Selection, synthesis, and anti-inflammatory evaluation of the arylidene malonate derivatives as TLR4 signaling inhibitors

Article abstract of DOI:10.1016/j.bmc.2012.08.022

Inhibition of TLR4 signaling is an important therapeutic strategy for intervention in the etiology of several pro-inflammatory diseases. There has been intensive research in recent years aiming to explore this strategy, and identify small molecule inhibitors of the TLR4 pathway. However, the recent failure of a number of advanced drug candidates targeting TLR4 signaling (e.g., TAK242 and Eritoran) prompted us to continue the search for novel chemical scaffolds to inhibit this critical inflammatory response pathway. Here we report the identification of a group of new TLR4 signaling inhibitors through a cell-based screening. A series of arylidene malonate analogs were synthesized and assayed in murine macrophages for their inhibitory activity against LPS-induced nitric oxide (NO) production. The lead compound 1 (NCI126224) was found to suppress LPS-induced production of nuclear factor-kappaB (NF-κB), tumor necrosis factor (TNF-α), interleukin-1β (IL-1β), and nitric oxide (NO) in the nanomolar-low micromolar range. Taken together, this study demonstrates that 1 is a promising potential therapeutic candidate for various inflammatory diseases.

Full text of DOI:10.1016/j.bmc.2012.08.022

Bioorganic & Medicinal Chemistry 20 (2012) 6073–6079
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Selection, synthesis, and anti-inflammatory evaluation of the arylidene
malonate derivatives as TLR4 signaling inhibitors
⇑
Shuting Zhang, Kui Cheng, Xiaohui Wang, Hang Yin
Department of Chemistry and Biochemistry and BioFrontiers Institute, 596 University of Colorado at Boulder, Boulder, CO 80309-0596, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Inhibition of TLR4 signaling is an important therapeutic strategy for intervention in the etiology of several
pro-inflammatory diseases. There has been intensive research in recent years aiming to explore this strat-
egy, and identify small molecule inhibitors of the TLR4 pathway. However, the recent failure of a number
of advanced drug candidates targeting TLR4 signaling (e.g., TAK242 and Eritoran) prompted us to con-
tinue the search for novel chemical scaffolds to inhibit this critical inflammatory response pathway. Here
we report the identification of a group of new TLR4 signaling inhibitors through a cell-based screening. A
series of arylidene malonate analogs were synthesized and assayed in murine macrophages for their
Received 2 June 2012
Revised 11 August 2012
Accepted 19 August 2012
Available online 25 August 2012
Keywords:
Toll-like receptor 4
NF-jB
Arylidene malonate
inhibitory activity against LPS-induced nitric oxide (NO) production. The lead compound
(NCI126224) was found to suppress LPS-induced production of nuclear factor-kappaB (NF- B), tumor
necrosis factor (TNF- ), interleukin-1b (IL-1b), and nitric oxide (NO) in the nanomolar-low micromolar
1
j
a
Anti-inflammation
range. Taken together, this study demonstrates that 1 is a promising potential therapeutic candidate
for various inflammatory diseases.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
family of inhibitory proteins known as Ij
Bs.7 Upon binding to
MD-2 in the presence of LPS, TLR4 initiates a series of phosphory-
The first line of defense in host protection against invading
microbial pathogens is the innate immune system, where the
Toll-like receptors (TLRs) play a critical role.1 TLRs function to de-
tect and respond to a series of structurally conserved molecules
known as pathogen-associated molecular patterns (PAMPs). Re-
sponse to PAMPs by TLRs leads to the up-regulation of pro-inflam-
matory cytokines and mediators, initiating the innate immune
response. The first identified and most well studied TLR is TLR4,
which recognizes lipopolysaccharide (LPS) or endotoxin, a major
component of the outer membrane of Gram-negative bacteria.2–4
LPS-induced TLR4 signal transduction requires the association of
the accessory protein myeloid differentiation factor 2 (MD-2) to
TLR4. Binding of LPS to the large hydrophobic pocket on the MD-
2 surface induces the homodimerization of two copies of the
MD-2–TLR4–LPS complex.5
lation events resulting in the phosphorylation of the cytoplasmic
I
j
Bs. These phosphorylated I
subsequent degradation by the proteasome, resulting in the trans-
location of NF- B promotes
B into the nucleus.8,9 Nuclear NF-
jBs then undergo ubiquitylation and
j
j
transcription of various proinflammatory cytokines including
inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2),
tumor necrosis factor (TNF)-a, interleukin-1b (IL-1b), and IL-
6.10,11 In contrast, TRIF-dependent signaling activates interferon
regulatory factor 3 (IRF3), which induces type I interferon (IFN)
expression.12,13
Although LPS-induced proinflammatory cytokine production
initiates the host defense against injury and infection,14,15 the dys-
regulation of TLR4 signaling contributes to an array of acute and
chronic human diseases such as septic shock, inflammatory arthri-
tis, atherosclerosis, and cancer.16–19 Its involvement in human dis-
ease makes the TLR4 signaling pathway an important therapeutic
target.20 In fact, several TLR4 signaling inhibitors have already been
investigated as potential anti-sepsis drugs. The most advanced of
these, TAK242 and Eritoran,21,22 were successful in pre-clinical tri-
als but both failed in Phase III clinical trials due to lack of effi-
cacy.23,24 Therefore, the identification of new TLR4 signaling
inhibitors which serve as novel therapeutics is still an urgent need.
