SAR studies on truxillic acid mono esters as a new class of antinociceptive agents targeting fatty acid binding proteins

Article abstract of DOI:10.1016/j.ejmech.2018.04.050

Fatty acid binding proteins (FABPs) serve as critical modulators of endocannabinoid signaling by facilitating the intracellular transport of anandamide and whose inhibition potentiates anandamide signaling. Our previous work has identified a novel small-m

Full text of DOI:10.1016/j.ejmech.2018.04.050

European Journal of Medicinal Chemistry 154 (2018) 233e252
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
SAR studies on truxillic acid mono esters as a new class of
antinociceptive agents targeting fatty acid binding proteins
,
,
Su Yan a 1, Matthew W. Elmes b 1, Simon Tong a, Kongzhen Hu e, Monaf Awwa a,
Gary Y.H. Teng e, Yunrong Jing e, Matthew Freitag a, Qianwen Gan a, Timothy Clement a,
Longfei Wei a, Joseph M. Sweeney b, Olivia M. Joseph b, Joyce Che b,
Gregory S. Carbonetti b, Liqun Wang b, Diane M. Bogdan c, Jerome Falcone c,
Norbert Smietalo c, Yuchen Zhou d, Brian Ralph d, Hao-Chi Hsu f, Huilin Li f,
Robert C. Rizzo d e, Dale G. Deutsch b e, Martin Kaczocha b e, Iwao Ojima a
,
,
,
c
,
, e, *
a Department of Chemistry, Stony Brook University, Stony Brook, NY, 11794-3400, United states
b Departments of Biochemistry and Cell Biology, Stony Brook University, Stony Brook, NY, 11794-5215, United states
c Department of Anesthesiology, Stony Brook University, Stony Brook, NY, 11794-8480, United states
d Applied Mathematics and Statistics, Stony Brook University, Stony Brook, NY, 11794-3600, United states
e Institute of Chemical Biology and Drug Discovery, Stony Brook University, Stony Brook, NY, 11794-3400, United states
f Cryo-EM Structural Biology Laboratory, Van Andel Research Institute, Grand Rapids, MI, 49503, United states
a r t i c l e i n f o
a b s t r a c t
Article history:
Fatty acid binding proteins (FABPs) serve as critical modulators of endocannabinoid signaling by facili-
tating the intracellular transport of anandamide and whose inhibition potentiates anandamide signaling.
Received 3 February 2018
Received in revised form
20 April 2018
Our previous work has identified a novel small-molecule FABP inhibitor,
a-truxillic acid 1-naphthyl
monoester (SB-FI-26, 3) that has shown efficacy as an antinociceptive and anti-inflammatory agent in
rodent models. In the present work, we have performed an extensive SAR study on a series of 3-analogs
as novel FABP inhibitors based on computer-aided inhibitor drug design and docking analysis, chemical
synthesis and biological evaluations. The prediction of binding affinity of these analogs to target FABP3, 5
and 7 isoforms was performed using the AutoDock 4.2 program, using the recently determined co-crystal
structures of 3 with FABP5 and FABP7. The compounds with high docking scores were synthesized and
evaluated for their activities using a fluorescence displacement assay against FABP3, 5 and 7. During lead
Accepted 23 April 2018
Keywords:
Fatty acid binding protein
FABP inhibitor
Anti-nociceptive agent
Anti-inflammatory agent
SAR study
Computer-aided design
Molecular docking
optimization, compound 3l emerged as a promising compound with the Ki value of 0.21
m
M for FABP 5,
M). Nine compounds exhibit similar or better binding affinity than 3,
M) and 4e (Ki, 0.68 M). Twelve compounds are selective for FABP5
4-fold more potent than 3 (Ki, 0.81
m
including compounds 4b (Ki, 0.55
m
m
and 7 with >10 mM Ki values for FABP3, indicating a safe profile to avoid potential cardiotoxicity concerns.
Compounds 4f, 4j and 4k showed excellent selectivity for FABP5 and would serve as other new lead
compounds. Compound 3a possessed high affinity and high selectivity for FABP7. Compounds with
moderate to high affinity for FABP5 displayed antinociceptive effects in mice while compounds with low
FABP5 affinity lacked in vivo efficacy. In vivo pain model studies in mice revealed that exceeding hy-
drophobicity significantly affects the efficacy. Thus, among the compounds with high affinity to FABP5
in vitro, the compounds with moderate hydrophobicity were identified as promising new lead com-
pounds for the next round of optimization, including compounds 4b and 4j. For select cases, compu-
tational analysis of the observed SAR, especially the selectivity of new inhibitors to particular FABP
isoforms, by comparing docking poses, interaction map, and docking energy scores has provided useful
insights.
© 2018 Elsevier Masson SAS. All rights reserved.
* Corresponding author. Department of Chemistry, Stony Brook University, Stony
Brook, NY, 11794, United states.
1. Introduction
E-mail address: iwao.ojima@stonybrook.edu (I. Ojima).
These two authors contributed equally.
1
The fatty acid binding proteins (FABPs) are a family of small
https://doi.org/10.1016/j.ejmech.2018.04.050
0223-5234/© 2018 Elsevier Masson SAS. All rights reserved.

234
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
chaperone proteins that act as cytosolic transporters for a wide
variety of lipophilic substances including fatty acids, N-acyletha-
nolamines (NAE), eicosanoids, and cannabinoids [1,2]. FABPs are
widely expressed throughout the body and play an integral role in a
multitude of physiological processes such as lipid metabolism,
inflammation and neuronal signaling [3]. FABPs of the mammalian
central and peripheral nervous systems have been shown to facil-
itate the intracellular transport of NAEs, particularly the endo-
cannabinoid arachidonoyl ethanolamide (anandamide, AEA), as
well as catabolism by the endoplasmic reticulum-localized enzyme
fatty acid amide hydrolase (FAAH) [4]. Genetic or pharmacological
inhibition of the FAAH enzyme or the FABPs results in a marked
elevation of brain AEA levels, which acts upon type-1 cannabinoid
receptors (CB1R) and thereby cause a suppression of pain trans-
mission and other therapeutically beneficial effects [5e7]. As such,
designing inhibitors of AEA inactivation is desirable. Over the years
numerous FAAH inhibitors have been developed and have generally
been well tolerated in the clinical setting [8,9]. However, focused
medicinal chemistry efforts on targeting FABPs may hold advan-
tages over direct FAAH inhibition because unlike FAAH, which is
distributed throughout the body, humans express multiple FABP
isoforms that exhibit tissue-specific expression patterns. Designing
small molecule inhibitors that selectively bind target FABP isoforms
will allow for drugs to preferentially act upon target tissues of in-
terest rather than systemic NAE upregulation, which may increase
the likelihood of off-target adverse events. This has led to the
pursuit of identifying novel compounds that are capable of selec-
tively inhibiting the FABP isoforms that are expressed in the
mammalian central and peripheral nervous systems. Three FABP
isoforms have been identified in these tissues, i.e., FABP3 (heart
FABP, H-FABP), FABP5 (epidermal FABP, keratinocyte FABP, E-FABP),
and FABP7 (brain FABP, B-FABP) [10].
antinociceptive activity studies on (ꢀ)-incarvillateine have found
that the cyclobutane moiety is required for its analgesic properties
[15]. To our knowledge, the mechanism of action for (ꢀ)-incarvil-
lateine-induced analgesia has not been formally elucidated in full,
though adenosine receptors likely play a role and there is some
contention in the literature pertaining to the involvement of the
opioid system [16,17]. Considering the structural similarity and
essentially overlapping reported pharmacological profiles of
(ꢀ)-incarvillateine and 3, it is tempting to speculate a possibility
that its effects are mediated, at least in part, by FABP inhibition and
subsequent NAE/endocannabinoid potentiation.
Compound 3 was further shown to be biologically active in a
FABP and CB1R-dependent manner. The compound has a half-life of
~3 h in vivo and is efficacious in producing anti-inflammatory and
anti-nociceptive effects in rodent models of visceral, thermal,
neuropathic, and inflammatory pain [6]. Furthermore, intraperito-
neal (i.p.) administration of 3 up to 40 mg/kg in mice showed no
conditioned place preference nor conditioned place aversion,
indicating a relatively low potential for addiction [18].
Despite promising efficacy in pain models, 3 requires further
preclinical optimization to improve potency, solubility, selectivity
and in vivo stability. To this end, the a-truxillic acid monoester core
structure was used as the scaffold for optimization in the present
study. Based on the prediction using the Autodock 4.2 program [19],
a series of novel 3-analogs have been designed, synthesized, and
their potencies evaluated to investigate SAR for enhanced potency
and selectivity.
2. Results and discussion
2.1. Optical resolution of 3
Previous work from our group has led to the identification of a
In the previous study, we examined the potency of racemic 3.
However, our recent protein X-ray structure determination of
FABP5/3 as well as FABP7/3 cocrystals revealed that (S,S,S,S)-3 was
incorporated into the canonical binding site. Accordingly, there is a
good possibility that the (S,S,S,S)-enantiomer may be substantially
more potent than the (R,R,R,R)-enantiomer. Thus, we set out to
optically resolve the two enantiomers of 3.
(1R,2S)-2-amino-1,2-diphenylethanol, (S)-1-(1-naphthyl)ethyl-
amine, and (S)-phenylalaninol (Fig. 2), were examined as resolving
agents. A mixture of 3 and a resolving agent was dissolved in
common lab solvents (i.e., ethanol, isopropanol, and acetonitrile)
and allowed to recrystallize as diastereomeric salts. Samples
showing crystal formation were acidified with 4 M HCl, extracted
with ethyl acetate, and analyzed by chiral HPLC using a Chiralcel
ODH column (iPrOH/hexane). The attempts using (1R,2S)-2-amino-
1,2-diphenylethanol, and (S)-1-(1-naphthyl)ethylamine resulted in
recovering only racemic 3. Fortunately, (S)-phenylalaninol was
found to be a suitable resolving agent in combination with meth-
anol. Thus, we were able to isolate two enantiomers, 3-A and 3-B.
(See Experimental for details.)
novel competitive FABP inhibitor,
a-truxillic acid 1-naphthyl
monoester (SB-FI-26, 3) [11]. It has been shown that 3 is a potent
inhibitor of FABP5 and FABP7, with sub-micromolar affinities re-
ported in vitro (Ki ¼ 0.9 0.1
m
M and 0.4 0.0
mM, respectively),
with weaker binding to FABP3 (Ki ¼ 3.9 0.7
mM) [6]. Selective in-
hibition for FABP5 over FABP3 is deemed desirable, as mice bearing
a knockout for FABP3 exhibited age-related cardiac hypertrophy,
and thus pharmacological inhibition of this protein may have a
potential to cause undesirable side effects [12].
Intriguingly, compound 3 shares the same a-truxillic acid skel-
eton with that of (ꢀ)-incarvillateine (Fig. 1), a natural monoterpene
alkaloid isolated from the Chinese herbal Incarvillei sinensis, that
has been used as a pain-reliever in traditional Eastern medicine (as
the dried plant matter ‘Jiaohao’), and more recently has been
shown to produce potent analgesic and anti-inflammatory effects
in formalin-induced mouse models [13,14]. Structure-
Since 3-A gave better crystals, it was subjected to an X-ray
crystallographic analysis and unambiguously determined to be the
(1R,2R,3R,4R)-enantiomer (Fig. 3). (See Supporting Information for
crystallographic data.) Thus, 3-B was assigned to the (1S,2S,3S,4S)-
enantiomer.
It should be noted, however, the fluorescence displacement
assay of the two enantiomers showed little difference in activity. Ki
values of (R,R,R,R)-3 were 0.71 0.08
FABP5 and FABP7, respectively, while Ki values of (S,S,S,S)-3 were
0.79 0.15 M and 0.45 0.01 M for FABP5 and FABP7, respec-
mM and 0.92 0.22 mM for
m
m
tively. The Ki values of two enantiomers for FABP5 are within an
error, although there is a recognizable difference in their values for
FABP7. Based on these data, we concluded that synthesis of
Fig. 1. Structures of (ꢀ)-incarvillateine and 3.

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
235
Fig. 2. Enantiopure amines examined for optical resolution of 3.
bridge with Arg129, one H-bond with Tyr131, and four H-bonds
with Arg109 via ordered water molecules, mimicking the natural
substrates of FABP5 (Fig. 4A zoom). Similar interactions are also
observed in the canonical-site of the FABP7/3 complex. The
carboxyl group of 3 forms a salt bridge with Arg127 and a water-
mediated network of hydrogen bonds with Arg127, Tyr129,
Arg107, and Thr54 (Fig. 4B zoom). It should be noted that FABP5 is
in an open-gate conformation as the S3eS4 and S5eS6 strands are
splayed away from the H1eH2 cap as compared to those in the
FABP7 structure.
The computational analysis of the FABP5/3 and FABP/3 com-
plexes revealed that the space around the two phenyl rings of 3
enables further modifications, with the purpose of generating
either new H-bond or hydrophobic interactions between the in-
hibitor and FABP5 or FABP7. Furthermore, the 1-naphthyl group of 3
is quite exposed in the binding pocket, which indicates that it is
worthwhile to replace the 1-naphthyl moiety by other groups for
improved potency, selectivity and structural diversity. The binding
pose was also considered as an important criterion. Thus, only
compounds with good overlap with the binding pose of 3 in FABP5
or FABP7 were selected for chemical synthesis and inhibitory ac-
tivity evaluation. In addition, the lipophilicity of the designed
compounds was also taken into account since the partition coeffi-
cient (LogP) provides important information about bioavailability,
as well as necessary drug formulation and delivery [21,22]. Thus,
the calculated LogP (cLogP) value for each designed compound was
obtained using the ChemDraw 15.0 software [23]. All docking en-
ergy scores and cLogP values for the novel 3-analogs, which were
designed and selected for chemical synthesis are summarized in
Table S7 in the Supporting Information. Fig. 5 shown the overlay of
the docked poses of representative FABP5 inhibitors, thus designed,
which were selected for synthesis, wherein all those molecules
possess critical canonical interactions with Tyr131 and Arg129
(AutoDock 4.2).
Fig. 3. X-ray crystal structure of (R,R,R,R)-3 (3-A).
enantiomerically pure analogs would not be necessary, and thus all
new analogs in the present work were synthesized as racemic
forms.
Currently we have two possible explanations why only (S,S,S,S)-
3 binds to both FABP5 and FABP7 in our co-crystal structures [20]:
(i) it is possible that the crystallization process may have selectively
incorporated (S,S,S,S)-3 complex into the crystal lattice, leaving the
R form in solution; (ii) it is also possible that (S,S,S,S)-3 may bind to
the portal site more rapidly than to the canonical site, leading to an
increased local concentration of (S,S,S,S)-3 for binding to the ca-
nonical site. Importantly, our computational analysis showed that
both (S,S,S,S)-3 and (R,R.R.R)-3 are geometrically and energetically
compatible in the binding pocket of the FABP7-3 complex [20].
Also, the docking energy scores of FABP5-(S,S,S,S)-3 and FABP5-
(R,R.R.R)-3 calculated by the AutoDock 4.2 program showed only a
very small difference (see Table S7, Supplementary Material). Thus,
the Ki values of these two enantiomers of 3 are consistent with the
computational analyses.
2.3. Chemical synthesis
The synthesis of novel 3-analogs of commenced with the
preparation of the
photodimerization of the corresponding E-cinnamic acids in solid
a
-truxillic acid components through [2 þ 2]
state [24]. A general procedure for the photochemical synthesis of
a-truxillic acids and its analogs from commercially available E-
cinnamic acids under the irradiation of UV light at 365 nm is shown
in Scheme 1.
2.2. Design of 3-analogs as FABP inhibitors
It was reported that direct [2 þ 2] photodimerization of E-ferulic
acid was very slow and the yield was very poor [25]. Thus, 4-OTs-E-
ferulic acid was synthesized first, following Kibayashi's procedure
and subjected to photodimerization at 360 nm [25]. Under the
optimized conditions, the desired product 1h was formed in 88%
yield. The tosyl group was removed by NaOH in MeOH to afford 1i
For designing novel FABP inhibitors, AutoDock 4.2 was used to
predict the binding affinity and interactions between the inhibitors
and FABP proteins. The co-crystal structures of FABP5 and FABP7
with 3 (PDB ID: 5UR9 and 5URA, respectively) [20] were used as
docking receptors. The structures of FABP5/3 complex (Fig. 4A) and
FABP7/3 complex (Fig. 4B) provide the structural basis for compu-
tational construction of new analogs.
for FABP binding assays. In a similar manner, a-truxillic acid analog
1j bearing 4-TBDMSO-phenyl moieties was synthesized from E-4-
At the canonical-site of the FABP5/3 complex, 3 forms a salt
coumaric acid in 84% yield (Scheme 2).

