared to
compound 3c with a preference for polar (compounds 4e−g)
over lipophilic (compounds 4b−d) residues. This is illustrated
by the high activity of the 4-dimethylamino and the 4-
methylester analogues 4f and 4g. While the para position of the
RHS phenyl ring accommodated a broader range of
substituents, their relative sizes appeared crucial for potency.
In fact, larger substituents such as heterocycles, cycloalkyls, or
aryls were found weaker or inactive on both mouse and human
EBI2 receptors. To further expand the SAR, we finally
synthesized different 3,4-disubstituted or bicyclic derivatives
by combining some of our preferred meta- and para-
substituents. All compounds made with this design motif
demonstrated a synergy between the different substituents,
leading overall to improved activity toward EBI2 compared to
their corresponding monosubstituted analogues (e.g., analogues
4h−j in Table 2). In summary, the optimization of the benzylic
amide of compound 3c delivered a set of potent compounds
with a good level of substituent diversity. Selected compounds
were tested in our in vitro metabolic clearance assay, and we
found that the nature of the substituent of the aryl amide tail
plays a significant role on the overall metabolic stability of our
compounds. Indeed, the compounds bearing lipophilic residues
such as 4b, 4d, and 4i were highly metabolized independently
from the electron withdrawing or donating properties of the
substituent. In contrast, compounds substituted with polar
residues such as 4e and 4h were found to be more stable
toward metabolic degradation while the intrinsic clearance
remained in the medium range. Being particularly interested in
the in vivo clearance properties of our compounds, we
conducted a rodent PK study of the most promising derivatives
by intravenous administration in mice. In all cases, we observed
medium to low in vivo blood clearances which were consistent
with the measured in vitro data. Of all the compounds tested,
4e showed the best PK profile (Table 3).
To follow up on these results, in a final step we synthesized
another library of compounds within the cinnamyl series by
combining our best LHS and RHS residues. These were
selected based on their intrinsic potency on human and mouse
EBI2 and their positive impact on metabolic stability. As
exemplified by compounds 4k, 4l, and especially 4m (Table 2),
with this approach we were able to generate a set of potent
EBI2 antagonists with similar affinities for the human and the
mouse receptor and overall improved metabolic stability
properties compared to compound 4e. To confirm the
3361
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
Figure 2. Pharmacological characterization of 4m by [S]GTPγS binding in CHO cells. (A) Dose titrations of 7α,25-OHC at fixed concentrations of
◆
◇
●
○
▽
4m. No antagonist ( ), or 4m at 10 nM ( ), 100 nM ( ), 1000 nM ( ), 10000 nM ( ). (B) Schild plot analysis for 4m. The compound displays
competitive antagonism with a Hill coefficient of 1. The intersection of the X-axis is at Ka = −8.6. (C) Dose response of 4m at fixed, submaximal
○
●
concentrations of 7α,25-OHC (0.33 nM ( ), 0.1 nM ( ).
observed in vitro data, we carried out a full PK evaluation of
compound 4m in mice by intravenous (iv) and oral (po)
administration and found that 4m showed a significantly lower
total blood clearance than 4e (Table 3), leading to an increased
half-life and a significantly higher exposure. Moreover, on the
basis of the good oral bioavailability observed for 4m, oral
administration of the compound could also be considered as an
attractive alternative route for in vivo delivery. Testing of 4m
against a panel of other GPCRs, transporters, and enzymes (see
Experimental Methods section for details) did not show any
activity at the maximal tested concentration of 30 μM. The
good mouse PK and selectivity profiles of 4m, associated with
high potency on EBI2, suggests that 4m is a suitable candidate
to support in vivo activities and for the pharmacological
validation of EBI2.
Pharmacological Characterization of 4m. While the
optimization of 4m was guided by functional assays measuring
release of intracellular calcium upon activation, we extended the
characterization the pharmacological properties of 4m by
[35S]GTPγS binding assays including a Schild plot analysis.
First we carried out dose titrations of 7α,25-OHC at fixed
concentrations of 4m (Figure 2A). 4m displays competitive
antagonism, and the Schild plot reveals a Hill coefficient of 1
(Figure 2B). The intersection of the X-axis corresponding to
the Ka value is at −8.6. Besides the dose-dependent right shift
in the 7α,25-OHC activation curves, the presence of 4m also
resulted in a downward shift of the activation curves. This
inhibition of the basal activity of EBI2 without addition of
7α,25-OHC indicates that the antagonist also act as an inverse
agonist. We also conducted dose titrations of 4m at fixed,
submaximal concentrations of oxysterol (0.33, 0.1 nM) and
found IC50s at 8.5 and 7.0 nM (log IC50s at −8.07 0.07 and
−8.15 0.11), respectively.
Figure 3. Pharmacological characterization of oxysterol-dependent
calcium release and cell migration of U937 cells. (A) 7α,25-OHC-
dependent release of intracellular calcium in U937 cells; 7α,25-OHC
◇
●
(
); 1 ( ). (B) Block of 7α,25-OHC-induced calcium release in
◇
●
U937 cells; 4m ( ); 2 ( ). The 7α,25-OHC agonist concentration
used in this experiment was 100 nM. (C) 7α,25-OHC-dependent
migration of U937 cells; 7α,25-OHC ( ). (D) Inhibition of 7α,25-
●
◇
●
OHC-dependent cell migration of U937 cells by 4m ( ) or 2 ( ).
The 7α,25-OHC agonist concentration used in this experiment was 20
nM. (D) All experiments were replicated three or more times, and the
graphs show data from one representative experiment.
While all assays described so far were conducted with cells
expressing recombinant EBI2 receptors, we extended our
analysis of 4m also to immune cells expressing endogenous
EBI2 receptor. Screening of EBI2 mRNA expression levels in
several cell lines suggested that the monocyte-derived U937 cell
expresses functional EBI2 levels. This was confirmed by
demonstrating 7α,25-OHC dependent release of intracellular
calcium with either the natural ligand 7α,25-OHC or the small
molecule activator 1 with half-maximal concentrations of
activation of 14 nM (log EC50 −7.95 0.33, n = 5) or 6662
nM (log EC50 −5.18 0.07, n = 4), respectively (Figure 3A).
4m and 2 can block oxysterol-dependent activation with an
IC50 of 9 nM (log IC50 −8.06 0.18, n = 3) or 91 nM (log IC50
−7.08 0.22, n = 3), respectively (Figure 3B). Similar to B
cells or dendritic cells, this activation leads to a directed
migration of U937 cells in a transwell assay. For 7α,25-OH, we
find an EC50 of 2.42 nM (log EC50 −8.64
0.18, n = 4)
(Figure 3C). The EBI2 antagonists 4m and 2 can block this
migration with an IC50 of 0.3 nM (log IC50 −9.65 0.34, n =
4) and 5.3 nM (log IC50 −8.31 0.20, n = 4), respectively
(Figure 3D).
3362
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
intensity peak by comparing the difference in intensity of each time
point with an earlier time point and determining if the difference
exceeds a defined threshold, which was usually set at 100. If it does,
then the intensity of the later time point is added into the Slope
Threshold’s sum.
CONCLUSION
■
Here we report for the first time identification of a small
molecule agonist 1 for EBI2, which was successfully used to
find receptor antagonists. While the utility of one of the initial
antagonist hit 2, due to potency and pharmacokinetic profile,
was limited, we conducted a medicinal chemistry campaign
which yielded a potent and selective EBI2 antagonist 4m.
Pharmacological characterization suggests that 4m is a potent
and selective EBI2 antagonist with pharmacokinetic properties
which should allow use for in vitro and in vivo experiments.
Given the selective expression of EBI2 in immune cells, much
of the initial testing was directed toward paradigms of
autoimmune disorders.4 Our results demonstrating a functional
role of the oxysterol/EBI2 pathway in monocytes suggest that
exploration of the physiological role of this system should be
extended beyond autoimmune disorders. This was further
substantiated by a recent report21 that showed that chemotactic
migration in THP1 cells is EBI2 dependent. Considering the
substantial concentration of oxysterols in atherosclerotic
lesion22 and the important role of monocyte migration for
the disease pathophysiology,23 our result should encourage
testing of EBI2 antagonists in paradigms relevant for
cardiovascular disease.