In the present study, we identified a group of novel TLR4 signal-
ing inhibitors, developed from our initial lead 2-(2-nitrobenzyli-
dene) malonate (1), and investigated their structure-activity
relationship. Further, we examined the inhibitory effects of 1 on
This homodimerization diverges to result in the activation of
either myeloid differentiation primary-response gene 88
(MyD88)-dependent, or Toll/interleukin-1 receptor domain-con-
taining adaptor inducing IFN-b (TRIF)-dependent signaling.6
MyD88-dependent signaling induces NF-
inflammatory response. Under normal circumstances, NF-
j
B
activation as an
B re-
j
mains sequestered in the cytoplasm as an inactive complex by a
⇑
Corresponding author. Tel.: +1 303 492 6786.
E-mail address: Hubert.Yin@Colorado.edu (H. Yin).
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2012.08.022

6074
S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
downstream NF-jB activation to elucidate the mechanism of its
inhibitory effects.
COOMe
COOMe
COOMe
COOMe
Cl
NaH
+
DMF, rt, 2 h
62%
NO2
NO2
22
2. Results and discussion
2.1. Screening for TLR4 signaling inhibitors
Scheme 2. Synthesis of compound 22.
We interrogated the 1363-member Diversity Set II library from
the National Cancer Institute (NCI) for inhibition of LPS-induced
NO production using murine macrophage RAW 264.7 cells as pre-
viously reported.25 This library consists of small molecules that
were selected from the larger 140,000-compound NCI library on
the basis of availability, purity, and other diversity criteria. After
the preliminary screening, a total of six compounds were identified
a
,b-double bond resulted in a complete loss of inhibitory activity
(IC50 >100 M), so compound 22 was used as the negative control
l
in the subsequent biological evaluations. This demonstrates that
the presence of Michael acceptor maybe essential for its TLR4 sig-
naling inhibitory activity. This property is consistent with TAK242,
a known TLR inhibitor that covalently binds to Cys747 in TLR4 via a
Michael reaction.28
that showed >80% inhibitory activity at a concentration of 1.0 lM.
Covalent inhibitors are rarely considered when initiating a tar-
get-directed drug discovery project due to safety concerns. None-
theless, a relatively potent inhibitor may render an activity
window that allows useful applications of specifically regulating
the TLR4-mediated inflammation response. Furthermore, the re-
cent revived interests of covalent drugs also suggest that such mol-
ecules could still serve as promising drug candidates in principle.29
Figure 1A showed a representative dose-response analysis of 1 to
assess the IC50 values for inhibition of LPS-induced NO production
in the RAW 264.7 cells in comparison with 22. The results showed
that 1 potently blocked LPS-induced NO production with an IC50 of
To ensure that the observed inhibition was not due to cell prolifer-
ation inhibition, we used the previously established WST-1 toxicity
assay to determine the cytotoxicity of the six selected hits.26 Only
one compound, NCI126224 (1, Scheme 1) based on an arylidene
malonate scaffold did not show significant toxicity up to
10.0 lM, prompting us to focus on this scaffold for further
investigation.
2.2. Structure-activity relationship studies of the arylidene
malonate derivatives
0.31 0.03 lM. By contrast, compound 22 showed negligible inhi-
bition at the tested concentrations.
Next, we performed structure-activity relationship (SAR) stud-
ies of the selected hit compound, 1, to identify the key structural
features essential for inhibitory activity. The representative syn-
thetic route for 1–21 is shown in Scheme 1. Commercially available
dimethyl malonate underwent a piperidine-catalyzed Knoevenagel
condensation with various aldehydes, affording 1–16 and 21 in
good yields. Then, the resultant Compounds 1–3 and 10 were con-
verted to the corresponding diacids, 17–20, by hydrolysis with
potassium hydroxide. Dimethyl malonate was treated with sodium
hydride, then 2-nitrobenzyl chloride to give the fully saturated
analog 22 (Scheme 2). Compound 23 was obtained from the con-
densation of 2-nitrobenzaldehyde and malonic acid followed by
methyl-esterification (Scheme 3).
Replacement of the phenyl ring with a cyclohexyl ring (21) re-
sulted in reduced activity. Additionally, the monoester (23) was
less potent than the corresponding diester (1). The effect of the
arylidene malonate pharmacophore on TLR4 signaling was also
investigated with regard to the electronic properties of phenyl
substituent. As shown in Table 1, the ortho-substituted arylidene
malonates (1, 3, 5) were remarkably more potent than their meta-
substituted counterparts (14, 15, 16). The corresponding para-
substituted analogs (6, 7, 10) were slightly less potent than the
ortho-substituted analogs. Further SAR studies revealed that the
presence of a strong electron-withdrawing group on the 2-position
of the benzene ring remarkably increased the inhibitory activity.
This suggests that the electron deficiency aromatic ring is critical
for compound efficacy. When the nitro group at the 2-position
was replaced with fluorine (3), no significant inhibitory potency
change was observed. By contrast, the introduction of a methoxy
group to the 2-position (5) led to a decreased inhibitory activity.