236
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
Fig. 4. Structures of FABP5 and FABP7 in complex with (S,S,S,S)-3. (A) (S,S,S,S)-3 bound at the canonical site of human FABP5. (B) (S,S,S,S)-3 bound at the canonical site of human
FABP7.
Fig. 5. Overlay of representative FABP5 inhibitors designed (Table S7).
Scheme 1. Synthesis of
of E-cinnamic acids.
a
-truxillic acids 1a~g via solid-state [2 þ 2] photodimerization
The resulting unsubstituted and substituted
a-truxillic acids
alcohols selected from the computational design to form the cor-
responding unsubstituted and substituted -truxillic acid mono-
esters (3 and 4) as shown in Scheme 3. In some cases, the formation
of -truxillic acid diesters was inevitable (5a~e) (Scheme 3) and
(1a~f) were reacted with thionyl chloride in the presence of a
catalytic amount of dimethylformamide (DMF) under reflux to
afford the corresponding diacid dichlorides (2a~f). The diacid
dichlorides thus formed were further reacted with a variety of
a
a

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
237
Scheme 2. Synthesis of 1h, 1i and 1j.
Scheme 3. Synthesis of a-truxillic acid monoesters 3x and 4x, as well as diesters 5x (for the R group, see Table S7).
thus these diesters were isolated and their FABP binding examined.
reacting 1h with 1-naphthol in the presence of EDC$HCl and DMAP,
followed by the removal of the tosyl group with Na/anthracene
(Scheme 4). In a similar manner, a simpler 3-analog bearing 4-
hydroxyphenyl groups, 4d, was also synthesized (Scheme 4).
Compound 3j was synthesized in a manner similar to that
described for 3a through mono-esterification of 1a, giving 3j-1,
As mentioned above, (ꢀ)-incarvillateine and 3 share the
a-
truxillic acid core structure, we mimicked the 4-hydroxy-3-
methoxy substitution pattern of (ꢀ)-incarvillateine on the phenyl
groups to examine the substituent effects on the FABP binding of
the corresponding 3-analog 4a. Compound 4a was synthesized by
Scheme 4. Synthesis of (ꢀ)-incarvillateine-mimicked 3-analogs, 4a and 4d.

238
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
followed by deprotection of the TIPS group with tetrabutylammo-
2.5. Structure-activity relationship (SAR) analysis
nium fluoride (TBAF), as shown in Scheme 5.
According to our docking study on
a
-truxic acid monoesters, the
Based on the Ki values obtained, the SAR of a series of new a-
ester moiety was found to be exposed in the FABP binding pockets.
This suggests that we may be able to attach a hydrophilic group to
the ester moiety without affecting binding affinity. Accordingly, we
designed compound 3s, bearing a short PEG chain. The docking
score of 3s for FABP5 was rather low, but favorable for FABP7 (see
Table S7). Compound 3s was prepared in high yield by reacting 3f
bearing an ethynyl group with azido-triethylene glycol under Cu-
catalyzed click reaction conditions (Scheme 6).
Compound 3l bearing a trans-2-phenylcyclohexyl monoester
gave considerably better docking scores than 3 for both FABP5 and
FABP7 (see Table S7), regardless of stereochemistry. Initially, we
synthesized 3l using racemic trans-2-phenylcyclohexanol for
mono-esterification with 1a. However, when we used optically
pure (1R,2S)-trans-2-phenylcyclohexanol, we were able to isolate
single stereoisomer 3l-A by recrystallization of a diastereomer
mixture 3l-A/C from EtOAc/hexanes (Scheme 7, equation a). In the
same manner, 31-B/D was synthesized from 1a and (1S,2R)-trans-2-
phenylcyclohexanol, and 3l-B was isolated as single isomer
(Scheme 7, equation b). Thus, 3l-A and 3l-B are enantiomers.
truxillic acid monoesters and their congeners can be analyzed. As
mentioned above, (S,S,S,S)-3 and (R,R,R,R)-3 are found to have the
same binding affinity (Ki) to all three FABPs within the standard
deviation. Some insight into this result was discussed in our recent
structural biology paper [20]. Our computational and structural
biology studies have clearly indicated that the canonical interaction
of the carboxylic acid moiety of 3 with Arg129 (FABP5) and Arg127
(FABP7) is essential for its binding to FABPs [11,20]. Accordingly, we
hypothesized that (ꢀ)-incarvillateine would be hydrolyzed to the
corresponding monoester, which is the putative active form of this
drug and binds to FABPs. Thus, a simplified mimic of this putative
monoester such as 4a should show a good affinity to FABPs. As
Table 1 shows, 4a exhibits fairly good affinity to FABP3 (Ki 1.06
mM)
and FABP7 (Ki 2.12 M), but not to FABP5 (Ki >10 M) (entry 31).
m
m
The results suggest that the 4-hydroxy-3-methoxyphenyl group in
4a somewhat compromises affinity as compared to the simple
phenyl group in 3, probably because of its bulkiness. Since the 4-
hydroxyl group could serve as a hydrogen bond donor and the
methoxy group at the 3-position might cause a steric clash, com-
pound 4d, bearing just 4-hydroxylphenyl group, is an interesting
compound to examine. However, 4d shows comparable affinity and
selectivity to 4a for the three FABPs (entry 34), which indicates that
the 4-hydroxyl group does not bring in additional favorable in-
teractions and rather compromises binding to FABPs, especially to
FABP5. These results indicate that the substituents at the meta- and
Although the absolute configuration of a-truxillic acid moiety of 3l-
A or 31-C has not been determined by X-ray crystallography, mo-
lecular mechanics calculations using the ChemBio 3D Ultra 14
program [26] indicate that (1R,2R,3R,4R) configuration is favorable
over (1S,2S,3S,4S) configuration (
D
E ¼ 1.7 kcal/mol) for 3l-A. Thus,
we have tentatively assigned (1R,2R,3R,4R,10R,20S) configuration to
3l-A and (1S,2S,3S,4S,10R,20S) configuration to 3l-C. Since 3l-B and
3l-D are enantiomers of 3l-A and 3l-C, respectively, the absolute
configurations of these compounds should be (1S,2S,3S,4S,10S,20R)
and (1R,2R,3R,4R,10S,20R), as shown in Scheme 7.
para-positions of a-truxillic acid core may cause unfavorable van
der Waals interactions with amino acid residues in the binding
sites, especially for FABP5.
Since
ported to show antinociceptive activities [14,28,29], we examined
the affinity of -truxillic acid (1a), its 3,4-dimethoxyphenyl deriv-
a-truxillic acid, its diesters and their congeners were re-
a
-Truxillic acid monoamides 6a and 6b were synthesized from
1a through the amide coupling with an aminothiazole and an
aminobiphenyl, as shown in Scheme 8. In addition to -truxillic
monoesters and monoamides, -truxillic acid mono ester 3o- and
monoamide 6a- were synthesized from -truxillic anhydride
(Scheme 9). -Truxillic anhydride (1a- ) was readily prepared by
reacting 1a with acetic anhydride in the presence of sodium acetate
[11]. Then, anhydride 1a- was reacted with a naphthol and an
aminothiazole to give the corresponding -monoester 3o- and
monoamide 6a- , respectively.
a
a
ative (1g) and 4-hydroxyphenyl derivative (1i) to the three FABPs.
However, none of these dicarboxylic acids show appreciable affinity
g
g
g
g
to FABPs (Ki > 10
by-products in the synthesis of the corresponding
monoesters, do not show appreciable affinity to FABPs (Ki > 10
(entries 43e46). These results unambiguously confirm that the a-
m
M) (entries 1e3). Also, diesters 5a~d, obtained as
-truxillic acid
M)
g
g
a
m
g
g
g
g-
truxillic acid monoester scaffold is essential for strong binding to
FABPs.
g
Among the new unsubstituted and substituted a-truxillic acid
derivatives, compound 3l shows substantially higher affinity to
FABP5 than the parent compound 3 b y a factor of 4 (entry 18). The
2.4. FABP-binding assay
Ki values of 3l for FABP5 and FABP7 are 0.21
respectively. Besides 3l, 8 compounds (3f, 3g, 3h, 3k, 3, 4b, 4e 4k)
exhibit better or similar affinity towards FABP5 (Ki 0.55e0.97 M)
mM and 0.40 mM,
All compounds thus synthesized were evaluated for their
binding affinity to human FABP3, FABP5 and FABP7 as previously
described (Table 1) [11,27].
m
Scheme 5. Synthesis of compound 3j.

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
239
Scheme 6. Synthesis of 3s.
Scheme 7. Synthesis of optically pure 3l isomers.
Scheme 8. Synthesis of a-truxillic acid monoamides 6a and 6b.
as compared to 3, while 6 compounds (3a, 3g, 3j, 3k, 4b and 4e)
show high affinities to FABP7 (Ki 0.35e0.67 M). For the stereo-
isomers of 3l (entries 19e22), all isomers exhibit the same Ki value
within standard deviation for FABP5 (Ki 0.20 M) and FABP3 (Ki
0.80 M). For FABP7, 3l-B/D showed a little higher affinity that 3l-A,
3l-B and 3l-A/C (Ki 0.16 M vs. 0.30 M), suggesting that 3l-D
than that of FABP5 [10], the selectivity to this FABP may not be of
medicinal importance in terms of pain management. Nevertheless,
selective FABP7 inhibitors would be useful as research tools. Since
FABP3 is primarily expressed in the heart and its inhibition might
cause cardiotoxicity [12], low affinity to FABP3 appears desirable.
m
m
m
m
m
The most selective FABP5 inhibitors appear to be 4f (Ki 1.70
and 4j (Ki 1.70 M), which practically do not bind to both FABP3
(Ki > 10 M) and FABP7 (Ki >10 M) (entries 36 and 40). Another
FABP inhibitor 4k (Ki 0.89 M) exhibits higher affinity to FABP5
than 4f and 4k and does not bind to FABP3 (Ki > 10 M), while binds
modestly to FABP7 (Ki 3.54 M) (entry 41). Compounds 4j and 4k
mM)
(1R,2R,3R,4R,10S,20R) may have better affinity than other isomers.
The selective affinity to FABP5 would be beneficial since this
FABP possesses significant potential as a drug target [30e32].
FABP7 does not seem to have particular clinical significance at
present and its expression in the adult brain is substantially lower
m
m
m
m
m
m