Membrane Preparation/Radioligand Binding Assay. Membranes
and assays were prepared as previously described.1,24 In brief, CHO
cells stably expressing human EBI2 were grown to confluence, washed
with phosphate buffered saline (PBS) once, and frozen at −80 °C. The
frozen cell pellets were resuspended in ice-cold homogenization buffer
(10 mL/g, 20 mM Tris pH 7.2) supplemented with a protease
inhibitor cocktail (Complete, Roche). The cell suspension was then
homogenized using a Dounce homogenizer (10 strokes) with a driller.
After centrifugation at 110000g for 45 min and removal of the
supernatant, pellets were resuspended in sucrose buffer (50 mM Tris
pH 7.2, 250 mM sucrose, 10% glycerol) by vortexing. The suspension
was again homogenize using a Dounce homogenizer (10 strokes) and
spun at 100000g for 45 min. The supernatant was discarded, the pellet
resuspended in sucrose buffer, protein concentration determined,
aliquoted, and frozen at −80 °C. These membrane preparations were
used for radioligand binding as well as GTPγS assays.
For the radioligand binding assay lyophilized wheat germ agglutinin
scintillation proximity assay (SPA) beads (RPNQ0001) were
purchased from GE Healthcare (Buckinghamshire, UK). One vial
(500 mg) was reconstituted by adding 5 mL of distilled water to give a
final concentration of 100 mg/mL. The radioligand (3H-7α, 25-OHC)
used in these experiments was batch RSE436-6 with a specific activity
of 972 GBq/mmol, which was stored at −20 °C. For the assay,
compounds were serially diluted into the assay buffer (HBSS (Gibco,
no. 14065) + 20 mM HEPES (Gibco, no. 1563) + 0.1% β-cyclodextrin
(Sigma-Aldrich, no. C4767), DMSO concentration <2%) before 10 μL
were transferred to the assay plate (Optiplate, white, flat square
bottom, PerkinElmer). In a second step, 20 μL of a membrane (10 μg/
well)/bead mixture (200 μg/well) was added before a final addition of
20 μL radioligand (10 nM/well). After 10 min of shaking, assay plates
were incubated at room temperature overnight and then read for 1
min with a scintillation counter (Topcount, PerkinElmer). IC50 values
were determined with using standard software (Prism, GraphPad
Software Inc.)
EXPERIMENTAL SECTION
Pharmacological Characterization of EBI2 Modulators.
Calcium Mobilization Assays. Methods on for calcium mobilization
assays were published previously.1 In brief:
■
FLIPR-Type Calcium Assay. Cells were seeded in black/clear
bottom, poly-D-Lysine coated 384-well plates at a cell density of 10000
cells/25 μL/well using culture medium. Cells were cultured for 24 h at
37 °C/5% CO2. Culture medium was removed, and for cell loading, 20
μL per well of loading buffer (Calcium-4 no wash dye, Molecular
Devices) were distributed into the wells and the plates were incubated
for 60 min at 37 °C in 5% CO2. With the FLIPR pipettor, 10 μL of the
different compounds diluted in FLIPR buffer (HBSS (Gibco, no.
14065) + 20 mM HEPES (Gibco, no. 1563) + 0.1% β-cyclodextrin
(Sigma-Aldrich, no. C4767)) were added to the cells, and the
subsequent fluorescent changes were monitored for 3 min. Using the
FLIPR software tool, two values of fluorescence calcium responses
were exported: Fmax, the fluorescence at the peak and Fbasal
corresponding to the value prior to compound injection. From these
two values, the calcium response is normalized with respect to the
calcium baseline values using dF/F = (Fmax − Fbasal)/Fbasal. The
[35S]GTPγS Binding Assay. [35S]GTPγS binding experiments were
carried out as previously described.3,24 In brief, a volume of membrane
preparation (corresponding to 20 μg protein/well) were diluted in
assay buffer (50 mM HEPES, 3 mM MgCl2, 100 mM NaCl, 1 mM
EGTA, 3 μM GDP, 10 μg/mL saponin, and Complete inhibitor mix,
pH 7.4). [35S]GTPγS (1250 Ci/mmol, 12.5 mCi/mL) diluted in assay
buffer was added to a final concentration of 1 nM and incubated 1 h at
30 °C. Subsequently, WGA PVT SPA-beads (PerkinElmer) were
added (final concentration of 2.8 mg/mL), followed by 30 min of
incubation at room temperature on a plate shaker. Finally, the plates
were centrifuged at 1500 rpm for 5 min and the amount of
radioactivity determined using a Top Count scintillation counter
(Packard Instruments). The level of unspecific binding was
determined by adding unlabeled GTPγS at a final concentration of
40 μM.
β-Arrestin Assay (Pathhunter, DiscoveRx). We applied the
Pathhunter technology (DiscoverRx) to probe for an interaction of
EBI2 with β-arrestin. For that, human EBI2 was cloned into the
Prolink vector (DiscoveRx) for GPCR-prolink fusion protein
production. Parental HEK293 cells that stably express the β-
arrestin2-β-gal-EA fusion protein (HEK293-BAEA) were detached
and transiently transfected with hEBI2-prolink vector using Fugene6
transfection reagent in suspension mode. Transfected cells in assay
medium (Phenol Red-free DMEM with 10% FBS) were plated into
white solid 384-well plates at 15000 cells/25 μL/well. After overnight
incubation, 200 nL of test molecules were transferred into the cell
plates by PinTool (GNF Systems), followed by 1.5 h incubation at 37
°C, 5% CO2. Flash detection reagents were added at 12.5 μL/well.
After a 30 min room temperature incubation, the cell plates were read
on CLIPR (PerkinElmer) or Acquest (Molecular Devices) for
luminescence signal. An algorithm similar to area under the curve
maximum dF/F is called Emax
.
For calcium mobilization assay of the U937 cell line the protocol
had to be modified. The number of cells plated for the experiment was
increased to 40000/well.
EC50 and IC50 values were determined with duplicate values in 8-
point dose titrations using standard software (Prism, GraphPad
Software Inc.). Values are given with the standard deviation of the
mean (SD) calculated from logarithmic values from two or more
independent experiments. Compounds with an * were determined in
screening mode with duplicate values in a single dose titration.
Aequorin Bioluminescence-Type Calcium Assay. The cells were
seeded into 384-well black clear bottom plates (Greiner Bio-one) at
10000 cells/25 μL/well in F12 medium containing 3% FBS and
incubated overnight. Coelenterazine (20 μM final) was added into the
cell plates at 25 μL/well. The cell plates were returned to incubator for
3 h. The compounds were diluted 1:20 into an intermediate plate with
Tyrode buffer (130 mM NaCl, 2 mM CaCl2, 5 mM NaHCO3, 5 mM
KCl, 1 mM MgCl2, 20 mM HEPES, pH 7.4). Then 12.5 μL/well of
prediluted compounds were transferred into the cell assay plate and
read on LumiLux (PerkinElmer) for flash luminescence. An algorithm
similar to area under curve (AUC) was created in-house to analyze the
aequorin kinetic data, which was named Slope Threshold (Slope 100).
The algorithm defines the beginning and the end of the luminescence
3363
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Compounds 2a−c and 3a−e
a
Reagents and conditions: (a) 2-phenylacetyl chloride, pyridine, CH2Cl2, RT, 1.5 h, 97%; (b) TFA, CH2Cl2, RT, 30 min, 80%; (c) chloroacetyl
chloride, Et3N, CH2Cl2, RT, 30 min, 57%; (d) K2CO3, DMF, RT, 16 h, 69%−89%; (e) Et3N, HATU, DMF, RT, 3 h, 48−85%.
(AUC) was created in-house to analyze the aequorin kinetic data,
which was named Slope Threshold (Slope 100). The algorithm defines
the beginning and the end of the luminescence intensity peak by
comparing the difference in intensity of each time point with an earlier
time point, and determining if the difference exceeds a defined
threshold, which was usually set at 100. If it does, then the intensity of
the later time point is added into the Slope Threshold’s sum.
Chemotaxis Assay for the U937 Cell Line. Prior to the migration
assay, U937 cells were kept overnight in lipid-depleted media (RPMI
(Gibco, no. 61870-010) + 1% NEAA (Gibco, no. 11140-035) + 1%
NaPyr (Gibco, no. 11360-039) + 0.1% 2-mercaptoethanol (Gibco, no.