As shown in Table 1 and 12 compounds exhibited submicrom-
olar IC50 values for inhibition of LPS-induced NO production in
RAW 264.7 cells. As a comparison, their inhibitory activities were
significantly higher than the widely used anti-inflammatory agent,
curcumin (IC50 = 6 l a,b-double
M).27 To study the influence of the
bond, reduced analog (22) of 1 was evaluated. The reduction of the
COOMe
piperidine/AcOH
CHO
COOMe
COOMe
COOH
COOH
KOH
THF, reflux, 16 h
+
R
R
R
benzene, reflux
COOMe
NCI126224 1
9
R = 2-NO2
R = H
R = 4-COOMe
10 R = 4-OMe
11
17 R = 2-NO2
18
19 R = 2-F
20
2
3
4
5
6
7
8
R = H
R = 2-F
R = 4-N(Me)2
R = 2-Cl
R = 2-OMe
R = 4-NO2
R = 4-F
12 R = 2,4-F
R = 4-OMe
13
14
R = 2,4-OMe
R = 3-NO2
15 R = 3-F
16
R = 4-Cl
R = 3-OMe
COOMe
COOMe
COOMe
piperidine/AcOH
benzene, reflux
CHO
+
COOMe
21
Scheme 1. Generic synthesis route of compounds 1–21.

S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
6075
O
O
COOH
COOH
piperidine
H2SO4
CHO
NO2
OH
OMe
+
MeOH
96%
pyridine, reflux, 3 h
73%
NO2
NO2
23
Scheme 3. Synthesis of compound 23.
Table 1
Inhibitory effects of the arylidene malontate derivatives on LPS-induced NO production in RAW 264.7 macrophage cells
O
O
2
1
COOMe
MeO
O
3
R2
OMe
R1
COOMe
O
4
O
R2
NO2
22
NO2
23
OMe
1-20
21
R1
R2
IC50 (lM)
1 (NCI126224)
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2-NO2
H
2-F
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OH
0.31 0.03
0.82 0.08
0.33 0.14
0.82 0.15
0.74 0.26
0.82 0.04
0.71 0.14
1.54 0.22
1.33 0.05
1.24 0.12
3.11 0.26
0.42 0.08
0.82 0.27
2.13 0.26
2.53 0.31
3.72 0.36
0.52 0.12
0.41 0.16
0.93 0.07
1.44 0.18
78.2 1.6
>100
2-Cl
2-OMe
4-NO2
4-F
4-Cl
4-COOMe
4-OMe
4-N(Me)2
2,4-F
2,4-OMe
3-NO2
3-F
3-OMe
2-NO2
2-F
OH
OH
OH
H
4-OMe
4.53 0.36
Figure 1. (A) Representative dose-dependent inhibitory response of the arylidene malonate analogs on LPS-induced NO production in RAW 264.7 macrophage cells. (B)
Effects of 1 and the negative control, 22, on LPS-induced activation of NF- B using a NF- B dual luciferase reporter assay in BV-2 microglial cells.
j
j
With the absence of any substituent on the benzene ring (2) the
activity decreased by greater than 2-fold. It was also determined
that the malonic acid derivatives (17–20) were less active than
the corresponding ester analogs (R2 = methoxy group). Neverthe-
less, these di-acid derivatives were still effective TLR4 signaling
inhibitors with IC50 values in the low micromolar range. Taken to-
gether, these results suggest that arylidene malontate derivatives
present a consistent SAR and small modifications of its core signif-
icantly affect its inhibitory potency, implying a near optimal recog-
nition of its potential target.
2.3. Effect of 1 on downstream NF-
jB activation
Induction of the TLR4 signaling pathway stimulates the activa-
tion of NF-
ways. NF-
j
B through both the MyD88- and TRIF-dependent path-
j
B activation upregulates iNOS, resulting in elevated
production of NO. To determine whether the inhibition of NO
production by 1 is due to the suppression of NF- B activation, an
NF- B luciferase reporter gene assay was performed. NF- B dual-
j
j
j
luciferase reporter in BV-2 cells, a widely used microglial cell line
expressing various TLR receptors,30 were incubated with LPS

6076
S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
(200 ng/mL) in the absence or presence of 1 or the negative control,
compound 22, for 24 h. NF- B reporter activity was increased by
j
45-fold after LPS treatment and this increase was diminished in
cells treated with 1 in a dose-dependent manner (Fig. 1B). The
IC50 value obtained for 1 was 5.92 0.14
tive compound, 22, did not alter the LPS-induced NF-
at concentrations up to 10 M. This result suggests that 1 inhibits
the TLR4 signaling pathway upstream of NF- B activation, which is
l
M. By contrast, the inac-
jB activation
l
j
in a good agreement with its inhibitory activity of the NO
production.
2.4. Effect of 1 on downstream cytokines production
LPS-induced TLR4 activation results in an increased production
of the proinflammatory cytokines, IL-1b and TNF-a. To further
understand the mechanism of TLR4 signaling inhibition by 1, we
examined its effect on LPS-induced production of two cytokines,
Figure 3. Inhibitory effects of 1 on the NO production induced by various TLR-
specific ligands in RAW 264.7 cells. LPS (lipopolysaccharide), R848 {4-amino-2-
(ethoxymethyl)-R,R-dimethyl-1H-imidazo[4,5-c]quinoline-1-ethanol}, Pam3CSK4
{N-palmitoyl-S-[2,3-bis(palmitoyloxy)-(2R,S)-propyl]-[R]-cysteinyl-[S]-seryl-[S]-
lysyl-[S]-lysyl-[S]-lysyl-[S]-lysineÁ3HCl}, poly(I:C) (polyriboinosinic:polyribocytidy-
lic acid), and FSL-1[(S,R)-(2,3-bispalmitoyloxypropyl)-Cys-Gly-Asp-Pro-Lys-His-
Pro-Lys-Ser-Phe], were used to selectively activate TLR4, TLR7/8, TLR2/1, TLR3,
IL-1b and TNF-
SA assay.31 LPS treatment resulted in the production of significant
elevation of the IL-1b and TNF- levels compared to vehicle treated
a, in macrophages using a previously developed ELI-
a
and TLR2/6, respectively, in the presence or absence of 0.6 lM 1. Data present the
cells, reaching a maximum of approximately 20-fold and 10-fold
after 24 h, respectively. LPS-induced IL-1b production in macro-
phages was potently inhibited by 1 in a dose-dependent manner,
mean values ( SD) of three separate experiments, each performed in triplication
(significance vs LPS alone treated group, ⁄⁄P < 0.01; significance vs FSL-1 alone
treated group, ⁄P < 0.01).