240
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
Scheme 9. Synthesis of 3o-g and amide 6a-g
bear 9-fluorenylmethyl as the ester moiety and 2-methoxyphenyl
and 2-chlorophenyl groups in the -truxillic acid core, respec-
structural mimic of 3l. Apparently, the para-substitution is not well
tolerated as evident from the Ki values of 3i (for FABP3, 5 and 7:
a
tively. Another compound, 3q (entry 28) also bears the 9-
fluorenylmethyl ester moiety and the Ki value for FABP3
is >10 mM. Thus, it is deduced that this rather bulky ester group is
3.93, 2.52 and 2.27 mM, respectively) and thus the affinity to all
three FABPs are compromised by the substitution at this position
(entry 15).
blocking the binding to FABP3. The ortho substitution on the phenyl
group also exerts substantial effects on the FABP selectivity. Thus,
the 2-methoxy group in 4j blocks the binding to FABP7
Previously, we reported that
[11], showed a slightly better Ki than that of 3, although the solu-
bility of 3- was very poor, which prevented us from selecting it as a
lead. In the present study, we examined a pair of 3-analogs, 3o and
3o- , which bear the 6-acetamidonaphth-1-yl groups as the ester
moieties. Although 3o keeps comparable affinities to those of 3 for
the three FABPs, 3o- clearly shows substantially weaker affinities,
especially to FABP5 and FABP7 (entry 26). Thus, the -isomer series
-isomer series so far, although there
g-isomer of 3, i.e., SB-FI-49 (3-g)
g
(Ki > 10
mM), while hydrogen (3q, Ki 2.70
mM) and 2-chloro group
(4k, Ki 3.54
mM) are moderately tolerated. Then, the 2-chlorophenyl
g
group is best tolerated among the three for FABP5 binding, realizing
very good affinity to FABP5. However, the 2-chlorophenyl group is
not always well tolerated. Thus, when the ester group is (1R,2S)-2-
phenylcyclohex-1-yl (4f), 2-chlorophneyl group reduces the affin-
g
a
appears to be better than the
might be exceptions.
g
ity to FABP5 (Ki 1.70 mM) (compare entry 18 and entry 36) although
it is clear that this group blocks the binding to FABP3 and FABP7 and
brings about selectivity towards FABP5.
We also reported previously that the amide analog of 3, i.e., SB-
FI-60 (6) [11], showed very good selectivity to FABP7 with high
When the ester moieties are a good fit to FABP binding sites, the
2-methoxyphenyl group increases the affinity to FABP3. Thus, 4b
(entry 32), an analog of 3, and 4e (entry 35), an analog of 3l, bearing
affinity (Ki, 0.3
not bind to FABP3 (Ki > 10
examined 4-(5,6,7,8-tetrahydronaphth-2-yl)thiazol-2-yl amide
with - and -truxillic acid cores, 6a and 6a- , as well as biphenyl-
4-yl amide, 6b. However, none of them exhibit appreciable affinity
to FABPs (i.e., Ki > 10 M) (entries 48e50). The affinities of 6b were
m
M), fairly good affinity to FABP5 (Ki 1.6
mM), but did
mM) [6]. In the present study, we
2-methoxyphenyl groups in place of phenyl groups at the
illic acid core, exhibit high affinity to all three FABPs (Ki
0.40e0.69 M), especially to FABP3 (Ki: 4b, 0.69 M; 4e, 0.40 M),
a
-trux-
a
g
g
m
m
m
m
implying a possible introduction of additional hydrogen bonding.
Other ortho substituted phenyl groups, i.e., 2,6-dichlorophenyl
(4g), 2-bromophenyl (4h) and 2-nitrophenyl (4i) groups, are fairly
tolerated by FABP5, but none of the compounds bearing those
much weaker than the correspond ester 3i. It appears that these
rather long amide moieties do not fit to the binding sites of FABPs.
Also, the inherent rigidity of amide as compared to ester may be the
reason for the difference. Although an amide could add hydrogen
bonding donor capability for additional interactions in the binding
site, it does not seem to provide beneficial effects in these systems.
groups show Ki values less than 1 mM (entries 37e39). These groups
clearly block the binding to FABP3. 2-Bromophenyl and 2-
nitrophenyl groups also block the binding to FABP7. Interestingly,
a quite bulky 2,6-dichlorophenyl group (4g) is moderately tolerated
by FABP7.
2.6. Computational screening and analysis of SAR
Two compounds, 3f and 3h, exhibit high selectivity to FABP5 (Ki
In general, the computational analysis for selecting promising
compounds for chemical synthesis and biological assays among the
truxillic acid congeners designed was found to be useful in the
majority of cases, but with clear exceptions. The most notable
failures are monoamide 6a, 2-nitrophenyl analogs 4c and 4i,
wherein docking scores are quite favorable (for FABP5:
6a, ꢀ9.60 kcal/mol; 4c, ꢀ9.37 kcal/mol; 4i: ꢀ10.01 kcal/mol), but Ki
0.89 and 0.85
respectively), but practically do not bind to FABP3 (Ki > 10 and
9.75 M, respectively) (entries 12 and 14). Compounds 3f and 3h
mM, respectively) and FABP7 (0.78 and 0.74 mM,
m
bear 3-ethynylphenyl and biphenyl-3-yl groups, respectively, as
their ester moieties and the substituents are placed at the meta
position of the phenyl moiety in both cases. The substitution
pattern, i.e., ortho, meta and para, on the phenyl group is found to be
critical to the affinity and selectivity to FABPs. Thus, 3g, 3h and 3i all
bear biphenyl groups as their ester moieties, wherein 3g has a
biphenyl-2-yl (ortho) group, 3h a biphenyl-3-yl (meta) group and 3i
a biphenyl-4-yl group (para). Then, 3g exhibits high affinity to all
values are >10 mM.
In spite of natural limitations in the accuracy of computational
screening based only on docking scores, computational SAR anal-
ysis for representative cases has provided critical insight into the
affinity and selectivity of selected FABP inhibitors as described
below. [Note: We used the co-crystal structures of FABP5/3 (PDB ID:
5UR9) [20] and FABP7/3 (PDB ID: 5URA) [20] for these docking
three FABPs (Ki for FABP3, 5 and 7: 0.70, 0.77 and 0.35
mM,
respectively) (entry 13). It should be noted that 3g is a good

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
241
Table 1
In vitro affinities (Ki, mM) of a-truxillic acid and its congeners.
Entry
Compound
Ar
X
R
FABP3 Ki
FABP5 Ki
FABP7 Ki
cLogP
1
2
3
4
5
6
7
8
1a
1g
1i
Ph
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
NH
NH
NH
H
H
H
>10
>10
>10
2.70 0.42
3.26 0.70
2.82 0.10
>10
>10
>10
>10
>10
>10
>10
>10
>10
3.07
2.38
1.43
6.97
6.97
6.97
5.28
5.20
5.42
6.13
5.77
5.06
6.12
6.68
6.68
5.19
6.67
7.17
7.17
7.17
7.17
7.17
5.56
3.87
5.10
6.23
7.10
5.61
4.49
5.10
4.33
5.00
5.29
4.63
6.20
8.59
10.01
8.89
6.49
6.14
8.52
4.89
7.50
7.33
7.78
3.97
10.29
7.48
7.48
6.46
3,4-(MeO)2Ph
3-OH-4-(MeO)Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
3 (racemic)
(R,R,R,R)- 3
(S,S,S,S)- 3
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
1-naphthyl
1-naphthyl
1-naphthyl
benzyl
4-MeO-benzyl
4-F-benzyl
4-Br-benzyl
2-iodophenyl
3-ethynylphenyl
biphenyl-2-yl
biphenyl-3-yl
biphenyl-4-yl
20eHO-biphenyl-2-yl
2,4,5-trichlorophenyl
trans-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
(1S,2R)-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
(1S,2R)-2-phenylcyclohex-1-yl
Indan-2-yl
0.81 0.09
0.78 0.14
0.80 0.14
3.81 0.53
2.15 0.10
2.42 0.18
1.58 0.16
1.34 0.21
0.89 0.15
0.77 0.08
0.85 0.22
2.52 0.36
1.59 0.43
0.80 0.11
0.21 0.02
0.21 0.02
0.20 0.03
0.18 0.03
0.21 0.02
1.57 0.15
>10
0.97 0.18
7.08 0.44
3.92 0.75
2.56 0.16
2.17 0.32
>10
>10
0.55 0.05
>10
0.45 0.07
0.89 0.24
0.66 0.16
0.53 0.12
1.14 0.06
1.65 0.21
1.25 0.03
0.94 0.34
0.78 0.12
0.35 0.12
0.74 0.17
2.27 0.03
0.54 0.18
0.54 0.02
0.40 0.03
0.33 0.05
0.25 0.12
0.33 0.15
0.15 0.02
2.41 0.09
1.59 0.24
1.12 0.45
7.43 1.11
1.03 0.22
2.70 0.62
0.50 0.11
1.06 0.07
2.12 0.19
0.67 0.04
>10
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
>10
1.18 0.10
>10
0.70 0.42
9.75 0.79
3.93 0.16
3.52 0.53
2.98 0.85
1.08 0.37
0.83 0.15
0.88 0.14
0.64 0.16
0.82 0.09
>10
3l-A
3l-B
3l-A/C
3l-B/D
3m
3n
CF3CH2e
>10
3
6-acetamidonaphth-1-yl
5-ethynylnaphth-1-yl
9-fluorenylmethyl
cyclohexyl
3-[1-(3,6,9-trioxa-dodecanyl)-1,2,3-triazol-4-yl]phenyl
6-acetamidonaphth-1-yl
1-naphthyl
1-naphthyl
1-naphthyl
1-naphthyl
(1R,2S)-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
(1R,2S)-2-phenylcyclohex-1-yl
9-fluorenylmethyl
9-fluorenylmethyl
quinolin-5-yl
benzyl
4-MeO-benzyl
4-F-benzyl
tetrahydropyran-4-ylmethyl
biphenyl-3-yl
2.82 0.18
3.56 0.58
4.94 0.31
>10
>10
>10
1.06 0.19
0.69 0.17
>10
2.30 0.47
0.40 0.08
>10
>10
>10
>10
>10
>10
>10
>10
>10
3o-
3p
3q
3r
3s
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
g
Ph
3-MeO-4-HO-Ph
2-MeO-Ph
2-O2N-Ph
4-HO-Ph
2-MeO-Ph
2-Cl-Ph
2,6-Cl2-Ph
2-Br-Ph
2-O2N-Ph
2-MeO-Ph
2-Cl-Ph
2-MeO-Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
>10
1.06 0.34
0.40 0.03
>10
6.32 0.96
>10
>10
>10
3.54 0.77
>10
>10
0.68 0.06
1.70 0.33
1.23 0.18
2.76 0.16
>10
1.72 0.12
0.89 0.05
3.93 0.51
>10
5a
5b
5c
5d
5e
6a
>10
>10
>10
>10
>10
>10
>10
>10
N.D.a
N.D.a
N.D.a
4-(5,6,7,8-tetrahydronaphth-2-yl)thiazol-2-yl
biphenyl-4-yl
4-(5,6,7,8-tetrahydronaphth-2-yl)thiazol-2-yl
>10
>10
>10
>10
>10
>10
>10
>10
>10
6a-
6b
g
Ki values represent an average S.E. of at least three independent experiments.
a
No data were obtained due to poor solubility.
analyses (See Experimental Section). For FABP3, however, no co-
crystal structure with 3 was available. Thus, we used the high-
resolution crystal structure of FABP3 apoprotein (PDB ID: 6AQ1)
and the best docking pose of 3 with canonical interaction as the
standard (see Supplementary Data).
The first analysis is for 4-hydroxyphenyl analog 4d. Although 4d
differs from 3 only with 4-hydroxyl groups at the phenyl moieities
of the a-truxillic acid core, this small change exerted detrimental
effects on the affinity of 4d to FABP5, although 4d keeps moderate
affinities to FABP3 and FABP7. Docking study of 4d using the local
energy minimization algorithm of AutoDock Vina to find a binding
pose starting from the pose of 3 in the co-crystal structure [20], 4d
was found to be 1.29 kcal/mol less favorable than 3 (Fig. 6A). The
interaction map analysis using AutoDock Vina Tool revealed that
one of the 4-hydroxyphenyl group of 4d causes a steric clash with
the hydrophobic alkyl moiety of Lys61 (Fig. 6B), which did not exist

242
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
Fig. 6. Docking analysis of 4d in FABP5. (A) Binding pose of 4d (gray) through the local search with AutoDock 4.2 in comparison with 3 (blue). (B) Steric clash of 4-hydroxyphenyl
group with Lys61 visualized by AutoDock Tool. (C) Docking pose of 4d (gray) by AutoDock 4.2 in comparison with 3. The binding pose of 3 is from the co-crystal structure13. (For
interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
in the case of 3. When 4d is docked to FABP5 from the FABP5/3 co-
crystal structure [20], the binding pose makes a substantial change,
wherein the critical canonical interaction of the carboxylate group
with Arg129 and Tyr131 is substantially weakened, and also the 1-
naphthyl group occupies a very different space than that of 3
(Fig. 6C) to avoid the steric clash with Lys61. Accordingly, the
computational analysis of 4d binding to FABP5 provides clear
explanation for the poor binding of 4d to FABP5 observed. Also, it
has turned out that the local energy minimization approach and
docking approach give complementary analyses, leading to the
same conclusion.
As described above, 4k is found to be one of the most FABP5-
selective inhibitors (4f and 4j are even more selective, but less
potent than 4k) among the 3-analogs examined in the present
study. The computational analysis of 4k-binding to FABPs has
provided clear rationale for the observed selectivity. The docking
poses of 4k in FABP3, FABP5 and FABP7 using AutoDock Vina are
shown in Fig. 7. In the docking pose in FABP3 (Fig. 7A), the critical
canonical interaction of the carboxylate group is substantially
weakened and the location of the 9-fluorenylmethyl group is also
very different from the 1-napthyl group of 3. Thus, this analysis
indicates very weak affinity of 4k to FABP3. In fact, the observed
affinity was >10 mM. In the case of FABP5 (Fig. 7B), however, the
canonical interaction is almost completely kept in contrast and an
excellent overlay with 3 is observed, which indicates very strong
affinity to FABP5. Indeed, the observed affinity of 4k to FABP5 was
0.89 mM. For FABP7, 4k shows moderate overlay with 3, as well as
reasonable preservation of the canonical interaction, suggesting
moderate affinity to FABP7. As anticipated, the observed affinity
was 3.54 mM.
Finally, the analysis of inhibitor 4b shows that it binds equally
well to all three FABPs. As Fig. 8 illustrates, 4b shows very good
overlay with 3 in FABP3 (Fig. 8A), FABP5 (Fig. 8B) and FABP7
(Fig. 8C), suggesting high affinities of 4b to all three FABPs. In fact,
4b exhibited high affinities to FABPs (Ki 0.55e0.69 mM) and thus not
selective to a particular FABP.
Consequently, those results demonstrate that the structure-
based computational analysis and design are reasonably reliable
and can be employed for further optimization of lead compounds
that were identified in the present SAR study.
Fig. 7. Docking poses of 4k (gray) in FABPs as compared to those of 3 (blue). (A) FABP3,
(B) FABP5, (C) FABP7. The binding poses of 3 in FABP5 and FABP7 are from co-crystal
structures,13 while that in FABP3 is predicted by the AutoDock 4.2. (For interpreta-
tion of the references to colour in this figure legend, the reader is referred to the Web
version of this article.)
potent antinociceptive effects in this model (Fig. 9A) [6]. Adminis-
tration of 3f and 3l-A resulted in antinociceptive effects (Fig. 9A).
Administration of 4a and 4d (which lack affinity for FABP5) was
without effect (Fig. 9A). Compound 1a also did not produce anti-
nociceptive effects, consistent with its low affinity for FABPs, con-
firming that the a-truxillic acid scaffold lacks efficacy in this pain
model.
These observations are in agreement with those reported in our
previous work, wherein antinociceptive effects following
2.7. In Vivo Efficacy evaluations
A subset of the 3-analogs were administered to mice (40 mg/kg,
i. p.) to probe for antinociceptive activity using the complete
Freund's adjuvant (CFA) model of inflammatory pain and mea-
surement of thermal hyperalgesia via the Hargreaves plantar
apparatus [33]. In agreement with previous reports, 3 displayed