31350-010) + 1% Pen/Strep (Bioconcept, no. 4-01F00-H) + 10%
charcoal-treated FCS (Gibco, no. 12676). Chemotaxis was performed
using HTS Transwell-96 plates with 5.0 μm pore polycarbonate
membranes (Corning, no. 3387) according to the manufacturer’s
protocol. In brief, lower chambers were filled with 240 μL of
compound solution in media in which the FCS was substituted with
1% BSA (Gibco, no. 15260-037), and after insertion of the filters, 0.75
× 105 cells in 75 μL media were added to the upper chamber. After 3 h
at 37 °C, cells in the lower chamber were analyzed by flow cytometry
using CyFlow Space with a RobbyWell plate reader (Partec).
Chemotaxis is expressed as percent of input cells.
Specificity Testing of EBI2 Antagonist 4m. To assess specificity of
4m, we have tested this compound in a panel of assays versus other
GPCRs, enzymes, transporters, and other receptors. In none of these
assays did we see significant interaction between these 4m and the
tested receptors at concentration below 30 μM. Binding assays
included 11 different GPCRs (acetylcholine (muscarinic), M1;
adenosine receptor, A2A, A3; adrenoreceptors, α2C, β1; cysteinyl
leukotriene receptors, CysLTR1, CysLTR2; dopamine receptor, D2;
histamine receptor, H1; melanocortin receptor, MC3; prostanoid
receptor, TP). Transporter assays included dopamine transporter,
DAT; norepinephrine transporter, NET. Enzymes included cyclo-
oxygenase, COX-1; monoaminoxidase, MAO-A; phosphodiesterases,
PDE3; PDE4D. Nuclear hormone receptor: SXR (NR1I2). Functional
assays included three different GPCRs (5-hydroxytryptamine receptor,
5HT2A; acetylcholine receptor (muscarinic), M2; adrenoreceptor,
α1A.
mM NADPH. At specific reaction time points (0, 5, 20, and 30 min),
aliquots (50 μL) were removed and reactions were terminated by the
addition to acetonitrile (100 μL) containing the analytical internal
standards (1 μM alprenolol and 1.6 μM chlorzoxazone) and stored at
−20 °C for at least 1 h to allow complete precipitation of proteins.
The samples were then centrifuged, and 20 μL of the supernatants
were analyzed by LC-MS/MS for quantitation of the remaining
compound. The percentage of compound remaining, relative to time
zero minute incubation, was used to estimate the in vitro elimination-
rate constant (kmic), which is used to calculate the in vitro metabolic
half-life (t1/2, min) and intrinsic clearance (CLint, μL·min−1·mg−1)
rates. In general, compounds with high in vitro CLint (>150 μL·min−1·
mg−1) are considered to be at high risk for exposure-limiting
metabolism in vivo. Compounds with low in vitro CLint (<50 μL·
min−1·mg−1) are not expected to be at significant risk for exposure-
limiting metabolism in vivo.
Pharmacokinetic Analysis. Pharmacokinetic parameters were
calculated from blood levels after iv (1 mg·kg−1) and po (3 mg·
kg−1) administration to conscious male mice (3 mice/time point;
strain, C57BL/6; age, 9−11 weeks; source, Charles River, Iffa Credo,
France). Cassette dosing was used with six compounds in maximum.
For iv application, compounds were formulated in N-methyl-2-
pyrrolidone (NMP):plasma (10:90), administration volume 5 mL·
kg−1. For po dosing, a compound suspension was formulated in Tween
80:carboxymethylcellulose (CMC05) (0.5:99.5), administration vol-
ume 10 mL·kg−1. The animals were housed in groups of three in
normal cages with free access to food and tap water throughout the
experiment. All animal procedures were approved by the Institutional
Animal Care and Use Committee (IACUC) and ethics committee.
Sampling time points for iv dosing were at 0.08, 0.5, 1.0, 2.0, 4.0, 8.0,
and 24.0 h and for po dosing at 0.25, 0.5, 1.0, 2.0, 4.0, 8.0, and 24.0 h.
Then 30 μL of blood (EDTA sample collection) was mixed with 200
μL of acetonitrile and centrifuged at 4 °C. Then 200 μL of supernatant
were transferred into a microtiter plate, solvent evaporated, and
residue dissolved in 60 μL of methanol and 40 μL of 0.1% formic acid.
An aliquot of each sample was injected into the LC-MS/MS system for
analysis.
Compounds. 7α,25-OHC (CAS no. 64907-22-8) was purchased
from Avanti Polar Lipids Inc. (Alabaster, AL). Compounds 1 and 2
were purchased from Maybridge (UK) and Chembridge (San Diego,
CA), respectively.
Synthesis. Compounds 2a−c and 3a−e were prepared as shown in
Scheme 1. The common intermediate 6 to both series was easily
synthesized in two steps by acylation of the commercially available tert-
butyl piperazine-1-carboxylate with 2-phenylacetyl chloride under
basic conditions to give 5, which was subsequently Boc deprotected
Mouse Liver Microsome Assay. Metabolic stability of compounds
was determined in mouse liver microsome assay using the compound
depletion approach quantified by LC-MS/MS as previously
described.24,25 In brief, 30 μL of compound (10 μM in 0.1%
DMSO/water) was added to 120 μL of human liver microsomal
protein (1.25 mg/mL; male mouse (CD-1) liver microsomes (BD
Biosciences)) suspended in phosphate buffer (pH 7.4). Reactions were
initiated by the addition of 150 μL of a cofactor solution containing 2
3364
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Compounds 4a−m
a
Reagents and conditions: (a) Et3N, HATU, DMF, RT, 2 h, 63%−96%; (b) HCl 4N, dioxane, RT, 16 h, 89%−quant; (c) Et3N, HATU, DMF, RT, 3
h, 27%−77%; (d) NMM, HATU, DMF, RT, 16 h, 69%.
with TFA. The intermediate 6 was further acylated with chloroacetyl
chloride under basic conditions to give 7, which reacted with
appropriate phenolates (generated in situ in the presence of K2CO3)
under nucleophilic reaction conditions to deliver the final compounds
2a−c in good yields. Starting from the same intermediate 6,
compounds 3a−e were obtained in one step by amide coupling
reactions with appropriate cinnamic acids, using HATU as a coupling
reagent.
A three-step reaction sequence was used for the preparation of
compounds 4a−l as disclosed in Scheme 2. Amide coupling reactions
performed on appropriate cinnamic acids with tert-butyl piperazine-1-
carboxylate, using the same conditions as for the preparation of
compounds 3, gave the intermediates 8a−c. Boc deprotection with
HCl in dioxane delivered the HCl salts 9a−c, which subsequently
reacted with various carboxylic acids under similar amide coupling
reaction conditions as described before to give the desired compounds
4a−l. As a higher amount of our tool compounds 4m was required to
support our biological activities, a specific one-step synthesis was
achieved from commercially available building blocks; the coupling of
(E)-3-(4-bromophenyl)acrylic acid with (4-methoxyphenyl)-
(piperazin-1-yl)methanone using previously described reaction con-
ditions delivered 4m in good yields.
Bruker Ultrashield 400 (400 MHz) or a 500 MHz DRX Bruker
CryoProbe (500 MHz) spectrometer using trimethylsilane as an
internal standard. Chemical shifts (δ values) were reported in parts per
million (ppm) downfield from tetramethylsilane, and coupling
constants (J) were given in Hz. LC-MS data were recorded on a
Waters 2795 Alliance HT instrument equipped with a Waters 2996
photodiode array detector (PDA MaxPlot detection at 210.0400.0
nm), a Micromass ZQ detector (positive ion electrospray ionization
detection), and a C-18 column (Sunfire C18, 4.6 mm × 20 mm, 3.5
μm) eluting with a linear gradient (5−100% acetonitrile in water both
containing 0.1% TFA) in 4 min with a flow rate of 3 mL/min at 45 °C.
Purity of all the tested compounds was determined by LC-MS and was
found to be ≥95% unless otherwise stated.
tert-Butyl 4-(2-Phenylacetyl)piperazine-1-carboxylate (5). To a
solution of tert-butyl piperazine-1-carboxylate (4.0 g, 21.48 mmol) and
pyridine (3.47 mL, 43.0 mmol) in dry CH2Cl2 (100 mL) was added
dropwise a solution of 2-phenylacetyl chloride (2.98 mL, 22.55 mmol)
in CH2Cl2 (10 mL) at 0 °C (ice bath). After the addition, the reaction
mixture was allowed to warm to RT and was stirred for 1.5 h. The
mixture was diluted with CH2Cl2 and washed successively with HCl
2N in water (2×) and Na2CO3 2 M in water (2×). The organic layer
was dried over Na2SO4, filtered, and evaporated to dryness to yield 5
(6.34 g, 20.83 mmol, 97%) as a white solid, which was used in the next
step without further purification. LC-MS (ESI): m/z 305.3 [M + H]+.