with an IC50 value of 0.42 0.15
duced TNF- production was decreased in the presence of 1, with
a measured IC50 value of 1.54 0.17 M (Fig. 2B). By contrast, the
inactive compound 22 did not affect the production of either IL-
1b or TNF- . Thus, 1 can efficiently block the LPS-induced produc-
tion of several different cytokines in macrophages, which is in good
lM (Fig. 2A). Similarly, LPS-in-
a
l
lated that a potential binding mode for 1 is that it might disrupt
the TLR4 signaling by interacting with MD-2, complex as TLR4
and TLR2 are the only TLRs that have been reported to require an
accessory protein, MD-2, to initiate their signaling.33 It would be
conceivable that an inhibitor that targets the MD-2 interface with
TLR4 or TLR2/6 might selectively block TLR2 and TLR4 over other
TLRs. To further explore this hypothesis, a computational docking
search was carried out to determine if there is a desired binding
mode of 1 to the TLR4/MD-2 protein interface (TLR2/MD-2 struc-
ture remains unsolved). As shown in Figure 4, in the most energet-
ically favorable predicted binding mode, 1 was found to fit into the
LPS-binding site of TLR4-MD-2 complex exhibiting close contacts
with Gln436 of TLR4, as well as Lys122 and Ser120 in the Phe126
loop of MD-2. The entire structure of 1 was buried inside the
LPS-binding pocket where the carbonyl group could form a hydro-
gen bond with the Gln436 residue on the TLR4 surface. Interest-
ingly, the hydrophilic residues in the Phe126 loop of MD-2 and
the Gln436 residue on TLR4 are known to be important for the
interaction between LPS and TLR4-MD-2. These docking results im-
plied a possible binding mode of 1 as a disruptor of the TLR4-MD-2
protein–protein interactions.
a
agreement with our observations of its activities in NO and NF-
inhibition.
jB
2.5. Selectivity and specificity of 1
As previously discussed, a potential pitfall for 1 is its specificity
due to its ability to serve as a Michael acceptor. In order to deter-
mine if 1 selectively inhibits the TLR4 signaling, the effects of 1 on
other murine analogous TLRs were investigated using a previously
reported method with RAW 264.7 macrophage cells that can be
activated by different TLR-specific ligands.32 At a concentration of
0.6 lM, 1 showed negligible inhibition to TLR1/TLR2, TLR3, or
TLR7/8, suggesting that 1 is TLR4-specific (Fig. 3). Interestingly, it
reduced the signal via the TLR2/TLR6 heterodimer for reasons yet
to be identified (see further discussion vide infra).
2.6. Molecular docking of 1 to the TLR4-MD-2 interface
In summary, we have identified, synthesized and evaluated a
series of arylidene malonate analogs as TLR4 signaling inhibitors.
SAR studies have determined the important structural requirements
Based on the evidence that 1 inhibits the TLR4 signaling path-
way but also affects the TLR2/TLR6 signaling pathway, we specu-
A
B
3000
2500
2000
1500
1000
500
0
(μM)
0.10.81.5 3
5
10 3 5 10
1
22
Figure 2. (A) Dose-dependence effects of 1 and 22 on LPS-induced IL-1b production in the RAW 264.7 cells. (B) Dose-dependence effects of 1 and 22 on the LPS-induced TNF-
production in the RAW 264.7 cells.
a

S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
6077
Figure 4. Molecular docking of 1 to the TLR4–MD-2 complex. (A) Docking simulation of 1 to the crystal structure of the human TLR4-MD-2 complex was performed using
Glide 5.6. Molecular modeling of 1 in the LPS-binding site of the TLR4–MD-2 complex is represented as 1 by the magenta sphere, TLR4 is shown in green ribbon and MD-2 in
light orange ribbon. (B) A close-up view of the predicted interaction between 1 and the LPS-binding site of the TLR4–MD-2 complex. TLR4 is shown in green and MD-2 in light
orange.
for the high potency observed with the lead compound, 1.
Furthermore, 1 was found to inhibit LPS-mediated NF- B activa-
tion and the cytokine production of IL-1b and TNF- . A possible
Dimethyl 2-(2-fluorobenzylidene) malonate 3: yield: 89%. 1H
NMR (300 MHz, CDCl3) d 7.96 (s, 1H), 7.45–7.37 (m, 2H), 7.19–
7.09 (m, 2H), 3.88 (s, 3H), 3.84 (s, 3H).
j
a
mechanism of 1 targeting the TLR4-MD-2 interface was proposed.
Dimethyl 2-(2-chlorobenzylidene) malonate 4: yield: 73%. 1H
NMR (300 MHz, CDCl3) d 8.09 (s, 1H), 7.47–7.34 (m, 3H), 7.33–
7.27 (m, 1H), 3.89 (s, 3H), 3.77 (s, 3H).
3. Materials and methods
3.1. Chemistry
Dimethyl 2-(2-methoxybenzylidene) malonate 5: yield: 66%. 1H
NMR (300 MHz, CDCl3) d 8.13 (s, 1H), 7.42–7.33 (m, 2H), 6.97–6.90
(m, 2H), 3.87 (s, 3H), 3.86 (s, 3H), 3.80 (s, 3H).