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
243
Fig. 8. Docking poses of 4b (gray) in FABPs in comparison with 3 (blue). (A) FABP3; (B) FABP5; (C) FABP7. The binding poses of 3 in FABP5 and FABP7 are from co-crystal structures,13
while that in FABP3 is predicted by the AutoDock 4.2. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Fig. 9. In Vivo Efficacy of Select Compounds. (A) Hyperalgesia was induced by intraplantar injection of CFA and thermal latencies were measured in mice following inhibitor
administration (20 mg/kg, i. p.). n ¼ 6e9 per group. **p < 0.01, ***p < 0.001 versus vehicle controls by one-way ANOVA with Dunnett post-hoc test. (B) Thermal latencies of wild-
type (WT) and FABP5 KO mice following administration of 3 (40 mg/kg, i. p.). n ¼ 6. ns indicates no significant difference, **p < 0.01 b y two-way ANOVA with Bonferonni post-hoc
test. All data are presented as mean standard error.
pharmacological FABP inhibition are correlated with inhibitor af-
finity for FABP5, while FABP3 or FABP7 inhibition are not predictors
of in vivo efficacy [6]. Furthermore, mice bearing a genetic knockout
for FABP5 (FABP5-KO) display an analgesic phenotype and no
further antinociceptive effects were observed in these animals
following administration of 3 (Fig. 9B).
Recent work indicates that FABP4 can modulate inflammation
[34] and its inhibition could contribute to the antinociceptive ef-
fects of our FABP inhibitors. However, this is unlikely because
FABP4 KO mice display normal pain responses [35] and inhibitors
that produce antinociceptive effects (e.g., 3f, 4b and 4j) do not
appreciably inhibit FABP4 (see Fig. S5). Taken together these data
lend further support to the model that FABP5 is the principal
cytosolic AEA transporter and the relevant FABP isoform for the
development of inhibitors to treat pain and inflammation.
Select compounds, exhibiting high affinity and/or selectivity to
FABP5 in vitro, were examined for their efficacy in vivo in the same
mouse model as that described above. As anticipated, 4b (FABP5 Ki
in vivo. This result may indicate that exceedingly high lipophilicity
causes unfavorable biodistribution problems in vivo even though
the compounds exhibit high affinity/potency in vitro. The estimated
cLogP values for 4b and 4j are 5.00 and 6.14, respectively, while
those for 4f and 4k are 8.59 and 8.52, respectively, which are
exceedingly high due to two 2-chlorophenyl moieties in these
molecules. A similar observation was made for the in vivo efficacy of
3l (FABP5 Ki 0.21 mM; cLogP 7.17), wherein its in vivo efficacy ap-
pears to be compromised by its high hydrophobicity despite much
stronger affinity to FABP5 in vitro.
Consequently, based on the in vivo efficacy, we have selected 4b
and 4j as new lead compounds for further optimization. In partic-
ular, 4j is a highly promising lead compound since it has almost
exclusive selectivity to FABP5 with no appreciable affinity to FABP3
and FABP7.
3. Conclusions
0.55 mM) exhibited strong antinociceptive effects, but 4b is non-
selective to FABP5, i.e., 4 b binds to FABPs 3,5, and 7 non-
discriminatively. The most FABP5-selective compound 4j (FABP5
In conclusion, we designed and synthesized a library of new
FABP inhibitors based on the
which are analogs of previously identified lead compound, 3 (SB-FI-
26), as well as some diesters, monoamides, -isomers and
substituted -truxillic acids. The SAR study clearly shows that the
-truxillic acid monoester scaffold is essential for high affinity
a-truxillic acid monoester scaffold,
Ki 1.72
nociceptive effects as observed with 3, despite a weaker affinity to
FABP5 than that of 3 (FABP5 Ki 0.81 M). Rather surprisingly,
however, 4k (FABP5 Ki 0.89 M), showing high selectivity to FABP5
(FABP3 >10 M; FABP7 3.54 M), did not exhibit good efficacy
mM; FABPs 3 and 7 Ki > 10 mM) showed potent anti-
g
a
m
a
m
binding to FABPs, because the canonical interaction of the carbox-
ylate moiety with Arg/Tyr residues is crucial for the inhibitor
m
m

244
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
binding. The critical importance of this canonical interaction is
strongly supported by the in-depth computational analysis of the
binding poses of new FABP inhibitors for their affinity and selec-
tivity to a particular FABP isoform. Among these new FABP in-
hibitors, compound 3l was found to possess 4 times higher affinity
to FABP5 than 3. However, 3l also exhibits high affinity to FABP3
and FABP7. In this category, compounds 3g, 4b and 4e are highly
potent as well. Compounds 4f and 4j are found to be exclusively
selective to FABP5, albeit their Ki values are not submicromolar
4.6 mm ꢂ 100 mm column, methanol and water, flow rate: 0.3 mL/
min, t ¼ 0e30 min, gradient of 40e95% MeOH; (3): Kinetex PFP,
2.6
m
m, 4.6 mm ꢂ 100 mm column, solvent A of water/acetonitrile
95:5 (25 mM ammonium acetate, pH 6.5), solvent B of water/
acetonitrile, 5:95 (25 mM ammonium acetate, pH 6.5), flow rate:
1.0 mL/min, t ¼ 0e15 min, gradient of 20e95% solvent B.
Single crystal X-ray crystallographic analysis was carried out on
an Oxford Gemini A Enhance diffractometer. The 2 þ 2 photo-
catalyzed cyclization was carried on Electro-Cure 4001 from
Electro-Lite Corporation.
(1.70 and 1.72 mM, respectively).
Compound 4k was found to be a highly selective inhibitor to
FABP5, which practically does not bind to FABP3, and possesses
only modest affinity to FABP7. Thus, 4f, 4j and 4k appear to be
promising FABP5-selective compounds. As noted above, the natural
product incarvillateine produces antinociceptive effects in rodents.
Our observation that 4a, a structural mimic of incarvillateine, does
not bind FABPs or produce antinociceptive effects indicates that
FABP inhibition may not account for the in vivo efficacy of incar-
villateine, unless it acts as a prodrug that is activated by meta-
bolism. Not surprisingly, in vivo efficacy evaluation of select
compounds revealed that the high affinity to FABP5 in vitro does not
necessarily be translated to high efficacy in vivo. It has turned out
that exceedingly high hydrophobicity of these compounds signifi-
cantly affects their in vivo efficacy. Thus, compounds with moderate
hydrophobicity, such as 4b and 4j (ClogP 5.00 and 6.14, respec-
tively) exhibited highly promising in vivo efficacy, while those with
exceedingly high hydrophobility, such as 4f and 4k (ClogP 8.59 and
8.52, respectively) only showed low efficacy in spite of their high
affinity to FABP5 in vitro. Based on the in vivo efficacy study, we
were able to select 4b and 4j as new and highly promising lead
compounds. Although the design of new analogs based on simple
docking has limitations, the computational analysis has provided
insight into the selectivity of certain compounds to the different
FABP isoforms.
4.2. Materials
All air- and moisture-insensitive reactions were carried out
under an ambient atmosphere, magnetically stirred, and monitored
by thin layer chromatography (TLC) using Agela Technologies TLC
plates pre-coated with 250
and visualized by fluorescence quenching under UV light. Flash
chromatography was performed on SiliaFlash® Silica Gel 40e63
60 Å particle size using a forced flow of eluent at 0.3e0.5 bar
pressure. All air- and moisture-sensitive manipulations were per-
formed using oven-dried glassware using the standard Schlenk and
glovebox techniques under nitrogen. Diethyl ether and THF were
distilled from deep purple sodium benzophenone ketyl. Dichloro-
methane, chloroform and acetonitrile were dried over CaH2 and
distilled. Dichloromethane was degassed via three freeze-pump-
thaw cycles. All other chemicals were used as received. All
deuterated solvents were purchased from Cambridge Isotope Lab-
oratories. Compound 3 (racemic) was synthesized, following the
mm thickness silica gel 60 F254 plates
mm
procedure previously reported by us [11]. a-2, 4-Di (3-methoxy-4-
tosyloxyphenyl)-cyclobutane-1,3-dicarboxylic acid (1h) was pre-
pared by the literature method [25]. Compound 5f was prepared by
the method reported previously from our laboratories [36].
g-
Truxillic anhydride was prepared by the literature method [11]. 12-
NBD-stearate [12-N-methyl-(7-nitrobenz-2-oxa-1,3-diazo)amino-
stearic acid] was purchased from Avanti Polar Lipids (Alabaster, AL,
4. Experimental section
USA).
11-[5-(Dimethylamino)-1-naphthalenesulfonyl-amino]
4.1. General methods
undecanoic acid (DAUDA) was purchased from Cayman Chemical
Company (Ann Arbor, MI, USA). 1-Anilinonaphthalene-8-sulfonic
Acid (ANS) were purchased from Cayman Chemical Company
(Ann Arbor, MI, USA).
Melting points were measured on a Thomas Hoover Capillary
melting point apparatus and are uncorrected. NMR spectra were
recorded on a Bruker Ascend 700 spectrometer operating at
700 MHz for 1H and 175 MHz for 13C, a Bruker 500 Advance spec-
trometer operating at 500 MHz and 125 MHz for 1H and 13C,
respectively, or a Bruker 400 Nanobay spectrometer operating at
400 MHz, 100 MHz, and 376 MHz for 1H, 13C, and 19F, respectively.
Chemical shifts were referenced to the residual proton and carbon-
13 peaks of solvents used for 1H and 13C NMRs, respectively (1H:
4.3. Interference compound analysis
All 57 compounds in the present work were analyzed for their
potential as pan assay interference compounds (PAINS) using the
publicly available ZINC15 web tool [37]. None of the compounds
were identified to contain PAINS patterns, nor were there any alerts
for predicted high risk of aggregation. Additionally, dynamic light
scattering (DLS) was used to experimentally probe two of the
CDCl3,
acetone-d6;
49.00); acetone-d6 d 29.84, 206.26. Signals are listed in ppm, and
d
7.26; (CD3)2SO,
d
2.50; CD3OD,
d
3.31; CD3CN,
d 1.94;
d
2.05; 13C: CDCl3,
d
77.16; DMSO‑d6,
d
39.52; CD3OD,
d
compounds in the series, 3 or 3l (20 mM), and there was no evidence
multiplicity identified as s ¼ singlet, br ¼ broad, d ¼ doublet,
t ¼ triplet, q ¼ quartet, m ¼ multiplet; J-coupling constants in Hz,
and integration. High resolution mass spectrometry analysis was
carried out on an Agilent LC-UV-TOF mass spectrometer at the
Institute of Chemical Biology and Drug Discovery, Stony Brook, NY
or on a Waters Q-TOF Ultima ESI mass spectrometer at the Mass
Spectrometry Laboratory, University of Illinois at Urbanaꢀ Cham-
paign, Urbana, IL. Purity of synthesized compounds was deter-
mined by a Shimadzu LC-2010A HT series HPLC assembly or Agilent
1100 series HPLC assembly. Purities of new compounds were all
>95%. Three analytical conditions were used for synthesized com-
pounds (25 ꢁC, 220 and 254 nm): (1) Chiralcel ODH, 250 ꢂ 4.6 mm
column, isopropanol (65%) and hexanes (35%), flow rate: 1 mL/min,
of colloidal aggregation. Furthermore, the lack of efficacy of SB-FI-
26 in FABP5-KO mice shown in Fig. 8, and the recent x-ray struc-
tures of 3 with FABP5 and FABP7 provides strong evidence that the
activity of this compound (and related analogs) are a result of
specificity for the target [20].
4.4. Chemistry
4.4.1. Optical resolution of racemic 3
To a solution of
a-truxillic acid mono-1-naphthyl ester (3)
(100 mg, 0.24 mmol) in acetone (3 mL) was added (S)-phenyl-
alaninol (36 mg, 0.24 mmol) and heated until the reaction mixture
was clear and homogeneous. The solution was allowed to cool to
room temperature and left on a benchtop overnight. After crystal
isocratic, at 220 and 254 nm; (2) Kinetex PFP, 2.6 mm,