1H NMR (400 MHz, DMSO-d6): 1.40 (s, 9H) 3.20−3.25 (m, 2H)
3.25−3.30 (m, 2H) 3.42−3.49 (m, 4H) 3.73 (s, 2H) 7.20−7.25 (m,
3H) 7.28−7.32 (m, 2H).
Chemistry. All commercial chemicals and solvents are reagent
grade and were used without further purification unless otherwise
indicated. All reactions were carried out under an inert atmosphere of
argon. Reactions were monitored by LC-MS on a Waters 2695
Alliance HT instrument equipped with a Waters 2996 photodiode
array detector (PDA MaxPlot detection at 210.0−400.0 nm) and a C-
18 column (Sunfire, 4.6 mm × 20 mm, 3.5 μm) eluting with a linear
gradient (5−100% acetonitrile in water containing 0.1% TFA) in 4
min with a flow rate of 3 mL/min and at 35 °C. TLC analysis were
performed using precoated silica gel Merck 60 F254 glass plates, and
spots were detected visually under UV irradiation (254 or 366 nm) or
after spraying with phosphomolybdic reagent (Aldrich) followed by
heating at 100 °C. Flash column chromatography purifications were
done using a medium-pressure Combi-Flash Companion (Isco Inc.)
instrument with commercial prepacked silica gel cartridges. Reverse
phase prep-HPLC purifications were performed using an optimized
gradient elution (acetonitrile/water containing 0.1% TFA) with a
Waters HPLC prep-system equipped with a UV detector Waters 2487
dual λ absorbance detector, a MS detector Waters Micromass ZQ, and
a reverse phase column SunFire Prep, C-18 OBD, 100 mm × 30 mm,
2-Phenyl-1-(piperazin-1-yl)ethanone (6). To a solution of 5 (6.34
g, 20.83 mmol) in CH2Cl2 (25 mL) was added TFA (24.07 mL, 312.0
mmol) at RT. The mixture was stirred for 30 min and evaporated to
dryness. The residue was dissolved in CH2Cl2, and the solution was
washed with Na2CO3 2 M in water (2×). The organic layer was dried
over Na2SO4, filtered, and evaporated to dryness to yield 6 (3.4 g,
16.64 mmol, 80%) as a yellow resin, which was used in the next step
1
without further purification. LC-MS (ESI): m/z 205.2 [M + H]+. H
NMR (400 MHz, DMSO-d6): 2.55−2.61 (m, 2H) 2.61−2.67 (m, 2H)
3.38−3.42 (m, 4H) 3.69 (s, 2H) 7.19−7.25 (m, 3H) 7.27−7.33 (m,
2H).
2-Chloro-1-(4-(2-phenylacetyl)piperazin-1-yl)ethanone (7). To a
solution of 6 (1.5 g, 7.34 mmol) and Et3N (2.05 mL, 14.69 mmol) in
CH2Cl2 (15 mL) was added dropwise chloroacetyl chloride (0.702
mL, 8.81 mmol) at 0 °C (ice bath). After the addition, the reaction
mixture was allowed to warm to RT and was stirred for 30 min. The
1
5 μm, or 100 mm × 19 mm, 5 μm. H NMR were measured on a
3365
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
mixture was diluted with CH2Cl2 and washed successively with HCl
2N in water (2×), Na2CO3 2 M in water (2×) and brine. The organic
layer was dried over Na2SO4, filtered, and evaporated to dryness. The
crude was purified by silica gel column chromatography (gradient
elution, 0.5−10% MeOH in CH2Cl2) to yield 7 (1.17 g, 4.17 mmol,
57%) as a brownish resin. TLC: Rf 1= 0.56 (CH2Cl2/MeOH 9:1). LC-
MS (ESI): m/z 281.2 [M + H]+. H NMR (400 MHz, DMSO-d6):
3.35−3.56 (m, 8H) 3.73−3.79 (m, 2H) 4.40 (s, 2H) 7.20−7.26 (m,
3H) 7.28−7.34 (m, 2H).
3.74 (m, 1H) 3.75−3.82 (m, 2H) 7.21−7.28 (m, 3H) 7.29−7.42 (m,
3H) 7.47 (d, J = 15.2, 1H) 7.64−7.75 (m, 2H) 8.12 (s, 1H).
(E)-3-(3-Fluoro-4-methylphenyl)-1-(4-(2-phenylacetyl)piperazin-
1-yl)prop-2-en-1-one (3e). Compound 3e (45.6 mg, 0.125 mmol,
85%) was synthesized from 6 (30 mg, 0.147 mmol) and (E)-3-(3-
fluoro-4-methylphenyl)acrylic acid (26.5 mg 0.147 mmol) as a white
solid, in a manner similar to 3a. LC-MS (ESI): m/z 367.4 [M + H]+.
1H NMR (400 MHz, DMSO-d6): 2.25 (s, 3H) 3.47−3.57 (m, 6H)
3.59−3.65 (m, 1H) 3.67−3.73 (m, 1H) 3.78 (br s, 2H) 7.21−7.35 (m,
7H) 7.40−7.49 (m, 2H) 7.62 (d, J = 11.4, 1H).
2-(4-Chlorophenoxy)-1-(4-(2-phenylacetyl)piperazin-1-yl)-
ethanone (2a). To a solution of 7 (30 mg, 0.107 mmol) in DMF (0.5
mL) were successively added K2CO3 (29.5 mg, 0.214 mmol) and 4-
chlorophenol (16.5 mg 0.128 mmol), and the mixture was stirred at
RT for 16 h. The reaction mixture was filtered through a Millipore
0.20 μM filter, and the filter cake was washed with DMF (0.5 mL).
The filtrate was directly subjected to purification by reverse phase
prep-HPLC (Waters system) to yield 2a (27.6 mg, 0.074 mmol, 69%)
(E)-tert-Butyl 4-(3-(4-Methoxyphenyl)acryloyl)piperazine-1-car-
boxylate (8a). To a solution of tert-butyl piperazine-1-carboxylate
(6.07 g, 32.6 mmol) in DMF (100 mL) were successively added Et3N
(9.08 mL, 65.2 mmol), (E)-3-(4-methoxyphenyl)acrylic acid (6.10 g,
34.2 mmol), and HATU (13.63 g, 35.8 mmol) at RT, and the mixture
was stirred for 2 h. The mixture was diluted with Et2O and washed
successively with HCl 2N in water (2×) and Na2CO3 2 M in water
(2×). The organic layer was dried over Na2SO4, filtered, and
evaporated to dryness to yield 8a (10.85 g, 31.3 mmol, 96%) as a
white solid, which was used in the next step without further
purification. LC-MS (ESI): m/z 347.3 [M + H]+. 1H NMR (400
MHz, DMSO-d6): 1.42 (s, 9H) 3.33−3.39 (m, 4H) 3.50−3.60 (m,
2H) 3.63−3.72 (m, 2H) 3.79 (s, 3H) 6.96 (d, J = 8.8, 2H) 7.10 (d, J =
15.4, 1H) 7.47 (d, J = 15.2, 1H) 7.67 (d, J = 8.8, 2H).
1
as a light-yellow solid. LC-MS (ESI): m/z 373.3 [M + H]+. H NMR
(400 MHz, DMSO-d6): 3.35−3.39 (m, 2H) 3.40−3.44 (m, 2H) 3.45−
3.57 (m, 4H) 3.76 (d, J = 6.8, 2H) 4.86 (s, 2H) 6.94 (d, J = 8.8, 2H)
7.22−7.26 (m, 3H) 7.29−7.33 (m, 4H).
2-(4-Bromophenoxy)-1-(4-(2-phenylacetyl)piperazin-1-yl)-
ethanone (2b). Compound 2b (33 mg, 0.079 mmol, 74%) was
synthesized from 7 (30 mg, 0.107 mmol) and 4-bromophenol (22.2
mg 0.128 mmol) as a yellow resin, in a manner similar to 2a. LC-MS
(ESI): m/z 419.2 [M + H]+. 1H NMR (400 MHz, DMSO-d6): 3.35−
3.58 (m, 8H) 3.72−3.79 (m, 2H) 4.85 (s, 2H) 6.90 (d, J = 8.7, 2H)
7.20−7.27 (m, 3H) 7.29−7.36 (m, 2H) 7.44 (d, J = 8.8, 2H).