Dimethyl 2-(4-nitrobenzylidene) malonate 6: yield: 79%. 1H
NMR (300 MHz, CDCl3) d 8.29–8.24 (m, 2H), 7.82 (s, 1H), 7.63–
7.58 (m, 2H), 3.90 (s, 3H), 3.86 (s, 3H).
Chemicals were purchased from Sigma Aldrich Chemical Co.
TLC was performed on glass plates precoated by silica gel with
visualization by UV-light. 1H NMR and 13C NMR spectra were re-
corded on a Bruker 300 or 400 MHz instrument and were refer-
enced internally to the residual solvent peak. A compound library
consisting of 1364 compounds was obtained from the National
Cancer Institute Development Therapeutics Program repository
(http://www.dtp.nci.nih.gov/index.html).
Compound 22 and 23 were prepared by following the literature
procedures (Schemes 2 and 3).34,35 The preparation of alkylidene
and arylidene malonates (1–16, 21) is through Knoevenagel Con-
densation of malonates and carbonyl compounds. The synthetic se-
quence is outlined in Scheme 1. A mixture of aldehyde (3.5 mmol),
Dimethyl 2-(4-fluorobenzylidene) malonate 7: yield: 74%. 1H
NMR (300 MHz, CDCl3) d 7.74 (s, 1H), 7.50–7.38 (m, 2H), 7.09 (m,
2H), 3.86 (s, 3H), 3.86 (s, 3H).
Dimethyl 2-(4-chlorobenzylidene) malonate 8: yield: 83%. 1H
NMR (300 MHz, CDCl3) d 7.73 (s, 1H), 7.37 (s, 4H), 3.86 (s, 3H),
3.86 (s, 3H).
Dimethyl 2-(4-(methoxycarbonyl)benzylidene) malonate 9:
yield:71%. 1H NMR (300 MHz, CDCl3) d 8.04 (d, J = 8.3 Hz, 2H),
7.79 (s, 1H), 7.48 (d, J = 8.3 Hz, 2H), 3.92 (s, 3H), 3.86 (s, 3H), 3.83
(s, 3H).
Dimethyl 2-(4-methoxybenzylidene) malonate 10: yield: 66%.
1H NMR (300 MHz, CDCl3) d 7.72 (s, 1H), 7.39 (d, J = 6.8 Hz, 2H),
6.90 (d, J = 6.8 Hz, 2H), 3.87 (s, 3H), 3.84 (s, 6H).
dimethyl malonate (3.5 mmol), acetic acid (10
lL) and piperdine
(20 L) in benzene (1.5 mL) was heated under reflux with azeotro-
l
pic removal of water overnight. Benzene was removed by rotava-
por, the residue was diluted with ethyl acetate (20 mL), washed
with 10 percent hydrochloric acid, saturated sodium bicarbonate
and brine. The organic layer was dried over anhydrous sodium sul-
fate, concentrated in vacuo and the residue was purified by silica
chromatography to provide pure product.
Dimethyl 2-(4-(dimethylamino)benzylidene)malonate
11:
yield: 63%. 1H NMR (300 MHz, CDCl3) d 7.69 (s, 1H), 7.35 (dd,
J = 9.1, 0.4 Hz, 2H), 6.65 (d, J = 9.0 Hz, 2H), 3.90 (s, 3H), 3.83 (s,
3H), 3.05 (s, 6H).
Dimethyl 2-(2.4-difluorobenzylidene) malonate 12: yield: 85%.
1H NMR (300 MHz, CDCl3) d 7.86 (s, 1H), 7.52–7.33 (m, 1H),
6.96–6.78 (m, 2H), 3.86 (s, 3H), 3.83 (s, 3H). 13C NMR (101 MHz,
CDCl3) d166.43 (s), 164.28 (dd, JCF = 253.0, 12.0 Hz), 164.09 (s),
161.39 (dd, JCF = 256.2, 12.2 Hz), 134.44 (dd, JCF = 4.6, 1.2 Hz),
130.48 (dd, JCF = 10.1, 3.6 Hz), 127.13 (s), 117.57 (dd, JCF = 12.5,
4.0 Hz), 112.05 (dd, JCF = 21.7, 3.7 Hz), 104.54 (t, JCF = 25.6 Hz),
52.78 (s), 52.68 (s). MS (ESI+) m/z: 279.0 (M+Na), 257.1 (M+H+).
Dimethyl 2-(2.4-dimethoxybenzylidene) malonate 13: yield:
75%. 1H NMR (300 MHz, CDCl3) d 8.09 (s, 1H), 7.32 (d, J = 8.5 Hz,
1H), 6.50–6.44 (m, 2H), 3.86 (s, 3H), 3.85 (s, 3H), 3.84 (s, 3H),
3.83 (s, 3H).
As shown in Scheme 1, the diester (2 mmol) was dissolved in
methanol (1.5 mL) and treated with a solution of KOH (0.23 g,
4 mmol) in water (1 mL), and the mixture was diluted with a
minimum volume of THF to provide for its homogeneity. The
reaction was heated under reflux for 16 h and concentrated in
vacuo. This residue was dissolved in a minimum of water, and
extracted with Et2O. The aqueous layer was acidified with
17.5% hydrochloric acid to pH 3–4 and extracted with Et2O.
The ethereal extract was washed with water, dried over anhy-
drous sodium sulfate and evaporated. Products were purified
by recrystallization or by precipitating them with hexane from
the ethereal solution.