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
245
formation, the mother liquor was removed, and the crystals were
washed with cold methanol (3 mL). Hydrochloric acid (12 N)
(1.0 mL) was added to the crystals and extracted thrice with ethyl
acetate (5 mL X 3). The organic layer was collected, dried over
magnesium sulfate, and analyzed by chiral HPLC on a Chiralcel ODH
column using 65% isopropanol and 35% hexanes as the eluent.
After the first resolution in methanol, the resulting 3 showed an
optical purity of 89% ee (er ¼ 94.5:5.5). The sample was resolved
again using the same procedure, which gave 3-A with 99% ee
(36 mg, 36% yield). The absolute configuration of this enantiomer 3-
A was unambiguously determined to be (R,R,R,R) by single crystal
X-ray crystallographic analysis (vide infra).
In the same manner, but using (R)-phenylalaninol instead of (S)-
phenylalaninol, the other enantiomer, 3-B, i.e., (S,S,S,S)-3, was ob-
tained with 95% ee (65 mg, 65% yield). When acetone was used in
place of methanol in the same procedure, 3-B was obtained with
93% ee after the first resolution. Thus, acetone appears to be a better
solvent for the optical resolution of 3.
acetone-d6)
7.20e7.24 (m, 2H), 7.25e7.28 (m, 2H), 11.90 (s, 2H). 13C NMR
(175 MHz, acetone-d6) 36.0, 44.6, 55.4, 110.5, 120.1, 127.1, 127.5,
127.9, 157.2, 173.4; HRMS (ESI) m/z: calcd for C20H20O6Hþ, 357.1338,
found, 357.1339 (
¼ 0.28 ppm).
d
3.74e3.79 (m, 8H), 4.42 (dd, J ¼ 10.1, 7.6 Hz, 2H),
d
D
4.4.3.3.
a
-2,4-Di(2-chlorophenyl)cyclobutane-1,3-dicarboxylic acid
(1c). White solid; 91% yield; m. p. 212e213 ꢁC;1H NMR (700 MHz,
acetone-d6)
d
3.89 (d, J ¼ 6.3 Hz, 2H), 4.63 (d, J ¼ 6.3 Hz, 2H), 7.08
(td, J ¼ 7.6, 1.3 Hz, 2H), 7.12 (t, J ¼ 7.0 Hz, 2H), 7.19e7.24 (m, 2H), 7.33
(d, J ¼ 7.6 Hz, 2H), 12.58 (s, 2H); 13C NMR (125 MHz, acetone-d6)
d
41.85, 42.64, 126.48, 128.01, 128.68, 129.11, 134.19, 136.53, 172.74;
HRMS (ESI) m/z: calcd for
365.0346 (
¼ 0.27 ppm).
C
18H14O4Cl2Hþ, 365.0347, found,
D
4.4.3.4.
a
-2,4-Di(2,6-dichlorophenyl)cyclobutane-1,3-dicarboxylic
acid (1d). White solid; 93% yield; m. p. decomposed at 210 ꢁC; 1H
NMR (300 MHz, acetone-d6)
d
4.80 (dd, J ¼ 4.6, 2.7 Hz, 2H), 5.31 (dd,
J ¼ 4.6, 2.7 Hz, 2H), 7.23 (m, J ¼ 9.0, 2.2 Hz, 6H), 11.18 (s, 2H); 13C
4.4.2. Single crystal X-ray diffraction analysis of 3-A (R,R,R,R)
Crystals of 3-A were selected and mounted on glass fibers using
epoxy adhesive. Each crystal was centered, and the X-ray intensity
data were measured on an Oxford Gemini A Enhance diffractom-
eter by using graphite-monochromated Cu radiation. The data was
collected using the Crysalis Pro 38.41 software [38], Wingx 2014.1
[39], Olex2 1.2 [40], and SHELX 2013 [41]. The crystal data and
structure refinement parameters are summarized in Table S1
(Supporting Information). Fractional atomic coordinates and
equivalent isotropic displacement parameters (Table S2), aniso-
tropic displacement parameters (Table S3), bond lengths (Table S4),
bond angles (Table S5), and hydrogen atom coordinates and
isotropic displacement parameters (Table S6) are also shown in the
Supporting Information. A ball and stick structure of 3-A, i.e.,
(R,R,R,R)-3, is shown in Fig. 2 and the ORTEP structure is shown in
Fig. S1 (Supporting Information). The CIF for 3-A is also available as
Supporting Information.
NMR (126 MHz, acetone-d6) d 42.40, 42.74, 128.87, 129.31, 134.33,
135.52, 172.83; HRMS (ESI) m/z: calcd for C18H12Cl4O4Hþ, 432.9568,
found, 432.9562 (
¼ 1.4 ppm).
D
4.4.3.5.
a
-2,4-Di(2-bromophenyl)cyclobutane-1,3-dicarboxylic acid
(1e). White solid; 91% yield; m. p. 218e220 ꢁC; 1H NMR (700 MHz,
DMSO‑d6)
d
3.86 (dd, J ¼ 4.0, 2.4 Hz, 2H), 4.63 (d, J ¼ 6.4 Hz, 2H),
6.99e7.04 (m, 2H), 7.16 (t, J ¼ 7.5 Hz, 2H), 7.33 (d, J ¼ 7.7 Hz, 2H),
7.40 (d, J ¼ 7.7 Hz, 2H), 12.58 (s, 2H); 13C NMR (175 MHz, DMSO‑d6)
d
42.6, 43.9, 124.7, 127.1, 128.5, 129.1, 132.3, 137.8, 173.6; HRMS (ESI)
m/z: calcd for
¼ 0.44 ppm).
C
18H14Br2O4Hþ, 451.9259, found, 451.9261;
(D
4.4.3.6.
a-2,4-Di(2-nitrophenyl)cyclobutane-1,3-dicarboxylic
acid
(1f). White solid; 55% yield; m. p. 206e208 ꢁC; 1H NMR (300 MHz,
MeOD) 3.20 (m, 2H), 3.81 (m, 2H), 7.15 (m, 2H), 7.29 (m, 4H), 7.53
(m, 2H); 13C NMR (125 MHz, MeOD)
40.90, 44.06, 123.13, 127.93,
129.13, 132.26, 133.74, 149.52; HRMS (ESI) m/z: calcd for
18H14N2O8Hþ, 387.0825, found, 387.0823 (
¼ 0.05 ppm).
d
d
4.4.3. Chemical synthesis
C
D
Note: For the nomenclatures of
derivatives and their monoesters, IUPAC nomenclatures are used,
but with the sign “ -“ or “ -“ designations, wherein “ -“ stands for
a- and g-truxillic acid, their
4.4.3.7.
a
-2,4-Di(3,4-dimethoxyphenyl)cyclobutane-1,3-dicarboxylic
acid (1g). White solid; 90% yield; 1H NMR (300 MHz, acetone-d6)
3.69e3.71 (m, 2H), 3.73 (s, 6H), 3.76 (s, 6H), 4.16e4.25 (m. 2H),
6.80e6.96 (m, 6H), 12.06 (s, 2H); 13C NMR (176 MHz, acetone-d6)
41.10, 47.05, 55.89, 55.92, 111.88, 112.30, 119.92, 132.41, 148.08,
148.75, 173.58; HRMS (ESI) m/z: calcd for C22H24O8Hþ, 417.1544,
found, 417.1548 (
¼ 0.96 ppm).
a
g
a
(1S*,2S*,3S*,4S*) or (1R*,2R*,3R*,4R*) stereochemistry, i.e., racemic
with all four carbon centers bearing the same S or R configuration,
d
while “g-“ stands for (1S*,2R*,3S*,4S*) or (1R*,2S*,3R*,4R*) stereo-
d
chemistry, i.e., racemic with only C2 carbon center bearing opposite
configuration.
D
4.4.3.1.
a-2,4-Diphenylcyclobutane-1,3-dicarboxylic
acid
(1a).
4.4.3.8. a-2,4-Di(3-methoxy-4-tosyloxyphenyl)cyclobutane-1,3-
(E)-Cinnamic acid (1.0 g, 6.7 mmol) was placed in a pyrex dish and
exposed to light at 360 nm and an intensity of 280 nW/cm2 for 5
days with periodic shaking. This process was performed in the solid
state and monitored by 1H NMR. After completion of the photore-
action, the white solid was washed with diethyl ether (20 mL) to
dicarboxylic acid (1h). To a solution of (E)-ferulic acid (7.00 g,
36.0 mmol) in H2O (15 mL) was added NaOH (7.21 g, 180 mmol) and
Tosyl chloride (21.75 g, 114 mmol) portions. The reaction mixture
was stirred overnight and was quenched with 6 M HCl (10 mL) and
diluted with water (50 mL). The precipitation was collected by
vacuum filtration and the solid was washed with water. The crude
solid was dried over phosphorus pentoxide in vacuo. The product
was purified by recrystallization in EtOH/H2O (9:1) to afford 3-
methoxy-4-tosyloxy-(E)-cinnamic acid (11.55 g, 92%) as a light
give
a-truxillic acid (0.99 g, 99% yield) as a white solid. m. p.
>230 ꢁC; 1H NMR (300 MHz, acetone-d6)
d
3.99 (dd, J ¼ 10.0, 7.6 Hz,
2H), 4.44 (dd, J ¼ 10.0, 7.6 Hz, 2H), 7.23 (t, J ¼ 7.1 Hz, 2H), 7.33 (t,
J ¼ 7.4 Hz, 4H), 7.42 (d, J ¼ 7.4 Hz, 4H), 10.62 (s, 1H); FIA (negative
mode), calcd C18H16O4 296.1, found (M ꢀ H)- 295.0. The analytical
data was consistent with literature values [11].
yellow crystal. m. p. 201e202 ꢁC; 1H NMR (300 MHz, CDCl3)
d2.47
(s, 3H), 3.63 (s, 3H), 6.40 (d, J ¼ 16.0 Hz, 1H), 7.01 (s, 1H), 7.10 (d,
J ¼ 8.2 Hz, 1H), 7.20 (d, J ¼ 8.2 Hz, 1H), 7.33 (d, J ¼ 8.2 Hz, 2H), 7.71
(d, J ¼ 16.0 Hz, 1H), 7.78 (d, J ¼ 8.2 Hz, 2H); FIA m/z: calcd for
In the same manner, other mono ester 1b to 1i were synthesized
and characterized, starting from corresponding substituted (E)-
cinnamic acids.
C
17H16O6SHþ, 349.1, found, 349.1. The analytical data was consistent
with literature values [25].
4.4.3.2.
a
-2,4-Di(2-methoxylphenyl)cyclobutane-1,3-dicarboxylic
3-Methoxy-4-tosyloxy-(E)-cinnamic acid (0.600 g, 0.86 mmol)
was placed in a pyrex dish and exposed to light at 350 nm and an
acid (1b). White solid; 92% yield; m. p. >230 ꢁC; 1H NMR (700 MHz,

246
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
intensity of 280 nW/cm2 for 4 days with periodic shaking. This
process was performed in the solid state and monitored by 1H NMR.
After completion of the photoreaction, the white solid was washed
with diethyl ether (20 mL) to give 1h (0.600 g, 88% yield) as a white
suspended in thionyl chloride (3 mL) was added three drops of DMF
and the mixture was heated to reflux for 3 h with stirring. Excess
thionyl chloride and DMF were removed in vacuo to give the cor-
responding
a-truxillic acid dichloride (2a) as light yellow solid,
solid. m. p.: >230 ꢁC; 1H NMR (300 MHz, acetone-d6)
d
2.45 (s, 6H),
3.51 (s, 6H), 3.98 (dd, J ¼ 10.5, 7.1 Hz, 2H), 4.41 (dd, J ¼ 10.5, 7.1 Hz,
2H), 6.93e7.11 (m, 6H), 7.40e7.48 (m, 4H), 7.62e7.73 (m, 4H). 13
NMR (126 MHz, acetone-d6) 21.63, 42.24, 47.25, 55.97, 113.62,
which was used directly in the subsequent reaction. To a solution of
2a in THF (10 mL) were added benzyl alcohol (0.53 mmol, 0.8 eq)
and pyridine (4.0 mmol), and the reaction mixture was stirred for
2 h. The reaction was quenched by adding distilled water (10 mL)
with stirring for 30 min. The resulted solution was diluted with
ethyl acetate (15 mL) and washed with aqueous copper sulfate
(5 mL x 3) and water (5 mL x 3). The organic layer was collected,
dried over MgSO4, and concentrated in vacuo. The crude mixture
was purified by flash chromatography on silica gel (hexanes/AcOEt/
AcOH ¼ 79/20/1) as the eluent to give mono-ester 3a and diester
5a.
C
d
120.40, 124.41, 129.39, 130.47, 134.16, 138.09, 140.88, 146.32, 152.43,
173.09; FIA m/z: calcd for C34H32O12S2Hþ, 697.1, found, 697.1. The
analytical data was consistent with literature values [25].
4.4.3.9.
a-2,4-Di(4-hydroxy-3-methoxyphenyl)cyclobutane-1,3-
dicarboxylic acid (1i). To a solution of 1h (1.00 g, 1.44 mmol) in
MeOH (20 mL) was added KOH (0.966 g, 17.3 mmol) and the reac-
tion mixture was stirred at room temperature overnight. The re-
action was quenched with 6 M HCl (pH 1) to form precipitate. The
product was collected by filtration and dried over P2O5 in vacuo to
afford 1i as a white solid (0.518 g, 83%): m. p. >240 ꢁC; 1H NMR
3a: White solid; 61% yield; m. p. 143e144 ꢁC; 1H NMR (500 MHz,
acetone-d6)
d
3.98e4.06 (m, 2H), 4.48 (m, 2H), 4.66 (d, J ¼ 12.3 Hz,
1H), 4.82 (d, J ¼ 12.3 Hz, 1H), 7.04 (m, 2H), 7.2e7.30 (m, 5H),
7.36e7.30 (m, 4H), 7.41 (m, 4H); 13C NMR (125 MHz, acetone-d6)
(300 MHz, acetone-d6)
d
3.32 (d, J ¼ 9.2 Hz, 2H), 3.85 (m, 8H),
d41.47, 41.76, 46.24, 46.68, 65.91, 126.80, 126.94, 127.65, 127.76,
127.85, 128.08, 128.20, 128.26, 128.26, 128.27, 128.30, 128.31, 135.99,
4.39e4.28 (m, 2H), 6.72e6.86 (m, 4H), 6.98 (m, 2H); 13C NMR
(125 MHz, acetone-d6)
d
41.29, 46.96, 55.35, 111.59, 114.59, 120.21,
139.34, 139.39, 171.54, 172.20; HRMS (ESI) m/z calculated for
130.98, 145.54, 147.13, 172.56; FIA m/z: calcd for C20H20O8Hþ, 389.1,
C
25H22O4Hþ: 387.1591, found 387.1593 (
D
¼ 0.52 ppm).
5a: White solid; 13% yield; m. p. 161e163 ꢁC; 1H NMR (500 MHz,
acetone-d6)
found, 389.1. These data are consistent with literature values [42].
d
3.99e4.17 (m, 2H), 4.52 (dd, J ¼ 10.38, 7.32 Hz, 2H),
4.4.3.10.
a
-2,4-Di(4-tert-butyldimethylsiloxylphenyl)cyclobutane-
4.64 (d, J ¼ 12.21 Hz, 2H), 4.81 (d, J ¼ 12.51 Hz, 2H), 6.98e7.08 (m,
1,3-dicarboxylic acid (1j). To a solution of (E)-4-hydroxycinnamic
acid (2.00 g, 12.2 mmol) in dichloromethane (20 mL) at 0 ꢁC was
added diisopropylethylamine (DIPEA, 4.72 g, 36.6 mmol) and tert-
butyldimethylsilyl chloride (TBDMSCl, 4.59 g, 30.5 mmol) drop-
wise. The reaction mixture was slowly warmed up to room tem-
perature and stirred overnight. The reaction mixture was diluted
with EtOAc (50 mL) then quenched with water (20 mL). The sepa-
rated organic layer was washed with 1 M HCl (2 ꢂ 30 mL), brine
(30 mL), then dried over MgSO4 and concentrated in vacuo. To the
resulting residue in THF (15 mL) was added a solution of K2CO3
(2.53 g, 18.3 mmol) in water (5 mL) and the reaction mixture was
stirred for 4 h. Upon completion, the reaction mixture was diluted
with EtOAc (60 mL) and quenched with slow addition of 1 M HCl
(15 mL). The layers were separated and the organic layer was
washed with 1 M HCl (2 ꢂ 20 mL), brine (30 mL), then dried over
MgSO4 and concentrated in vacuo. The crude product was purified
by column chromatography on silica gel to afford (E)-4-tert-butyl-
dimethylsilyloxycinnamic acid as a white solid (2.624 g, 77%): m. p.
4H), 7.23e7.30 (m, 7H), 7.30e7.36 (m, 4H), 7.36e7.45 (m, 4H); 13C
NMR (126 MHz, acetone-d6)
129.1, 129.2, 129.3, 136.9, 140.1, 172.3; HRMS (ESI) m/z calculated for
32H28O4Hþ: 477.206, found 477.2059 (
¼ 0.34 ppm).
In the same manner, other mono esters 3b to 3i, 3k to 3r and di-
d 42.6, 47.5, 66.9, 128.0, 128.7, 128.8,
C
D
esters 5b to 5e were synthesized and characterized, starting from
2,4-diphenylcyclobutene-1,3-dicarboxylic acid 1a.
a-
4.4.3.12. 1-(4-Methoxybenzyl)
diphenylcyclobutane-1-carboxylate (3b) and 1,3-di(4-methoxy)
benzyl -2,4-diphenylcyclobutane-1,3-dicarboxylate (5b). 3b: White
solid; 40% yield; m. p. 99e102 ꢁC; 1H NMR (500 MHz, acetone-d6)
3.78 (s, 3H), 3.93e4.07 (m, 2H), 4.40e4.50 (m, 2H), 4.58 (d,
a-3-hydroxycarbonyl-2,4-
a
d
J ¼ 11.9 Hz, 1H), 4.74 (d, J ¼ 11.9 Hz, 1H), 6.79e6.87 (m, 2H),
6.93e7.03 (m, 2H), 7.21e7.39 (m, 8H), 7.41 (d, J ¼ 7.32 Hz, 2H); 13C
NMR (126 MHz, acetone-d6)
d 42.4, 42.7, 47.1, 47.6, 55.6, 66.7, 114.6,
127.7, 127.8, 128.6, 128.7, 129.1, 129.2, 130.9, 140.3, 160.6, 172.5,
173.1; HRMS (ESI) m/z calcd for C26H24O5NHþ4 : 434.1962, found
127e128 ꢁC 1H NMR (300 MHz, CDCl3)
6.33 (d, J ¼ 15.9 Hz, 1H), 6.87 (d, J ¼ 8.5 Hz, 2H), 7.47 (d, J ¼ 8.6 Hz,
2H), 7.76 (d, J ¼ 15.9 Hz, 1H); FIA m/z: calcd for C15H22O3SiHþ, 279.1,
found, 279.1. These data are consistent with literature values [42].
d0.24 (s, 6H), 1.00 (s, 9H),
434.1964 (
5b: White solid; 12% yield; m. p.145e148 ꢁC; 1H NMR (500 MHz,
acetone-d6)
D
¼ 0.42 ppm).
d
3.78 (s, 6H), 3.97e4.08 (m, 2H), 4.47 (dd, J ¼ 10.38,
7.32 Hz, 2H), 4.56 (d, J ¼ 12.2 Hz, 2H), 4.73 (d, J ¼ 12.2 Hz, 2H),
(E)-4-Tert-butyldimethylsilyloxycinnamic
acid
(1.00 g,
6.78e6.86 (m, 4H), 6.93e7.01 (m, 4H), 7.23e7.30 (m, 2H), 7.30e7.40
3.59 mmol) was placed in a pyrex dish and exposed to light at
360 nm and an intensity of 280 nW/cm2 for 4 days with periodic
shaking. This process was performed in the solid state and moni-
tored by 1H NMR. After completion of the photoreaction, the re-
action mixture was purified by column chromatography to afford 1j
as white solid (0.846 g, 84%): m. p. >230 ꢁC; 1H NMR (500 MHz,
(m, 8H); 13C NMR (125 MHz, acetone-d6)
d
42.5, 47.5, 55.6, 66.7,
114.6, 127.9, 128.6, 129.3, 130.9, 140.1, 160.6, 172.3; HRMS (ESI) m/z
calcd for C34H32O6NHþ4 554.2536, found 544.2537 (
¼ 0.2 ppm).
D
4.4.3.13. 1-(4-Fluorobenzyl)
a
-3-hydroxycarbonyl-2,4-
diphenylcyclobutane-1-dicarboxylate (3c) and di(4-fluorobenzyl)
2,4-diphenylcyclobutane-1,3-dicarboxylate (5c). 3c: White solid;
34% yield; m. p. 145e148 ꢁC; 1H NMR (500 MHz, acetone-d6)
4.03
a
-
acetone-d6)
d
0.22 (s, 12H), 1.01 (s, 18H), 3.90 (dd, J ¼ 10.4, 7.3 Hz,
2H), 4.36 (dd, J ¼ 10.4, 7.3 Hz, 2H), 6.87e6.80 (m, 4H), 7.30 (d,
d
J ¼ 8.5 Hz, 4H). 13C NMR (126 MHz, acetone-d6)
d
ꢀ5.18, 17.86, 25.14,
(dd, J ¼ 10.38, 7.32 Hz, 2H), 4.35e4.58 (m, 2H), 4.68 (d, J ¼ 12.21 Hz,
40.88, 46.74, 119.61, 128.86, 132.41, 154.48, 172.36; HRMS (ESI) m/z
1H), 4.80 (d, J ¼ 12.21 Hz, 1H), 6.93e7.14 (m, 4H), 7.21e7.45 (m,
calculated for
¼ 0.54 ppm).
C
30H44O6Si2Hþ: 557.2749, found 557.2952
10H), 10.63 (br. s., 1H); 13C NMR (125 MHz, acetone-d6)
d 42.4, 42.7,
(D
47.2, 47.6, 66.1, 115.8, 116.0, 127.8, 127.9, 128.6, 128.7, 129.2, 129.3,
131.2, 131.3, 133.1, 140.2, 140.3, 162.4, 164.3, 172.4, 173.0; HRMS (ESI)
m/z calculated for C25H21FO4 (M þ H)þ: 405.1497, found 405.1502
4.4.3.11. 1-Benzyl
carboxylate (3a) and 1,3-dibenzyl
dicarboxylate (5a). To -truxillic acid (1a, 200 mg, 0.66 mmol)
a
-3-hydroxycarbonyl-2,4-diphenylcyclobutane-1-
a
-2,4-diphenylcyclobutane-1,3-
(D
¼ 1.3 ppm).
a
5c: White solid; 15% yield; m. p. 103e105 ꢁC; 1H NMR (500 MHz,