2-(3,4-Dimethylphenoxy)-1-(4-(2-phenylacetyl)piperazin-1-yl)-
ethanone (2c). Compound 2c (34.9 mg, 0.095 mmol, 89%) was
synthesized from 7 (30 mg, 0.107 mmol) and 3,4-dimethylphenol
(15.7 mg 0.128 mmol) as a white solid, in a manner similar to 2a. LC-
(E)-tert-Butyl 4-(3-(4-Chlorophenyl)acryloyl)piperazine-1-carbox-
ylate (8b). Intermediate 8b (2.38 g, 6.78 mmol, 63%) was synthesized
from tert-butyl piperazine-1-carboxylate (2.0 g, 10.74 mmol) and (E)-
3-(4-chlorophenyl) acrylic acid (1.96 g, 10.74 mmol) as a white solid,
1
in a manner similar to 8a. LC-MS (ESI): m/z 351.4 [M + H]+. H
NMR (400 MHz, DMSO-d6): 1.42 (s, 9H) 3.33−3.39 (m, 4H) 3.52−
3.59 (m, 2H), 3.66−3.72 (m, 2H) 7.29 (d, J = 15.4, 1H) 7.43−7.53
(m, 3H) 7.77 (d, J = 8.6, 2H).
1
MS (ESI): m/z 367.3 [M + H]+. H NMR (400 MHz, DMSO-d6):
(E)-tert-Butyl 4-(3-(4-Bromophenyl)acryloyl)piperazine-1-carbox-
ylate (8c). Intermediate 8c (3.36 g, 8.50 mmol, 79%) was synthesized
from tert-butyl piperazine-1-carboxylate (2.0 g, 10.74 mmol) and (E)-
3-(4-bromophenyl) acrylic acid (2.44 g, 10.74 mmol) as a white solid,
2.13 (s, 3H) 2.17 (s, 3H) 3.35−3.56 (m, 8H) 3.72−3.78 (m, 2H) 4.74
(s, 2H) 6.63 (dd, J = 8.3, 2.2, 1H) 6.73 (br s, 1H) 7.01 (d, J = 8.3 Hz,
1H) 7.20−7.26 (m, 3H) 7.28−7.35 (m, 2H).
1
in a manner similar to 8a. LC-MS (ESI): m/z 397.3 [M + H]+. H
(E)-3-(4-Bromophenyl)-1-(4-(2-phenylacetyl)piperazin-1-yl)prop-
2-en-1-one (3a). To a solution of 6 (30 mg, 0.147 mmol) in DMF
(0.5 mL) were successively added Et3N (0.041 mL, 0.294 mmol), (E)-
3-(4-bromophenyl)acrylic acid (33.3 mg, 0.147 mmol), and HATU
(67.0 mg, 0.176 mmol) at RT, and the mixture was stirred for 3 h. The
reaction mixture was directly subjected to purification by reverse phase
prep-HPLC (Waters system) to yield 3a (31.9 mg, 0.072 mmol, 49%)
NMR (400 MHz, DMSO-d6): 1.42 (s, 9H) 3.33−3.42 (m, 4H) 3.47−
3.62 (m, 2H) 3.63−3.76 (m, 2H) 7.30 (d, J = 15.4, 1H) 7.47 (d, J =
15.4, 1H) 7.61 (d, J = 8.3, 2H) 7.69 (d, J = 8.3, 2H).
(E)-3-(4-Methoxyphenyl)-1-(piperazin-1-yl)prop-2-en-1-one (9a).
To a solution of 8a (10.85 g, 31.3 mmol) in dioxane (100 mL) was
added HCl 4N in dioxane (66 mL, 264 mmol) at RT, and the mixture
was stirred for 16 h. The resulting precipitate was filtered, washed with
Et2O, collected, and dried under high vacuum to yield the HCl salt of
9a (8.88 g, 31.3 mmol, quant) as an off-white solid. LC-MS (ESI): m/z
247.2 [M + H]+. 1H NMR (400 MHz, DMSO-d6): 3.06−3.17 (m, 4H)
3.80 (s, 3H) 3.81−4.06 (m, 4H) 6.98 (d, J = 8.8, 2H) 7.12 (d, J = 15.4,
1H) 7.50 (d, J = 15.2, 1H) 7.68 (d, J = 8.8, 2H) 9.40−9.48 (m, 1H).
(E)-3-(4-Chlorophenyl)-1-(piperazin-1-yl)prop-2-en-1-one (9b).
Intermediate 9b (1.74 g, 6.06 mmol, 89%) was synthesized from 8b
(2.38 g, 6.78 mmol) as a white solid (HCl salt), in a manner similar to
9a. LC-MS (ESI): m/z 251.3 [M + H]+. 1H NMR (400 MHz, DMSO-
d6): 3.03−3.22 (m, 4H) 3.76−3.86 (m, 2H) 3.90−4.03 (m, 2H) 7.31
(d, J = 15.4, 1H) 7.44−7.58 (m, 3H) 7.78 (d, J = 8.4, 2H) 9.27−9.48
(m, 2H).
(E)-3-(4-Bromophenyl)-1-(piperazin-1-yl)prop-2-en-1-one (9c).
Intermediate 9c (2.58 g, 7.78 mmol, 92%) was synthesized from 8c
(3.36 g, 8.5 mmol) as a white solid (HCl salt), in a manner similar to
9a. LC-MS (ESI): m/z 295.2 [M + H]+. 1H NMR (400 MHz, DMSO-
d6): 3.05−3.21 (m, 4H) 3.75−3.84 (m, 2H) 3.89−4.00 (m, 2H) 7.32
(d, J = 15.4, 1H) 7.50 (d, J = 15.3, 1H) 7.62 (d, J = 8.7, 2H) 7.71 (d, J
= 8.1, 2H) 9.23−9.41 (m, 2H).
1
as a white solid. LC-MS (ESI): m/z 415.2 [M + H]+. H NMR (400
MHz, DMSO-d6): 3.47−3.57 (m, 6H) 3.59−3.65 (m, 1H) 3.66−3.72
(m, 1H) 3.77 (br s, 2H) 7.21−7.27 (m, 3H) 7.27−7.35 (m, 3H) 7.47
(d, J = 15.4, 1H) 7.61 (d, J = 8.4, 2H) 7.70 (d, J = 8.4, 2H).
(E)-1-(4-(2-Phenylacetyl) piperazin-1-yl)-3-(p-tolyl)prop-2-en-1-
one (3b). Compound 3b (31.0 mg, 0.089 mmol, 61%) was synthesized
from 6 (30 mg, 0.147 mmol) and (E)-3-(p-tolyl)acrylic acid (23.8 mg
0.147 mmol) as a white solid, in a manner similar to 3a. LC-MS (ESI):
1
m/z 349.3 [M + H]+. H NMR (400 MHz, DMSO-d6): 2.33 (s, 3H)
3.47−3.58 (m, 6H) 3.59−3.64 (m, 1H) 3.66−3.72 (m, 1H) 3.77 (br s,
2H) 7.16−7.28 (m, 6H) 7.29−7.36 (m, 2H) 7.47 (d, J = 15.4, 1H)
7.61 (d, J = 8.0, 2H).
(E)-3-(4-Methoxyphenyl)-1-(4-(2-phenylacetyl)piperazin-1-yl)-
prop-2-en-1-one (3c). Compound 3c (24.8 mg, 0.068 mmol, 48%)
was synthesized from 6 (30 mg, 0.147 mmol) and (E)-3-(4-
methoxyphenyl)acrylic acid (26.2 mg 0.147 mmol) as a white solid,
1
in a manner similar to 3a. LC-MS (ESI): m/z 365.3 [M + H]+. H
NMR (400 MHz, DMSO-d6): 3.48−3.57 (m, 6H) 3.59−3.64 (m, 1H)
3.65−3.71 (m, 1H) 3.77 (br s, 2H) 3.80 (s, 3H) 6.96 (d, J = 8.7, 2H)
7.10 (d, J = 15.4, 1H) 7.20−7.28 (m, 3H) 7.29−7.35 (m, 2H) 7.46 (d,
J = 15.4, 1H) 7.67 (d, J = 8.6, 2H).