Dimethyl 2-(3-nitrobenzylidene) malonate 14: yield: 69%. 1H
NMR (400 MHz, CDCl3) d 8.38–8.31 (m, 1H), 8.31–8.25 (m, 1H),
7.82 (s, 1H), 7.78–7.74 (m, 1H), 7.65–7.59 (m, 1H), 3.92 (s, 3H),
3.91 (s, 3H).
Dimethyl 2-(2-nitrobenzylidene)malonate 1: yield: 75%. 1H
NMR (300 MHz, CDCl3) d 8.25–8.22 (m, 2H), 7.76–7.53 (m, 2H),
7.44–7.41 (m, 1H), 3.90 (s, 3H), 3.62 (s, 3H).
Dimethyl 2-(3-fluorobenzylidene) malonate 15: yield: 73%. 1H
NMR (400 MHz, CDCl3) d 7.73 (s, 1H), 7.41–7.34 (m, 1H), 7.25–
7.19 (m, 1H), 7.17–7.07 (m, 2H), 3.87 (s, 6H).
Dimethyl 2-benzylidenemalonate 2: yield: 86%. 1H NMR
(300 MHz, CDCl3) d 8.09 (s, 1H), 7.48–7.29 (m, 4H), 7.28–7.23 (m,
1H), 3.89 (s, 3H), 3.77 (s, 3H).

6078
S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
Dimethyl 2-(3-methoxybenzylidene) malonate 16: yield: 85%.
3.5. Screening for NO production inhibitors
1H NMR (400 MHz, CDCl3) d 7.77 (s, 1H), 7.36–7.29 (m, 1H),
7.07–7.01 (m, 1H), 7.01–6.92 (m, 2H), 3.87 (s, 6H), 3.83 (s, 3H).
2-(2-nitrobenzylidene)malonic acid 17: yield: 52%. 1H NMR
(400 MHz, DMSO) d 8.24–8.15 (m, 1H), 7.93 (s, 1H), 7.84–7.79
(m, 1H), 7.72–7.67 (m, 1H), 7.54–7.51 (m, 1H).
For screening of the 1364-compound NCI Diversity Set II li-
brary, the murine macrophage RAW 264.7 based nitric oxide
(NO) assay was used. The library compounds (final concentration
were 10
compounds that reduced LPS-induced nitric oxide (NO) produc-
tion by 90% or more at a concentration of 10 M. The inhibition
lM) were added in duplicate. Hits were qualified as
2-benzylidene malonic acid 18: yield: 55%. 1H NMR (400 MHz,
DMSO) d 7.63–7.55 (m, 2H), 7.54 (s, 1H), 7.48–7.42 (m, 3H).
2-(2-fluorobenzylidene)malonic acid 19: yield: 46%. 1H NMR
(400 MHz, DMSO) d 7.62 (s, 1H), 7.61–7.56 (m, 1H), 7.54–7.48
(m, 1H), 7.35–7.26 (m, 2H).
l
rate (%) of NO release was determined using the following for-
mula: inhibition (%) = (OD450 value of LPS and vehicle treated
group–OD450 value of compound treated group)/(OD450 value
of LPS and vehicle treated group-OD450 value of vehicle treated
group) Â 100. For validation, positive and negative control wells
were also included that consisted of LPS-activated cells without
inhibitor and LPS-activated cells with TAK242.
To ascertain relative potencies of the most effective hits, we fur-
ther assessed the activities of the compounds that showed more
than 90% inhibition of the NO production at the initial concentra-
2-(4-methoxybenzylidene) malonic acid 20: yield: 54%. 1H NMR
(400 MHz, DMSO) d 7.78 (s, 1H), 7.50–7.46 (m, 1H), 7.46–7.41 (m,
1H), 7.14–7.07 (m, 1H), 7.01–6.96 (m, 1H).
Dimethyl 2-(cyclohexylmethylene) malonate 21: yield: 88%. 1H
NMR (300 MHz, CDCl3) d 6.86 (d, J = 10.5 Hz, 1H), 3.84 (s, 3H), 3.78
(s, 3H), 2.39 (dd, J = 10.8, 3.3 Hz, 1H), 1.79–1.63 (m, 5H), 1.37–1.11
(m, 5H).
Dimethyl 2-(2-nitrobenzyl) malonate 22: yield: 62%. 1H NMR
(300 MHz, CDCl3) d 8.04-8.01 (m, 1H), 7.58–7.53 (m, 1H), 7.47–
7.37 (m, 2H), 3.94 (t, J = 7.6 Hz, 1H), 3.72 (s, 6H), 3.53 (d,
J = 7.6 Hz, 2H).
tion of 10
lM. Secondary screening was performed using the Nitric
Oxide assay at a compound concentration of 1
lM in triplicate.
Compounds exhibiting the inhibitory by 80% or more were sub-
jected to a toxicity analysis.
(E)-Methyl 3-(2-nitrophenyl) acrylate 23: 1H NMR (300 MHz,
CDCl3) d 8.14 (d, J = 15.8 Hz, 1H), 8.09–8.04 (m, 1H), 7.69–7.64
(m, 2H), 7.60–7.54 (m, 1H), 6.39 (d, J = 15.8 Hz, 1H), 3.85 (s, 3H).