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
247
acetone-d6)
d
4.06 (dd, J ¼ 10.53, 7.48 Hz, 2H), 4.51 (dd, J ¼ 10.38,
2H), 6.53 (m, 2H), 6.54e6.58 (m, 2H), 7.34 (t, J ¼ 7.6 Hz, 2H), 7.39 (m,
4H), 7.43e7.55 (m, 14H), 7.65 (d, J ¼ 7.6 Hz, 4H); 13C NMR (175 MHz,
7.32 Hz, 2H), 4.66 (d, J ¼ 12.36 Hz, 2H), 4.78 (d, J ¼ 12.36 Hz, 2H),
6.93e7.17 (m, 8H), 7.20e7.44 (m,10H); 13C NMR (125 MHz, acetone-
acetone-d6)
d 41.60, 46.38, 120.06, 120.52, 124.02, 126.86, 127.50,
d6)
133.1, 133.1, 140.0, 162.4, 164.3, 172.2; HRMS (ESI) m/z calcd for
32H26F2O4Hþ 513.1872, found 513.1882 (
¼ 2.0 ppm).
d
42.5, 47.4, 66.1, 115.8, 116.0, 128.0, 128.6, 129.3, 131.3, 131.3,
127.72, 128.31, 128.62, 128.83, 129.53, 139.03, 139.62, 142.16, 151.21,
170.29; HRMS (ESI) m/z calcd for C42H32O4Hþ 601.2373, found
C
D
601.2371 (
D
¼ 0.33 ppm).
4.4.3.14. 1-(4-Bromobenzyl)
a
-3-hydroxycarbonyl-2,4-
diphenylcyclobutane-1-carboxylate (3d). White solid; 44% yield;
m. p. 175e177 ꢁC; 1H NMR (500 MHz, acetone-d6)
3.92e4.11 (m,
4. 4. 3.19. 1-(Biphenyl-4-yl) a-3-hydroxycarbonyl-2, 4-
diphenylcyclobutane-1-carboxylate (3i). White solid; 40% yield; m.
p. 190 ꢁC (decomp.); 1H NMR (700 MHz, acetone-d6) 3.97 (t,
J ¼ 10.5 Hz, 1H), 4.17 (t, J ¼ 10.5 Hz, 1H), 4.52 (t, J ¼ 10.5 Hz, 1H), 4.78
(t, J ¼ 10.5 Hz, 1H), 6.50 (d, J ¼ 8.5 Hz, 2H), 7.29 (t, J ¼ 7.3 Hz, 1H),
d
2H), 4.40e4.57 (m, 2H), 4.69 (d, J ¼ 12.66 Hz, 1H), 4.78 (d,
J ¼ 12.66 Hz,1H), 6.96 (d, J ¼ 8.54 Hz, 2H), 7.20e7.29 (m, 2H), 7.32 (t,
J ¼ 7.48 Hz, 4H), 7.35e7.40 (m, 2H), 7.40e7.49 (m, 4H); 13C NMR
7.33e7.46 (m, 8H), 7.48e7.53 (m, 5H), 7.59 (d, J ¼ 7.3 Hz, 2H),
d
10.78
(125 MHz, acetone-d6)
d
42.4, 42.7, 47.2, 47.5, 66.0, 122.3, 127.8,
(s, 1H). 13C NMR (175 MHz, acetone-d6)
d 42.10, 44.52, 46.03, 46.28,
127.9, 128.6, 128.7, 129.2, 129.3, 131.0, 132.3, 136.4, 140.2, 140.3,
121.93, 126.67, 126.78, 127.24, 127.30, 127.54, 128.38, 128.45, 128.81,
128.85, 138.47, 138.50, 140.07, 142.22, 150.16, 170.09, 171.76: HRMS
172.4, 173.1; HRMS (ESI) m/z calcd for C25H21BrO4Hþ 465.0696,
found 465.0697 (
D
¼ 0.27 ppm).
(ESI) m/z calcd for
¼ 1.11 ppm).
C
30H24O4Hþ: 449.1747, found 449.1752
(D
4 . 4 . 3 .15 . 1, 3 - B i s ( t e t ra h y d r o p y ra n - 4 - m e t h y l )
a
- 2 , 4 -
diphenylcyclobutane-1,3-dicarboxylate (5d). White solid; 21%
4.4.3.20. 1-(20-Hydroxybiphenyl-2-yl)
diphenylcyclobutane-1-carboxylate (3j). 20-TIPSO-biphenyl-2-yl
a-3-hydroxycarbonyl-2,4-
yield; m. p. 143e144 ꢁC; 1H NMR (500 MHz, acetone-d6)
d
0.91e1.10
a-
(m, 4H), 1.16e1.31 (m, 4H), 1.38e1.55 (m, 2H), 3.17 (m, 4H), 3.58 (d,
J ¼ 6.41 Hz, 4H), 3.67e3.81 (m, 4H), 3.90e4.08 (m, 2H), 4.48 (m,
2H), 7.19e7.31 (m, 2H), 7.31e7.46 (m, 8H); 13C NMR (125 MHz,
2, 4-diphenylcyclobutane-1, 3-dicarboxylate (3j-i) was prepared in
the same manner as that for 3a. The crude product was purified by
column chromatography on silica gel to afford 3j-i as white solid
(59% yield). The TIPS protecting group of 3j-i was removed by 1 M
tetrabutylammonium fluoride (TBAF) in THF at room temperature
to give 3j as a white solid (99% yield): m. p. 55e57 ꢁC; 1H NMR
acetone-d6)
140.3, 172.3; HRMS (ESI) m/z calculated for C30H36O6Hþ: 493.2585,
found 493.2591 (
¼ 1.24 ppm).
d 35.2, 42.5, 47.7, 67.8, 67.8, 69.4, 127.9, 128.7, 129.3,
D
(500 MHz, acetone-d6)
d
3.99 (dd, J ¼ 10.5, 7.0 Hz, 1H), 4.11 (dd,
4.4.3.16. 1-(3-ethynylphenyl)
diphenylcyclobutane-1-carboxylate (3f). White solid; 67% yield;
m.]p. 168e169 ꢁC; 1H NMR (700 MHz, CDCl3)
3.06 (s, 1H).
a
-3-hydroxycarbonyl-2,4-
J ¼ 10.5, 7.0 Hz, 1H), 4.37 (dd, J ¼ 10.5, 7.0 Hz, 1H), 4.51 (dd, J ¼ 10.5,
7.0 Hz, 1H), 6.15 (m, 1H), 6.93e6.87 (m, 1H), 7.03 (d, J ¼ 8.1 Hz, 1H),
7.11 (m, 1H), 7.26 (m, 12H), 7.43 (t, J ¼ 7.5 Hz, 2H), 7.48 (d, J ¼ 7.5 Hz,
d
4.12e4.07 (m, 1H), 4.19 (m, 1H), 4.53e4.60 (m, 2H), 6.32 (d,
J ¼ 8.1 Hz, 1H), 6.39 (s, 1H), 7.14 (t, J ¼ 8.1 Hz, 1H), 7.25 (d, J ¼ 7.7 Hz,
1H), 7.28 (s, 2H), 7.33e7.46 (m, 9H). 13C NMR (175 MHz, CDCl3)
2H), 10.67 (s, 1H), 8.21 (s, 1H); 13C NMR (175 MHz, CDCl3)
d 41.10,
41.54, 45.88, 46.34, 116.27, 120.57, 122.47, 123.75, 125.32, 126.69,
126.90, 127.07, 127.46, 127.55, 128.25, 128.30, 128.43, 128.74, 129.06,
129.55, 129.65, 129.91, 130.10, 130.57, 131.58, 138.03, 138.35, 148.24,
152.88, 170.41; HRMS (ESI) m/z: calcd for C30H25O5Hþ, 465.1697,
d
41.30, 41.80, 46.04, 46.93, 77.83, 82.41, 122.15, 123.22, 125.08,
127.43, 127.49, 127.73, 127.87, 128.63, 128.83, 129.12, 129.60, 138.09,
149.92, 170.38; HRMS (ESI) m/z calculated for
26H20O4Hþ:
397.1434, found 397.1446 (
¼ 3.02 ppm).
C
found, 465.1699 (
D
¼ 0.43 ppm).
D
4.4.3.21. 1-(2,4,5-Tricholrobenzyl)
a
-3-hydroxycarbonyl-2,4-
4. 4. 3.17. 1-(Biphenyl-2-yl)
a
-3-hydroxycarbonyl-2, 4-
diphenylcyclobutane-1-carboxylate (3k). White solid; 70% yield;
diphenylcyclobutane-1-carboxylate (3g). White solid; 21% yield; m.
m. p. 191e193 ꢁC; 1H NMR (700 MHz, acetone-d6)
d
4.14 (dd,
p. 195e196 ꢁC; 1H NMR (500 MHz, acetone-d6)
d
3.81 (dt, J ¼ 11.67,
J ¼ 10.7, 7.2 Hz, 1H), 4.38 (dd, J ¼ 10.7, 7.2 Hz, 1H), 4.60 (dd, J ¼ 10.7,
7.2 Hz, 1H), 4.68 (dd, J ¼ 10.7, 7.2 Hz, 1H), 6.03 (s, 1H), 7.27 (t,
J ¼ 7.4 Hz, 1H), 7.36 (t, J ¼ 7.6 Hz, 2H), 7.41 (t, J ¼ 7.3 Hz, 1H), 7.48 (dt,
J ¼ 15.1, 9.8 Hz, 4H), 7.57 (d, J ¼ 7.5 Hz, 2H), 7.70 (d, J ¼ 7.6 Hz, 1H),
10.49 Hz, 2H), 4.39 (t, J ¼ 10.07 Hz, 1H), 4.68 (t, J ¼ 10.68 Hz, 1H),
5.99 (dd, J ¼ 8.09, 1.07 Hz, 1H), 7.11e7.19 (m, 3H), 7.20e7.28 (m, 2H),
7.28e7.40 (m, 11H), 7.40e7.46 (m, 2H); 13C NMR (126 MHz,
acetone-d6)
d
42.3, 45.4, 46.8, 47.0, 123.4, 127.1, 127.4, 128.1, 128.4,
10.75 (s,1H); 13C NMR (175 MHz, acetone-d6)
d 41.2, 42.0, 45.9, 46.5,
129.2, 129.3, 129.6, 129.8, 131.3, 135.8, 138.5, 139.2, 143.2, 148.7,
125.0, 126.3, 127.0, 127.5, 127.8, 128.8, 129.0, 129.8, 130.7, 130.9,
170.6, 172.9; HRMS (ESI) m/z calcd for C30H24O4Hþ: 449.1747, found
138.8, 139.1, 145.9, 169.2, 171.9; HRMS (ESI) m/z calcd for
449.1754 (
D
¼ 1.55 ppm).
C
24H18Cl3O4Hþ: 475.0265; found, 475.0264 (
D
¼ 0.21 ppm).
4. 4. 3.18. 1-(Biphenyl-3-yl)
a
-3-hydroxycarbonyl-2, 4-
4.4.3.22. 1-[(1R,2S)-2-phenylcyclohexyl] a-3-hydroxycarbonyl-2,4-
diphenylcyclobutane-1-carboxylate (3h) and 1,3-di(biphenyl-3-yl)
-2,4-diphenylcyclobutane-1,3-dicarboxylate (5e). 3h: white solid;
42% yield; m. p. 208e209 ꢁC; 1H NMR (500 MHz, acetone-d6)
4.14
diphenylcyclobutane-1-carboxylate (3l). White solid; 61% yield; m.
p. 192e193 ꢁC; 1H NMR (700 MHz, CDCl3)
0.80 (m, 1H), 1.22e1.39
a
d
d
(m, 2H), 1.47e1.59 (m, 2H), 1.64e1.76 (m, 2H), 1.90 (d, J ¼ 13.4 Hz,
1H), 2.57 (m, 1H), 3.65 (m, 2H), 3.96 (m, 1H), 4.24e4.32 (m, 1H),
4.69 (m, 1H), 7.00 (d, J ¼ 7.0 Hz, 2H), 7.07 (d, J ¼ 7.0 Hz, 2H), 7.15 (d,
J ¼ 7.0 Hz, 2H), 7.20e7.29 (m, 7H), 7.31 (m, 2H), 10.64 (s, 1H). 13C
(dd, J ¼ 10.8, 7.2 Hz, 1H), 4.29e4.35 (m, 1H), 4.61 (dd, J ¼ 10.8,
7.2 Hz, 1H), 4.67 (dd, J ¼ 10.8, 7.2 Hz, 1H), 6.53 (t, J ¼ 1.9 Hz, 1H),
6.55e6.57 (m, 1H), 7.29 (d, J ¼ 7.4 Hz, 1H), 7.33 (d, J ¼ 7.9 Hz, 1H),
7.36e7.41 (m, 4H), 7.44e7.55 (m, 9H), 7.59 (d, J ¼ 7.5 Hz, 2H); 13C
NMR (175 MHz, CDCl3)
d 24.7, 25.8, 31.4, 33.7, 40.6, 42.3, 46.1, 47.3,
NMR (175 MHz, CDCl3)
d41.34, 41.79, 45.78, 46.95, 76.86, 77.04,
49.9, 76.9, 126.7, 126.8, 127.1, 127.3, 127.6, 128.5, 128.6, 138.6, 139.2,
143.2, 171.0, 177.7; HRMS (ESI) m/z: calcd for
30H31O4Hþ:
455.2217; found, 455.2221 (
¼ 0.81 ppm).
77.22. 120.10, 120.12, 124.55, 127.16, 127.49, 127.56, 127.62, 127.72,
127.99, 128.67, 128.83, 129.50, 138.14, 138.18, 139.91, 142.55, 150.58,
170.65, 174.19; HRMS (ESI) m/z calcd for C30H24O4Hþ: 449.1747,
C
D
found 449.1753 (
5e: white solid; 15% yield; m. p. 184e185 ꢁC; 1H NMR (500 MHz,
acetone-d6)
4.45 (dd, J ¼ 10.8, 7.3 Hz, 2H), 4.81 (dd, J ¼ 10.8, 7.3 Hz,
D
¼ 1.33 ppm).
4 . 4 . 3 . 2 3 . 1 - ( I n d a n - 2 - y l ) a - 3 - h y d r o x y c a r b o n y l - 2 , 4 -
diphenylcyclobutane-1-carboxylate (3m). White solid; 34% yield;
d
m. p. 174e175 ꢁC; 1H NMR (500 MHz, acetone-d6)
d
2.05 (dt,