(E)-3-(3,4-Dichlorophenyl)-1-(4-(2-phenylacetyl)piperazin-1-yl)-
prop-2-en-1-one (3d). Compound 3d (50.0 mg, 0.124 mmol, 84%)
was synthesized from 6 (30 mg, 0.147 mmol) and (E)-3-(3,4-
dichlorophenyl)acrylic acid (31.9 mg 0.147 mmol) as a white solid, in
a manner similar to 3a. LC-MS (ESI): m/z 403.3 [M + H]+. 1H NMR
(400 MHz, DMSO-d6): 3.47−3.58 (m, 6H) 3.60−3.66 (m, 1H) 3.67−
(E)-3-(4-Methoxyphenyl)-1-(4-(3-methylbenzoyl)piperazin-1-yl)-
prop-2-en-1-one (4a). To a solution of 9a (34 mg, 0.120 mmol) in
DMF (0.5 mL) were successively added Et3N (0.050 mL, 0.361
mmol), 3-methylbenzoic acid (16.6 mg, 0.120 mmol), and HATU
(54.9 mg, 0.144 mmol) at RT, and the mixture was stirred for 3 h. The
reaction mixture was directly subjected to purification by reverse phase
prep-HPLC (Waters system) to yield 4a (29.5 mg, 0.081 mmol, 67%)
3366
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
as a yellow resin. LC-MS (ESI): m/z 365.3 [M + H]+. 1H NMR (400
MHz, DMSO-d6): 2.35 (s, 3H) 3.49−3.70 (m, 8H) 3.80 (s, 3H) 6.97
(d, J = 8.6, 2H) 7.07−7.17 (m, 1H) 7.20−7.26 (m, 2H) 7.27−7.32 (m,
1H) 7.32−7.38 (m, 1H) 7.48 (d, J = 15.3, 1H) 7.63−7.71 (m, 2H).
(E)-3-(4-Methoxyphenyl)-1-(4-(4-(trifluoromethyl)benzoyl)-
piperazin-1-yl)prop-2-en-1-one (4b). Compound 4b (17.5 mg, 0.042
mmol, 39%) was synthesized from 9a (30 mg, 0.106 mmol) and 4-
(trifluoromethyl)benzoic acid (21.2 mg 0.110 mmol) as a white solid,
(m, 6H) 3.80 (s, 3H) 6.97 (d, J = 8.6, 2H) 7.06−7.20 (m, 1H) 7.29 (d,
J = 8.0, 1H) 7.42−7.54 (m, 3H) 7.67 (d, J = 7.6, 2H).
(E)-1-(4-(3,4-Dimethylbenzoyl)piperazin-1-yl)-3-(4-
methoxyphenyl)prop-2-en-1-one (4j). Compound 4j (22.1 mg, 0.058
mmol, 55%) was synthesized from 9a (30 mg, 0.106 mmol) and 3,4-
dimethylbenzoic acid (16.7 mg 0.110 mmol) as a white solid, in a
1
manner similar to 4a. LC-MS (ESI): m/z 379.5 [M + H]+. H NMR
(400 MHz, DMSO-d6): 2.26 (s, 6H) 3.36−3.78 (m, 8H) 3.79 (s, 3H)
6.97 (d, J = 8.6, 2H) 7.06−7.19 (m, 2H) 7.19−7.25 (m, 2H) 7.48 (d, J
= 15.3, 1H) 7.67 (d, J = 8.0, 2H).
(E)-3-(4-Chlorophenyl)-1-(4-(2,3-dihydrobenzofuran-6-carbonyl)-
piperazin-1-yl)prop-2-en-1-one (4k). Compound 4k (29.2 mg, 0.074
mmol, 40%) was synthesized from 9b (52.6 mg, 0.183 mmol) and 2,3-
dihydrobenzofuran-5-carboxylic acid (30.0 mg 0.183 mmol) as a white
solid, in a manner similar to 4a. LC-MS (ESI): m/z 397.3 [M + H]+.
1H NMR (400 MHz, DMSO-d6): 3.22 (t, J = 8.7, 2H) 3.47−3.67 (m,
6H) 3.73−3.82 (m, 2H) 4.59 (t, J = 8.7, 2H) 6.81 (d, J = 8.2, 1H) 7.21
(d, J = 8.2, 1H) 7.27−7.37 (m, 2H) 7.45−7.55 (m, 3H) 7.77 (d, J =
8.4, 2H).
(E)-3-(4-Bromophenyl)-1-(4-(2,3-dihydrobenzofuran-6-carbonyl)-
piperazin-1-yl)prop-2-en-1-one (4l). Compound 4l (41.9 mg, 0.095
mmol, 52%) was synthesized from 9c (60.7 mg, 0.183 mmol) and 2,3-
dihydrobenzofuran-5-carboxylic acid (30.0 mg 0.183 mmol) as a white
solid, in a manner similar to 4a. LC-MS (ESI): m/z 443.2 [M + H]+.
1H NMR (400 MHz, DMSO-d6): 3.21 (t, J = 8.7, 2H) 3.48−3.58 (m,
1
in a manner similar to 4a. LC-MS (ESI): m/z 419.4 [M + H]+. H
NMR (400 MHz, DMSO-d6): 3.28−3.36 (m, 2H) 3.55−3.75 (m, 5H)
3.76−3.86 (m, 4H) 6.97 (d, J = 8.3, 2H) 7.05−7.22 (m, 1H) 7.48 (d, J
= 15.3, 1H) 7.62−7.73 (m, 4H) 7.85 (d, J = 8.0, 2H).
(E)-1-(4-(4-Bromobenzoyl)piperazin-1-yl)-3-(4-methoxyphenyl)-
prop-2-en-1-one (4c). Compound 4c (31.2 mg, 0.073 mmol, 60%)
was synthesized from 9a (34 mg, 0.120 mmol) and 4-bromobenzoic
acid (24.5 mg 0.120 mmol) as a white solid, in a manner similar to 4a.
LC-MS (ESI): m/z 431.2 [M + H]+. 1H NMR (400 MHz, DMSO-d6):
3.35−3.41 (m, 2H) 3.53−3.74 (m, 5H) 3.75−3.84 (m, 4H) 6.97 (d, J
= 8.7, 2H) 7.05−7.19 (m, 1H) 7.41 (d, J = 8.2, 2H) 7.48 (d, J = 15.3,
1H) 7.62−7.73 (m, 4H).
(E)-3-(4-Methoxyphenyl)-1-(4-(4-methylbenzoyl)piperazin-1-yl)-
prop-2-en-1-one (4d). Compound 4d (30.7 mg, 0.084 mmol, 70%)
was synthesized from 9a (34 mg, 0.120 mmol) and 4-methylbenzoic
acid (16.6 mg 0.120 mmol) as a white solid, in a manner similar to 4a.
LC-MS (ESI): m/z 365.3 [M + H]+. 1H NMR (400 MHz, DMSO-d6):
2.36 (s, 3H) 3.35−4.00 (m, 11H) 6.97 (d, J = 8.6, 2H) 7.06−7.17 (m,
1H) 7.24−7.30 (m, 2H) 7.31−7.37 (m, 2H) 7.48 (d, J = 15.3, 1H)
7.67 (d, J = 7.7, 2H).
4H) 3.58−3.65 (m, 2H) 3.73−3.80 (m, 2H) 4.58 (t, J = 8.7, 2H) 6.81
(d, J = 8.3, 1H) 7.21 (dd, J = 8.2, 1.6, 1H) 7.28−7.36 (m, 2H) 7.49 (d,
J = 15.4, 1H) 7.60−7.63 (m, 2H) 7.68−7.71 (m, 2H).