3.6. TLR specificity test
This assay was performed using the same protocol with ‘‘3.4
Measurement of NO’’ as previously described. Instead of LPS, polyr-
iboinosinic:polyribocytidylic acid (Poly(I:C)), FSL-1 ((S,R)-(2,3-bis-
palmitoyloxypropyl)-Cys-Gly-Asp-Pro-Lys-His-Pro-Lys-Ser-Phe),
3.2. Cell culture and inhibitor treatment
Each compound is dissolved at a concentration of 10 mM in
DMSO. Murine macrophage RAW 264.7 (American Type Culture
Collection, Rockville, MD) were routinely cultured at 37 °C in a
humidified 5% CO2 atmosphere in RPMI 1640 supplemented with
10% fetal bovine serum, penicillin (100 units/mL), and streptomy-
R848 (4-amino-2-(ethoxymethyl)-
c]quinoline-1-ethanol) and Pam3CSK4
a
,
a
-dimethyl-1H-imidazo[4,5-
(N-palmitoyl-S-[2,3-
bis(palmitoyloxy)-(2RS)-propyl]-[R]-cysteinyl-[S]-seryl-[S]-lysyl-
[S]-lysyl-[S]-lysyl-[S]-lysineÁ3HCl) were used to selectively activate
TLR3, TLR2/6, TLR7/8 and TLR1/2, respectively.
cin sulfate (100 lg/mL). Cells placed in a 96-well plate at a density
of 7.5 Â 104 cells/well were incubated for 24 h. Cultured cells were
treated with vehicle (control) and various concentrations of com-
pound and then stimulated with 20 ng/mL of LPS for 24 h.
3.7. Dual luciferase report assay
NF-
jB dual luciferase reporter BV-2 cells were cultured in
3.3. Cell viability assay
DMEM medium supplemented with 10% FBS, penicillin (100 unit/
mL), streptomycin (100 lg/mL) and puromycin (3 lg/mL). BV-2
Cell viability was determined by (4-[3-(4-iodophenyl)-2-(4-
nitrophenyl)-2H-5-tetrazolio]-1, 3-benzene disulfonate, WST-1)
assay using Clontech premixed WST-1 cell proliferation reagent
according to the manufacturer’s instructions. Briefly, cells were
inoculated at a density of 2 Â 104 cells/well into 96-well plate
and cultured at 37 °C for 24 h. The culture medium was replaced
reporter cells were seeded at a density of 1 Â 104 cells/well in
96-well plates. After 24 h incubation, medium was changed to
Opti-MEM medium supplemented with 0.5% FBS and indicated
concentration of compound was added, and then stimulated with
200 ng/mL of LPS. After further 24 h treatment, the NF-jB activity
was analyzed by Dual-Glo Luciferase Assay System. The ratio of
Firefly luciferase activity to Renilla luciferase activity represents
the NF-jB activity.
with 100
with vehicle (control) and various concentrations of compound.
After 24 h, 10 L premixed WST-1 solution was added to each well.
lL serum free medium and cultured cells were treated
l
After incubation at 37 °C for 30 min, the absorbance at 490 nm was
measured using a microplate reader.
3.8. Measurement of cytokines
RAW 264.7 cells (5 Â 105/well) pretreated with or without
LPS, followed by treatment with indicated compound in 6-well
plates. After 24 h, supernatants were harvested, clarified by cen-
trifugation, and stored at À80 °C prior to analysis. Cells were col-
lected and lysed by mammalian protein extraction reagent
(Thermo Scientific, Rockford, IL, USA). Cell lysates were centri-
fuged at 1.3k rpm for 30 min at 4 °C, the supernatant were col-
lected and stored at À80 °C prior to analysis. RAW 264.7 cell
3.4. Measurement of NO
RAW 264.7 cells were placed in a 96-well plate at a density of
7.5 Â 104 cells/well and incubated for 24 h. On the treatment
day, media was removed and replaced with RPMI 1640 medium.
Cultured cells were treated with vehicle or various concentration
of compound then stimulated with 20 ng/mL LPS for 24 h. The ni-
trite concentration in the cultured media was measured as an indi-
lysate IL-1b levels and RAW 264.7 cell media TNF-
a levels were
cator of NO secretion. Culture media (100
10 L of 2,3-diaminonaphthalene (0.05 mg/mL in 0.62 M aqueous
HCl solution). After 15 min incubation in the dark, 5 L of a 3 M
lL) were mixed with
determined by enzyme-linked immunosorbent assay (ELISA) (BD
Bioscience, San Diego, CA, USA) following the manufacturer’s
protocol. The total amount of the IL-1b was normalized to the
total protein concentration.
l
l
aqueous NaOH solution was added to each well. Then, absorbance
of the mixture at 450 nm was measured with a microplate reader.

S. Zhang et al. / Bioorg. Med. Chem. 20 (2012) 6073–6079
11. Tian, B.; Brasier, A. R. Recent Prog. Horm. Res. 2003, 58, 95.
6079
3.9. Molecular modeling
12. Hennessey, E. J.; Parker, A. E.; O’Neill, L. A. Nat. Rev. Drug Disc. 2010, 9, 293.
13. Guo, L. H.; Schluesener, H. J. Cell. Mol. Life Sci. 2007, 64, 1128.
14. Friedman, W. J. Exp. Neurol. 2001, 168, 23.
15. Sian-Marie Lucas, S.; Rothwell, N. J.; Gibson, R. M. Br. J. Pharmacol. 2006, 147,
S232.
Docking simulation of 1 was carried out using the Glide 5.6 pro-
gram,31,36 and the crystal structure of human TLR4–MD-2–LPS was
cited from Protein Data Bank (3FXI).5 The TLR4–MD-2–LPS com-
plex orientations and conformations were prepared using standard
Glide protocol and compound 1 was created with multiple confor-
mational states and all possible rotational orientations.37 1 was fi-
nal docked using standard Glide.