248
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
J ¼ 4.35, 2.25 Hz, 1H), 2.10 (dd, J ¼ 17.09, 2.44 Hz, 1H), 2.74 (dd,
J ¼ 16.94, 2.59 Hz, 1H), 2.87 (dd, J ¼ 16.94, 6.26 Hz, 1H), 3.11 (dd,
J ¼ 17.09, 6.41 Hz, 1H), 3.90 (dd, J ¼ 10.68, 6.71 Hz, 1H), 3.97e4.07
(m, 1H), 4.36e4.48 (m, 2H), 5.14e5.21 (m, 1H), 7.05e7.18 (m, 4H),
7.19e7.25 (m, 1H), 7.27e7.37 (m, 7H), 7.40 (d, J ¼ 7.32 Hz, 2H); 13C
4.4.3.28. 1-(9-Fluorenylmethyl)
diphenylcyclobutane-1-carboxylate (3q). White solid; 35% yield;
m. p. 180e181 ꢁC; 1H NMR (700 MHz, CDCl3)
a-3-hydroxycarbonyl-2,4-
d
3.85 (t, J ¼ 6.7 Hz,
1H), 4.00 (m, 2H), 4.08 (m, 2H), 4.37 (m,1H), 4.47 (m,1H), 7.25e7.39
(m, 12H), 7.40e7.48 (m, 4H), 7.80 (d, J ¼ 7.5 Hz, 2H); 13C NMR
NMR (125 MHz, acetone-d6)
d
39.8, 40.0, 42.3, 42.6, 46.9, 47.5, 76.0,
(175 MHz, CDCl3) d 41.6, 41.8, 46.3, 46.7, 47.2, 66.6, 120.1, 125.0,
125.3, 125.5, 127.3, 127.4, 127.7, 127.9, 128.7, 128.7, 129.1, 129.2, 140.1,
125.1, 127.2, 127.4, 127.5, 128.6, 128.7, 138.3, 138.4, 141.3, 141.4, 143.6,
140.4, 141.4, 173.1; HRMS (ESI) m/z calcd for C27H24O4Hþ: 413.1747,
144.1, 171.9, 177.2; HRMS (ESI) m/z: calcd for C32H27O4Hþ 475.1904;
found 413.1749 (
D
¼ 0.43 ppm).
found, 475.1907 (
D
¼ 0.63 ppm).
4.4.3.29. 1-(2,2,2-Trifluoroethyl)
a
-3-hydroxycarbonyl-2,4-
4.4.3.24. 1-(2,2,2-Trifluoroethyl)
diphenylcyclobutane-1-carboxylate (3n). White solid, 51% yield;
mp 158e159 ꢁC; 1H NMR (500 MHz, CDCl3)
3.84 (m, 1H),
4.00e4.11 (m, 2H), 4.19 (m, 1H), 4.47 (m, 2H), 7.24e7.39 (m, 10H);
13C NMR (125 MHz, CDCl3)
41.32, 41.53, 46.00, 46.44, 60.04, 60.33,
a-3-hydroxycarbonyl-2,4-
diphenylcyclobutane-1-carboxylate (3n). White solid, 51% yield; m.
p. 158e159 ꢁC; 1H NMR (300 MHz, (acetone-d6)
3.98e4.22 (m,
3H), 4.29e4.64 (m, 3H), 7.16e7.55 (m, 10H); 13C NMR (75 MHz,
acetone-d6) 41.4, 41.7, 46.0, 46.2, 59.4, 59.7, 60.1, 60.4, 121.9, 124.6,
d
d
d
d
126.9, 127.0, 127.4, 127.6, 128.2, 128.3, 138.7, 138.9, 170.3, 172.0;
60.63, 60.92, 127.28, 127.33, 127.47, 127.55, 128.62, 128.65, 137.61,
HRMS (ESI) m/z: calcd for C20H17O4F3Hþ 379.1152; found 379.1155
137.69, 170.35, 176.49; HRMS (ESI) m/z: calcd for C20H17O4F3Hþ
(D
¼ 0.79 ppm).
379.1152; found 379.1155 (
D
¼ 0.79 ppm).
4 . 4 . 3 . 3 0 . 1 - C y c l o h e x y l
a
- 3 - h y d r o x y c a r b o n y l - 2 , 4 -
4.4.3.25. 1-(6-Acetamidonaphth-1-yl)
diphenylcyclobutane-1-carboxylate (3o). White solid; 37% yield;
m. p. 131e132 ꢁC; 1H NMR (500 MHz, MeOD)
2.18 (s, 3H), 4.08 (dd,
a
-3-hydroxycarbonyl-2,4-
diphenylcyclobutane-1-carboxylate (3r). White solid; 68% yield; m.
p. 152e153 ꢁC; 1H NMR (500 MHz, CDCl3)
d
0.83 (m, 1H), 1.09e1.25
d
(m, 5H), 1.38e1.49 (m, 2H), 1.59 (m, 2H), 3.90 (dd, J ¼ 10.6, 6.8 Hz,
J ¼ 10.5, 7.0 Hz, 1H), 4.49 (dd, J ¼ 10.5, 7.0 Hz, 1H), 4.65 (m, 2H), 6.32
(d, J ¼ 7.5 Hz, 1H), 6.96 (d, J ¼ 9.1 Hz, 1H), 7.35e7.23 (m, 3H), 7.
37e7.53 (m, 7H), 7.58 (d, J ¼ 7.0 Hz, 3H), 8.20 (m, 1H); 13C NMR
1H), 4.01 (dd, J ¼ 10.6, 7.8 Hz, 1H), 4.40 (m, 1H), 4.50e4.42 (m, 2H),
7.21e7.37 (m,11H); 13C NMR (125 MHz, CDCl3)
d 23.53, 23.60, 25.25,
30.92, 31.41, 41.45, 41.58, 46.25, 46.98, 72.90, 127.18, 127.47, 127.59,
(176 MHz, acetone-d6)
d
24.42, 42.59, 43.08, 47.72, 115.76, 117.38,
128.41, 128.50, 138.50, 138.52, 171.32, 176.77; HRMS (ESI) m/z calcd
121.02, 123.03, 124.33, 126.14, 126.74, 127.84, 128.28, 128.79, 129.20,
129.21, 129.67, 136.11, 138.50, 140.21, 140.48, 147.72, 169.29, 171.61;
HRMS (ESI) m/z calcd for C27H24O4Hþ: 480.1811, found 480.1809
for C24H27O4Hþ: 379.1909, found 379.1909 (
D
¼ 0 ppm).
4.4.3.31. 1-{3-[1-(3,6,9-Trioxadodecanyl)-1,2,3-triazol-4-yl]phenyl}
(D
¼ 0.42 ppm).
a-3-hydroxycarbonyl-2,4-diphenylcyclobutane-1-carboxylate (3s).
To a solution of 3f in THF (8 mL) and water (1.5 mL), was added
cupric sulfate pentahydrate (81 mg, 0.32 mmol), ascorbic acid
(57 mg, 0.32 mmol), and 1-azido-3,6,9-dodecane (60 mg,
0.30 mmol) and the reaction mixture was stirred at room temper-
ature overnight. Upon completion, the reaction mixture was
diluted with water (20 mL) and extracted thrice with dichloro-
methane (25 mL x 3). The crude mixture was purified by flash
chromatography on silica gel with 3.5% methanol in dichloro-
methane as the eluent to give 3s (180 mg, 94% yield) as a white
4.4.3.26. 1-(6-Acetamidonaphth-1-yl)
diphenylcyclobutane-1-carboxylate (3o-
g
-3-hydroxycarbonyl-2,4-
g). To a solution of -trux-
g
illic anhydride (100 mg, 0.36 mmol) and 6-acetamido-1-naphthol
(80 mg, 0.40 mmol) in THF (2.0 ml), was added diisoproplethyl-
amine (0.06 ml, 0.40 mmol) dropwise. The reaction mixture was
stirred at room temperature overnight. Upon completion, the re-
action mixture was diluted with water (2 mL) and extracted thrice
with dichloromethane (10 mL x 3). The crude mixture was purified
solid: m. p. 93e95 ꢁC; 1H NMR (500 MHz, acetone-d6)
d 1.07 (t,
by flash chromatography on silica gel to give 3o-
yield) as a white solid. m. p. decompose at 165 ꢁC; 1H NMR
(500 MHz, acetone-d6)
g (36 mg, 82%
J ¼ 7.02 Hz, 6H), 3.39 (q, J ¼ 7.02 Hz, 4H), 3.43e3.49 (m, 4H),
3.49e3.55 (m, 4H), 3.55e3.61 (m, 4H), 3.61e3.71 (m, 4H), 3.93 (t,
J ¼ 5.19 Hz, 4H), 4.06e4.19 (m, 2H), 4.27e4.36 (m, 2H), 4.54e4.72
(m, 8H), 6.28e6.43 (m, 2H), 6.96 (t, J ¼ 1.83 Hz, 2H), 7.21e7.32 (m,
4H), 7.33e7.44 (m, 6H), 7.44e7.54 (m, 8H), 7.57 (d, J ¼ 7.32 Hz, 4H),
7.69 (d, J ¼ 7.93 Hz, 2H), 8.22 (s, 2H); 13C NMR (126 MHz, acetone-
d
2.04 (s, 3H), 3.90 (t, J ¼ 9.8 Hz, 1H), 4.28 (t,
J ¼ 9.8 Hz, 1H), 4.52 (t, J ¼ 9.8 Hz, 1H), 4.74 (t, J ¼ 9.8 Hz, 1H), 6.30 (d,
J ¼ 7.6 Hz, 1H), 6.89 (d, J ¼ 9.0 Hz, 1H), 7.21 (dd, J ¼ 14.6, 7.0 Hz, 2H),
7.37e7.26 (m, 6H), 7.43e7.55 (m, 5H), 8.27 (s, 1H), d C
9.24 (s, 1H); 13
NMR (176 MHz, acetone-d6)
d23.46, 42.27, 44.65, 46.32, 46.43,
d6) d 15.6, 42.3, 42.9, 47.0, 47.6, 51.0, 66.8, 70.1, 70.6, 71.2, 71.3, 119.5,
121.7, 122.4, 123.4, 127.8, 128.3, 128.7, 129.0, 129.2, 129.5, 130.4,
114.76, 116.40, 120.03, 121.92, 123.36, 125.18, 125.80, 126.70, 126.72,
126.73, 127.43, 128.48, 128.59, 129.06, 135.17, 137.56, 138.60, 142.18,
146.71, 168.24, 170.20, 171.70; HRMS (ESI) m/z calcd for
133.7, 140.1, 140.2, 146.8, 152.1, 171.3, 173.1; HRMS (ESI) m/z calcu-
lated for C34H37N3O7Hþ: 600.2704, found 600.2705 (
D
¼ 0.13 ppm).
C
30H26NO5Hþ: 480.1811, found 480.1809 (
D
¼ 0.45 ppm).
4.4.3.32. 1-Naphthyl
a-3-hydroxycarbonyl-2,4-di(3-methoxy-4-
4.4.3.27. 1-(5-Ethynylnaphth-1-yl)
nylcyclobutane-1-carboxylate (3p). White solid; 44% yield; m. p.
220e221 ꢁC; 1H NMR (500 MHz, acetone-d6)
4.06 (s, 1H), 4.11 (dd,
a
-2,4-3-hydroxycarbonyl-diphe-
tosyloxy)cyclobutane-1-carboxylate (4a-i). To 1h (105 mg,
0.15 mmol) in dichloromethane (1.5 mL), was added EDC. HCl
(87 mg, 0.45 mmol), DMAP (22 mg, 0.16 mmol) and 1-naphthol
(26 mg, 0.16 mmol). The reaction mixture was stirred at room
temperature overnight. Upon completion, the reaction mixture was
diluted with water (2 mL) and extracted thrice with dichloro-
methane (5 mL x 3). The crude mixture was purified by flash
chromatography on silica gel to give 4a-i (36 mg, 29% yield) as a
d
J ¼ 10.7, 7.2 Hz, 1H), 4.53 (dd, J ¼ 10.7, 7.2 Hz, 1H), 4.64 (dd, J ¼ 10.7,
7.2 Hz,1H), 4.68 (dd, J ¼ 10.7, 7.2 Hz,1H), 6.48 (d, J ¼ 7.5 Hz,1H), 7.26
(m, 3H), 7.34 (t, J ¼ 7.6 Hz, 2H), 7.42 (m, 4H), 7.51 (d, J ¼ 7.5 Hz, 2H),
7.60 (d, J ¼ 7.2 Hz, 2H), 7.67 (d, J ¼ 6.8 Hz,1H), 8.11 (d, J ¼ 8.4 Hz,1H),
10.70 (s, 1H); 13C NMR (176 MHz, acetone-d6)
d 41.61, 42.11, 46.41,
46.77, 80.96, 83.68, 118.80, 119.77, 122.72, 123.45, 125.70, 126.49,
126.72, 126.92, 127.46, 127.85, 128.26, 128.27, 128.80, 131.41, 134.15,
139.18, 139.47, 147.18, 170.69, 172.12; HRMS (ESI) m/z: calcd for
white solid: m. p. 97e99 ꢁC; 1H NMR (500 MHz, acetone-d6)
d 2.43
(s, 3H), 2.48 (s, 3H), 3.55 (s, 3H), 3.58 (s, 3H), 4.13 (dd, J ¼ 10.5,
7.5 Hz, 1H), 4.58 (dd, J ¼ 10.5, 7.5 Hz, 1H), 4.68 (m, 2H), 6.56 (m, 1H)
7.22e7.09 (m, 5H), 7.28 (m, 1H), 7.34e7.54 (m, 8H), 7.69e7.74 (m,
C
30H22O4Hþ 447.1596; found 447.1591, (
D
¼ 1.1 ppm).