(E)-3-(4-Bromophenyl)-1-(4-(4-methoxybenzoyl)piperazin-1-yl)-
prop-2-en-1-one (4m). To a solution of (4-methoxyphenyl)-
(piperazin-1-yl)methanone (0.970 g, 4.40 mmol) in DMF (17 mL)
were successively added N-methylmorpholine (2.4 mL, 22.02 mmol),
(E)-3-(4-bromophenyl)acrylic acid (1.0 g, 4.40 mmol), and HATU
(1.84 g, 4.84 mmol) at RT, and the mixture was stirred for 16 h. The
mixture was diluted with AcOEt, and the solution was washed
successively with HCl 2N in water and Na2CO3 2 M in water and
brine. The organic layer was dried over Na2SO4, filtered, and
evaporated to dryness. The resulting crude residue was purified by
silica gel column chromatography (gradient elution, 0.5−10% MeOH
in CH2Cl2) to give the title compound as a light-yellow solid. The
solid was recrystallized in AcOEt/MeOH (1:1, 100 mL) to yield 4m
(1.30 g, 3.03 mmol, 69%) as a colorless crystalline solid. LC-MS (ESI):
m/z 431.2 [M + H]+. 1H NMR (400 MHz, DMSO-d6): 3.47−3.59 (m,
3H) 3.59−3.66 (m, 2H) 3.73−3.79 (m, 2H) 3.81 (s, 3H) 7.00 (d, J =
8.6, 2H) 7.32 (d, J = 15.4, 1H) 7.42 (d, J = 8.6, 2H) 7.49 (d, J = 15.4,
1H) 7.59−7.64 (m, 2H) 7.67−7.73 (m, 2H).
(E)-1-(4-(4-Methoxybenzoyl)piperazin-1-yl)-3-(4-
methoxyphenyl)prop-2-en-1-one (4e). Compound 4e (35.2 mg,
0.093 mmol, 77%) was synthesized from 9a (34 mg, 0.120 mmol) and
4-methoxybenzoic acid (18.5 mg 0.120 mmol) as a white solid, in a
1
manner similar to 4a. LC-MS (ESI): m/z 381.3 [M + H]+. H NMR
(400 MHz, DMSO-d6): 3.47−3.58 (m, 4H) 3.59−3.67 (m, 2H) 3.72−
3.78 (m, 2H) 3.80 (s, 3H) 3.81 (s, 3H) 6.97 (d, J = 8.8, 2H) 7.01 (d, J
= 8.8, 2H) 7.12 (d, J = 15.2, 1H) 7.42 (d, J = 8.8, 2H) 7.49 (d, J = 15.4,
1H) 7.67 (d, J = 8.6, 2H).
(E)-1-(4-(4-(Dimethylamino)benzoyl)piperazin-1-yl)-3-(4-
methoxyphenyl)prop-2-en-1-one (4f). Compound 4f (15.5 mg, 0.039
mmol, 37%) was synthesized from 9a (30 mg, 0.106 mmol) and 4-
(dimethylamino)benzoic acid (18.4 mg 0.110 mmol) as a white solid,
1
in a manner similar to 4a. LC-MS (ESI): m/z 394.4 [M + H]+. H
NMR (400 MHz, DMSO-d6): 2.96 (s, 6H) 3.48−3.66 (m, 6H) 3.70−
3.78 (m, 2H) 3.80 (s, 3H) 6.73 (d, J = 8.7, 2H) 6.97 (d, J = 8.6, 2H)
7.14 (d, J = 15.4, 1H) 7.32 (d, J = 8.7, 2H) 7.48 (d, J = 15.3, 1H) 7.67
(d, J = 8.6, 2H).
(E)-Methyl 4-(4-(3-(4-Methoxyphenyl)acryloyl)piperazine-1-
carbonyl)benzoate (4g). Compound 4g (11.8 mg, 0.029 mmol,
27%) was synthesized from 9a (30 mg, 0.106 mmol) and 4-
(methoxycarbonyl)benzoic acid (20.1 mg 0.110 mmol) as a white
solid, in a manner similar to 4a. LC-MS (ESI): m/z 409.4 [M + H]+.
1H NMR (400 MHz, DMSO-d6): 3.52−3.84 (m, 8H) 3.79 (s, 3H)
3.89 (s, 3H) 6.97 (d, J = 8.3, 2H) 7.04−7.21 (m, 1H) 7.48 (d, J = 15.4,
1H) 7.59 (d, J = 8.1, 2H) 7.63−7.73 (m, 2H) 8.04 (d, J = 8.1, 2H).
(E)-1-(4-(2,3-Dihydrobenzofuran-5-carbonyl)piperazin-1-yl)-3-(4-
methoxyphenyl)prop-2-en-1-one (4h). Compound 4h (14.5 mg,
0.037 mmol, 35%) was synthesized from 9a (30 mg, 0.106 mmol) and
2,3-dihydrobenzofuran-5-carboxylic acid (18.3 mg 0.110 mmol) as a
white solid, in a manner similar to 4a. LC-MS (ESI): m/z 393.4 [M +
H]+. 1H NMR (400 MHz, DMSO-d6): 3.21 (t, J = 8.8, 2H) 3.46−3.58
(m, 4H) 3.58−3.66 (m, 2H) 3.71−3.78 (m, 2H) 3.80 (s, 3H) 4.58 (t, J
= 8.7, 2H) 6.81 (d, J = 8.2, 1H) 6.97 (d, J = 8.7, 2H) 7.13 (d, J = 15.4,
1H) 7.21 (d, J = 8.1, 1H) 7.34 (s, 1H) 7.48 (d, J = 15.4, 1H) 7.67 (d, J
= 8.6, 2H).
ASSOCIATED CONTENT
* Supporting Information
Dose−response curves for key compounds measuring release of
intracellular calcium in CHO cells expressing human EBI2
receptor. [3H] 7α,25-OHC radioligand binding data for all
compounds at the human EBI2 receptor. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*Phone: +41 79 5500941. Fax: +41 61 3246105. E-mail:
andreas.sailer@novartis.com. Address: Developmental & Mo-
lecular Pathways, Novartis Institutes for BioMedical Research,
Forum 1, Novartis Campus, WSJ-355.4025.8 4002 Basel,
Switzerland.
■
(E)-1-(4-(4-Chloro-3-methylbenzoyl)piperazin-1-yl)-3-(4-
methoxyphenyl)prop-2-en-1-one (4i). Compound 4i (23.0 mg, 0.058
mmol, 54%) was synthesized from 9a (30 mg, 0.106 mmol) and 4-
chloro-3-methylbenzoic acid (19.0 mg 0.110 mmol) as a white solid, in
a manner similar to 4a. LC-MS (ESI): m/z 399.4 [M + H]+. 1H NMR
(400 MHz, DMSO-d6): 2.37 (s, 3H) 3.35−3.44 (m, 2H) 3.52−3.78
Present Addresses
#For D.G.N: Organovo, Inc., 6275 Nancy Ridge Drive, San
Diego, California, 92121, United States.
∇For Y.A.C.: NGM Biopharmaceuticals, Inc., 630 Gateway
Blvd., South San Francisco, California, 94080, United States.
3367
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368

Journal of Medicinal Chemistry
Article
RORalpha and RORgamma activity by 24S-hydroxycholesterol
(cerebrosterol). Biochim. Biophys. Acta 2010, 1801, 917−923.
(9) Kallen, J.; Schlaeppi, J. M.; Bitsch, F.; Delhon, I.; Fournier, B.
Crystal structure of the human RORalpha ligand binding domain in
complex with cholesterol sulfate at 2.2 A. J. Biol. Chem. 2004, 279,
14033−14038.
Notes
The authors declare the following competing financial
interest(s): All authors, except for M. M. Rosenkilde, are
current or previous employees of Novartis and some of them
do hold stock or stock options in their company.
(10) Kallen, J. A.; Schlaeppi, J. M.; Bitsch, F.; Geisse, S.; Geiser, M.;
Delhon, I.; Fournier, B. X-ray structure of the hRORalpha LBD at 1.63
A: structural and functional data that cholesterol or a cholesterol
derivative is the natural ligand of RORalpha. Structure 2002, 10, 1697−
1707.
(11) Raychaudhuri, S.; Prinz, W. A. The diverse functions of
oxysterol-binding proteins. Annu. Rev. Cell Dev. Biol. 2010, 26, 157−
177.
(12) Ohgane, K.; Karaki, F.; Dodo, K.; Hashimoto, Y. Discovery of
oxysterol-derived pharmacological chaperones for NPC1: implication
for the existence of second sterol-binding site. Chem. Biol. 2013, 20,
391−402.
(13) Dwyer, J. R.; Sever, N.; Carlson, M.; Nelson, S. F.; Beachy, P. A.;
Parhami, F. Oxysterols are novel activators of the hedgehog signaling
pathway in pluripotent mesenchymal cells. J. Biol. Chem. 2007, 282,
8959−8968.