16. Carcillo, J. A. Crit. Care Clin. 2003, 19, 413.
17. Serhan, C. N.; Savill, J. Nat. Immunol. 2005, 6, 1191.
18. Rocha, V. Z.; Libby, P. Nat. Rev. Cardiol. 2009, 6, 399.
19. Rakoff-Nahoum, S. Yale J. Biol. Med. 2006, 79, 123.
20. Buchanan, M. M.; Hutchinson, M. R.; Watkins, L. R.; Yin, H. J. Neurochem. 2010,
114, 13.
21. Ii, M.; Matsunaga, N.; Hazeki, K.; Nakamura, K.; Takashima, K.; Seya, T.; Hazeki,
O.; Kitazaki, T.; Iizawa, Y. Mol. Pharmacol. 2006, 69, 1288.
22. Mullarkey, M.; Rose, J. R.; Bristol, J.; Kawata, T.; Kimura, A.; Kobayashi, S.;
Przetak, M.; Chow, J.; Gusovsky, F.; Christ, W. J.; Rossignol, D. P. J. Pharmacol.
Exp. Ther. 2003, 304, 1093.
23. Rice, T. W.; Wheeler, A. P.; Bernard, G. R.; Vincent, J. L.; Angus, D. C.; Aikawa, N.;
Demeyer, I.; Sainati, S.; Amlot, N.; Cao, C.; Ii, M.; Matsuda, H.; Mouri, K.; Cohen,
J. Crit. Care Med. 2010, 38, 1685.
24. Bloomberg web site. http://www.bloomberg.com/news/2011-01-25/eisai-s-
sepsis-drug-eritoran-fails-to-save-more-lives-in-final-stage-study.html.
Posted 25.01.11.
25. Nussler, A. K.; Glanemann, M.; Schirmeier, A.; Liu, L.; Nüssler, N. C. Nat. Protoc.
2006, 1, 2223.
Acknowledgments
We thank the National Institutes of Health (DA026950,
DA025740, NS067425, and DA029119) for financial support of this
work. S.Z. thanks the support of China Scholarship Council (Grant
No. 2009604039).
Supplementary data
26. Ngamwongsatit, P.; Banada, P. P.; Panbangred, W.; Bhunia, A. K. J. Microbiol.
Methods 2008, 73, 211.
27. Mullarkey, M.; Rose, J. R.; Bristol, J.; Kawata, T.; Kimura, A.; Kobayashi, S.;
Przetak, M.; Chow, J.; Gusovsky, F.; Christ, W. J.; Rossignol, D. P. J. Pharmacol.
Exp. Ther. 2003, 304, 1093.
28. Takashima, K.; Matsunaga, N.; Yoshimatsu, M.; Hazeki, K.; Kaisho, T.; Uekata,
M.; Hazeki, O.; Akira, S.; Iizawa, Y.; Ii, M. Br. J. Pharmacol. 2009, 157, 1250.
29. Guterman, L. Chem. Eng. News 2011, 89, 19.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2012.08.022.
References and notes
1. Akira, S.; Uematsu, S.; Takeuchi, O. Cell 2006, 124, 783.
2. Akira, S. Curr. Top. Microbiol. Immunol. 2006, 311, 1.
3. Dauphinee, S. M.; Karsan, A. Lab Invest. 2006, 86, 9.
4. Raetz, C. R.; Whitfield, C. Annu. Rev. Biochem. 2002, 71, 635.
5. Park, B. S.; Song, D. H.; Kim, H. M.; Choi, S.; Lee, H.; Lee, J. Nature 2008, 458,
1191.
6. Saitoh, S.; Akashi, S.; Yamada, T.; Tanimura, N.; Kobayashi, M.; Konno, K.;
Matsumoto, F.; Fukase, K.; Kusumoto, S.; Nagai, Y.; Kusumoto, Y.; Kosugi, A.;
Miyake, K. Int. Immunol. 2004, 16, 961.
7. Collins, T.; Read, M. A.; Neish, A. S.; Whitley, M. Z.; Thanos, D.; Maniatis, T.
FASEB 1995, 9, 899.
8. Zhang, G.; Ghosh, S. J. Clin. Invest. 2001, 107, 13.
9. Karin, M.; Ben-Neriah, Y. Annu. Rev. Immunol. 2000, 18, 621.
10. Hawiger, J. Immunol. Res. 2001, 23, 99.
30. Liu, S.; Kielian, T. J. Immunol. 2009, 183, 5537.
31. Cheng, K.; Wang, X.; Yin, H. J. Am. Chem. Soc. 2011, 133, 3764.
32. Bevan, D. E.; Martinko, A. J.; Loram, L. C.; Stahl, J. A.; Taylor, F. R.; Joshee, S.;
Watkins, L. R.; Yin, H. ACS Med. Chem. Lett. 2010, 1, 194.
33. Dziarski, R.; Gupta, D. J. Endotoxin Res. 2000, 6, 401.
34. Adediran, S. A.; Cabare, D.; Lohier, J.-F.; Wakselman, M.; Pratt, R. F. Bioorg. Med.
Chem. 2010, 18, 282.
35. Chang, S.; Yin, S.-L.; Wang, J.; Jing, Y.-K.; Dong, J.-H. Molecules 2009, 14, 4166.
36. Glide, version 5.6; Schrödinger, LLC: New York, NY, 2010.
37. Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W.
T.; Banks, J. L. J. Med. Chem. 2004, 47, 1750.