ACKNOWLEDGMENTS
■
We thank Emilie Joly, Pierre Nimsgern, and Inger Smith
Simonson for their excellent technical assistance. We are
grateful to Drs. Gunther Scheel, Shin Numao, and John Priestle
̈
for supporting activities closely connected to this project.
ABBREVIATIONS USED
■
7α,25-OHC, 7α,25-dihydroxycholesterol; RHS, right-hand side;
LHS, left-hand side; EBI2, Epstein−Barr virus-induced gene 2;
Clint, intrinsic clearance; SD, standard deviation; CHO, Chinese
hamster ovary; SPA, scintillation proximity assay; WGA, wheat
germ agglutinin; PVT, polyvinyltoluene; Vss, apparent volume
of distribution at steady state; Cmax, maximal blood concen-
tration after po administration; Tmax, time of peak blood
concentration after po administration; HATU, (1-[bis-
(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]-
pyridinium3-oxide hexafluorophosphate)
(14) Nachtergaele, S.; Mydock, L. K.; Krishnan, K.; Rammohan, J.;
Schlesinger, P. H.; Covey, D. F.; Rohatgi, R. Oxysterols are allosteric
activators of the oncoprotein Smoothened. Nature Chem. Biol. 2012, 8,
211−220.
(15) Nedelcu, D.; Liu, J.; Xu, Y.; Jao, C.; Salic, A. Oxysterol binding
to the extracellular domain of Smoothened in Hedgehog signaling.
Nature Chem. Biol. 2013, 9, 557−564.
REFERENCES
■
(1) Hannedouche, S.; Zhang, J.; Yi, T.; Shen, W.; Nguyen, D.;
Pereira, J. P.; Guerini, D.; Baumgarten, B. U.; Roggo, S.; Wen, B.;
Knochenmuss, R.; Noel, S.; Gessier, F.; Kelly, L. M.; Vanek, M.;
Laurent, S.; Preuss, I.; Miault, C.; Christen, I.; Karuna, R.; Li, W.; Koo,
D. I.; Suply, T.; Schmedt, C.; Peters, E. C.; Falchetto, R.; Katopodis,
A.; Spanka, C.; Roy, M. O.; Detheux, M.; Chen, Y. A.; Schultz, P. G.;
Cho, C. Y.; Seuwen, K.; Cyster, J. G.; Sailer, A. W. Oxysterols direct
immune cell migration via EBI2. Nature 2011, 475, 524−527.
(2) Liu, C.; Yang, X. V.; Wu, J.; Kuei, C.; Mani, N. S.; Zhang, L.; Yu,
J.; Sutton, S. W.; Qin, N.; Banie, H.; Karlsson, L.; Sun, S.; Lovenberg,
T. W. Oxysterols direct B-cell migration through EBI2. Nature 2011,
475, 519−523.
(3) Rosenkilde, M. M.; Benned-Jensen, T.; Andersen, H.; Holst, P. J.;
Kledal, T. N.; Luttichau, H. R.; Larsen, J. K.; Christensen, J. P.;
Schwartz, T. W. Molecular pharmacological phenotyping of EBI2. An
orphan seven-transmembrane receptor with constitutive activity. J.
Biol. Chem. 2006, 281, 13199−13208.
(4) Heinig, M.; Petretto, E.; Wallace, C.; Bottolo, L.; Rotival, M.; Lu,
H.; Li, Y.; Sarwar, R.; Langley, S. R.; Bauerfeind, A.; Hummel, O.; Lee,
Y. A.; Paskas, S.; Rintisch, C.; Saar, K.; Cooper, J.; Buchan, R.; Gray, E.
E.; Cyster, J. G.; Erdmann, J.; Hengstenberg, C.; Maouche, S.;
Ouwehand, W. H.; Rice, C. M.; Samani, N. J.; Schunkert, H.; Goodall,
A. H.; Schulz, H.; Roider, H. G.; Vingron, M.; Blankenberg, S.;
Munzel, T.; Zeller, T.; Szymczak, S.; Ziegler, A.; Tiret, L.; Smyth, D. J.;
Pravenec, M.; Aitman, T. J.; Cambien, F.; Clayton, D.; Todd, J. A.;
Hubner, N.; Cook, S. A. A trans-acting locus regulates an anti-viral
expression network and type 1 diabetes risk. Nature 2010, 467, 460−
464.
(5) Gill, S.; Chow, R.; Brown, A. J. Sterol regulators of cholesterol
homeostasis and beyond: the oxysterol hypothesis revisited and
revised. Prog. Lipid Res. 2008, 47, 391−404.
(6) Janowski, B. A.; Shan, B.; Russell, D. W. The hypocholester-
olemic agent LY295427 reverses suppression of sterol regulatory
element-binding protein processing mediated by oxysterols. J. Biol.
Chem. 2001, 276, 45408−45416.
(7) Janowski, B. A.; Grogan, M. J.; Jones, S. A.; Wisely, G. B.;
Kliewer, S. A.; Corey, E. J.; Mangelsdorf, D. J. Structural requirements
of ligands for the oxysterol liver X receptors LXRalpha and LXRbeta.
Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 266−271.
(8) Wang, Y.; Kumar, N.; Crumbley, C.; Griffin, P. R.; Burris, T. P. A
second class of nuclear receptors for oxysterols: regulation of
(16) Yi, T.; Wang, X.; Kelly, L. M.; An, J.; Xu, Y.; Sailer, A. W.;
Gustafsson, J. A.; Russell, D. W.; Cyster, J. G. Oxysterol gradient
generation by lymphoid stromal cells guides activated B cell movement
during humoral responses. Immunity 2012, 37, 535−548.
(17) Benned-Jensen, T.; Smethurst, C.; Holst, P. J.; Page, K. R.;
Sauls, H.; Sivertsen, B.; Schwartz, T. W.; Blanchard, A.; Jepras, R.;
Rosenkilde, M. M. Ligand modulation of the Epstein−Barr virus-
induced seven-transmembrane receptor EBI2: identification of a
potent and efficacious inverse agonist. J. Biol. Chem. 2011, 286,
29292−29302.
(18) Benned-Jensen, T.; Madsen, C. M.; Arfelt, K. N.; Smethurts, C.;
Blanchard, A.; Jepras, R.; Rosenkilde, M. M. Small molecule
antagonism of oxysterol-induced Epstein−Barr virus induced gene 2
(EBI2) activation. FEBS Open Bio 2013, 3, 156−160.
(19) Hammer, M. M.; Wehrman, T. S.; Blau, H. M. A novel enzyme
complementation-based assay for monitoring G-protein-coupled
receptor internalization. FASEB J. 2007, 21, 3827−3834.
(20) Yin, H.; Chu, A.; Li, W.; Wang, B.; Shelton, F.; Otero, F.;
Nguyen, D. G.; Caldwell, J. S.; Chen, Y. A. Lipid G protein-coupled
receptor ligand identification using beta-arrestin PathHunter assay. J.
Biol. Chem. 2009, 284, 12328−12338.
(21) Eibinger, G.; Fauler, G.; Bernhart, E.; Frank, S.; Hammer, A.;
Wintersperger, A.; Eder, H.; Heinemann, A.; Mischel, P. S.; Malle, E.;
Sattler, W. On the role of 25-hydroxycholesterol synthesis by
glioblastoma cell lines. Implications for chemotactic monocyte
recruitment. Exp. Cell Res. 2013, 319, 1828−1838.
(22) Brown, A. J.; Jessup, W. Oxysterols and atherosclerosis.
Atherosclerosis 1999, 142, 1−28.
(23) Moore, K. J.; Tabas, I. Macrophages in the pathogenesis of
atherosclerosis. Cell 2011, 145, 341−355.
(24) Benned-Jensen, T.; Norn, C.; Laurent, S.; Madsen, C. M.;
Larsen, H. M.; Arfelt, K. N.; Wolf, R. M.; Frimurer, T.; Sailer, A. W.;
Rosenkilde, M. M. Molecular characterization of oxysterol binding to
the Epstein−Barr virus-induced gene 2 (GPR183). J. Biol. Chem. 2012,
287, 35470−35483.
(25) Trunzer, M.; Faller, B.; Zimmerlin, A. Metabolic soft spot
identification and compound optimization in early discovery phases
using MetaSite and LC-MS/MS validation. J. Med. Chem. 2009, 52,
329−335.
3368
dx.doi.org/10.1021/jm4019355 | J. Med. Chem. 2014, 57, 3358−3368