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
249
4H), 7.78 (d, J ¼ 8.3 Hz, 1H), 7.89e7.94 (m, 1H); 13C NMR (126 MHz,
7.49e7.43 (m, 7H), 7.73 (d, J ¼ 8.1 Hz, 1H), 7.87 (d, J ¼ 8.2 Hz, 1H),
8.30 (s, 1H), 8.49 (s, 1H); 13C NMR (101 MHz, acetone-d6)
40.97,
acetone-d6)
d
20.64, 20.69, 41.34, 41.83, 46.51, 46.53, 55.10, 55.26,
d
112.84, 113.36, 117.99, 119.59, 119.98, 121.24, 123.55, 123.93, 125.59,
125.79,126.30,126.42, 126.77, 127.80,128.40,128.48,129.57,129.58,
133.31, 133.53, 134.56, 137.31, 137.75, 139.61, 139.71, 145.41, 146.66,
151.59, 152.09, 170.40, 171.93; HRMS (ESI) m/z calculated for
41.40,47.03, 47.40, 115.06, 115.54, 118.05, 121.56, 125.25, 125.62,
126.14, 126.24, 126.93, 127.56, 128.99, 129.42, 129.99, 130.25, 134.47,
146.95, 156.45, 156.94, 170.85, 172.27; HRMS (ESI) m/z: calcd for
C
28H22O6Hþ, 455.1489, found, 455.1491 (
D
¼ 0.43 ppm).
In a manner similar to that described for the synthesis of 3a, the
following -truxillic acid mono esters bearing substituted phenyl
moieties, 4b, 4c and 4e ~4k, were synthesized from the corre-
sponding -truxillic acids and characterized.
C
44H38O12S2Hþ: 823.1867, found 823.1877 (
D
¼ 1.21 ppm).
a
4.4.3.33. 1-Naphthyl
a
-3-hydroxycarbonyl-2,4-di(3-methoxy-4-
hydroxy)cyclobutane-1-carboxylate (4a). To solution of 4a-i
a
a
(71 mg, 0.085 mmol), in MeOH (9.0 ml), was added 20% sodium
amalgam (283 mg) and the reaction mixture was stirred at room
temperature for 4 h. The reaction was quenched by 12 ml saturated
aqueous NaHCO3, and the mixture was extracted with AcOEt
(5 ꢂ 20 ml). The combined organic layers were washed with brine
(5 ml), dried over MgSO4, and concentrated in vacuo. The crude
mixture was purified using flash column chromatography on silica
gel with 2% methanol in dichloromethane as the eluent to give 4a
(18 mg, 40% yield) as a white solid: m. p. >230 ꢁC; 1H NMR
4.4.3.36. 1-Naphthyl
cyclobutane-1-carboxylate (4b). White solid; 35% yield; m. p.
184e185 ꢁC; 1H NMR (500 MHz, acetone-d6)
3.90 (s, 1H),
a-3-hydroxycarbonyl-2,4-di(2-methoxylphenyl)
d
4.16e4.21 (dd, J ¼ 8.00, 10.50 Hz, 1H), 4.47 (dd, J ¼ 8.00, 10.50 Hz,
1H), 4.86 (dd, J ¼ 8.00, 10.50 Hz, 1H), 4.86 (dd, J ¼ 8.00, 10.50 Hz,
1H), 4.93 (dd, J ¼ 8.00, 10.50 Hz, 1H), 6.47 (d, J ¼ 7.5 Hz, 1H),
7.01e7.04 (m, 2H), 7.07e7.12 (m, 2H), 7.28e7.31 (td, J ¼ 4.45, 24 Hz,
1H), 7.33e7.36 (t, J ¼ 7.70 Hz, 1H), 7.39e7.41 (m, 3H), 7.47e7.52 (m,
2H), 7.54e7.56 (d, J ¼ 7.00 Hz, 1H), 7.72e7.74 (d, J ¼ 8.00 Hz, 1H),
(500 MHz, acetone-d6)
d 3.77 (s, 3H), 3.89 (s, 3H), 3.90e3.94 (m,
1H), 4.28 (t, J ¼ 10.0 Hz, 1H), 4.51 (t, J ¼ 10.0 Hz, 1H), 4.73 (t,
J ¼ 10.0 Hz, 1H), 6.72 (d, J ¼ 7.5 Hz, 1H), 6.86 (m, 2H), 6.95e7.03 (m,
3H), 7.13 (m, 1H), 7.18 (m, 1H), 7.33e7.27 (m, 1H), 7.37 (t, J ¼ 7.5 Hz,
1H), 7.46 (t, J ¼ 7.5 Hz, 1H), 7.50 (s, 1H), 7.70 (s, 1H), 7.74 (d,
J ¼ 8.0 Hz, 1H), 7.87 (d, J ¼ 8.0 Hz, 1H); 13C NMR (125 MHz, acetone-
7.88e7.89 (d, J ¼ 8.00 Hz, 1H); 13C NMR (125 MHz, CDCl3)
d 37.33,
37.62, 44.32, 45.46, 55.12, 55.49, 110.32, 110.48, 117.59, 120.74,
125.30, 134.41, 146.62, 157.55, 157.78, 171.30, 177.82. HRMS (ESI) m/
z: calcd for C30H26O6Hþ, 483.1802; found, 483.1807 (
D
¼ 1.03 ppm).
d6)
d
41.91, 44.34, 46.89, 47.22, 55.39, 110.40, 112.65, 114.83, 115.03,
4.4.3.37. 1-Naphthyl
a-3-hydroxycarbonyl-2,4-di(2-nitrophenyl)
118.01, 119.19, 121.44, 122.13, 125.31, 125.67, 126.11, 126.27, 126.93,
127.57, 134.50, 144.80, 146.12, 146.97, 147.16, 170.48, 171.89; HRMS
(ESI) m/z calculated for C30H26O8Hþ: 515.1706, found 515.1717
cyclobutane-1-carboxylate (4c). White solid; 38% yield; m. p.
189e190 ꢁC; 1H NMR (500 MHz, DMSO‑d6)
4.37 (s, 1H). 4.68 (s,
1H), 4.89 (s, 1H), 5.04 (s, 1H), 7.79 (m, 15H), 13.23 (s, 1H); 13C NMR
(125 MHz, DMSO‑d6) 40.3, 40.7, 42.2, 43.0,118.1,121.3,124.5,125.7,
d
(D
¼ - 2.14 ppm).
d
In a manner similar to that described for the synthesis of 4a-i,
126.4, 128.0, 129.6, 132.2, 132.4, 133.0, 146.3, 148.9, 171.1, 173.6;
4d-i was synthesized from the corresponding
a
-truxillic acids 1j
HRMS (ESI) m/z: calcd for C28H20N2O8Hþ, 513.1292; found, 513.1302
and characterized.
(D
¼ 1.84 ppm).
4.4.3.34. 1-Naphthyl
a-3-hydroxycarbonyl-2,4-di(4-tert-butyldime-
4.4.3.38. (1R,2S)-2-phenylcyclohexyl a-3-hydroxycarbonyl-2,4-di(2-
thylsilylphenyl)- cyclobutane-1-carboxylate (4d-i). To 1j (200 mg,
0.36 mmol) in DCM (15 mL) was added DMAP (3 mg, 0.026 mmol)
and EDC (50 mg, 0.26 mmol), and the reaction was stirred at room
temperature for 15 min. 1-naphthol in DCM (5 mL) was added to
the solution by syringe pump at 2 ml/h and the reaction was stirred
at room temperature for overnight. The reaction mixture was then
concentrated in vacuo. The crude mixture was purified using flash
column chromatography on silica gel with 30% ethyl acetate in
hexanes as the eluent to give 4d-i (94 mg, 64%) as a white solid; m.
methoxylphenyl)-cyclobutane-1-carboxylate (4e). White solid; 48%
yield; m. p. 163e165 ꢁC; 1H NMR (500 MHz, CDCl3)
0.72 (m, 1H),
d
1.27 (m, 2H), 1.39e1.56 (m, 2H), 1.66 (m, 2H), 1.82 (m, 1H),
2.51e2.40 (m, 1H), 3.60 (s, 3H), 3.67 (s, 4H), 3.81 (dd, J ¼ 10.5,
6.2 Hz, 1H), 3.85e3.93 (m, 1H), 4.39e4.29 (m, 1H), 4.69 (dd, J ¼ 10.7,
6.6 Hz, 1H), 6.71 (d, J ¼ 8.1 Hz, 1H), 6.76 (d, J ¼ 8.1 Hz, 1H), 6.82 (t,
J ¼ 7.4 Hz, 1H), 6.87 (d, J ¼ 6.8 Hz, 1H), 6.94 (t, J ¼ 7.4 Hz, 1H), 7.09
(dd, J ¼ 16.6, 7.2 Hz, 3H), 7.13e7.24 (m, 5H). 13C NMR (125 MHz,
CDCl3)
120.2, 126.4, 127.5, 128.0, 143.2, 157.3, 171.6, 178.5. HRMS (ESI) m/z:
calcd for C32H35O6Hþ, 515.2428, found, 515.2434 (
¼ 1.2 ppm).
d 24.7, 25.8, 31.4, 37.2, 44.5, 45.8, 54.9, 55.2, 76.0, 110.0, 110.1,
p. 84e86 ꢁC; 1H NMR (500 MHz, CDCl3)
d0.19 (s, 6H), 0.23 (d,
J ¼ 4.1 Hz, 6H), 0.99 (s, 9H), 4.05 (dd, J ¼ 10.5, 7.5 Hz, 1H), 4.32 (dd,
J ¼ 10.5, 7.5 Hz, 1H), 4.52e4.63 (m, 2H), 6.42 (d, J ¼ 7.5 Hz, 1H), 6.88
(m, 4H), 7.31e7.23 (m, 5H), 7.36 (m, 3H), 7.44 (t, J ¼ 7.5 Hz, 1H), 7.65
(d, J ¼ 8.2 Hz, 1H), 7.80 (d, J ¼ 8.2 Hz, 1H). 13C NMR (125 MHz, CDCl3)
D
4.4.3.39. (1R,2S)-2-phenylcyclohexyl
a
-3-hydroxycarbonyl-2,4-di(2-
chlorophenyl)-cyclobutane-1-carboxyliate (4f). White solid; 30%
yield; m. p. 163e165 ꢁC; 1H NMR (500 MHz, CDCl3)
1.32e1.6 (m,
d
-4.41, ꢀ4.36, 18.22, 25.70, 40.99, 41.23, 46.87, 47.46, 117.72, 120.25,
d
120.32, 121.17, 125.20, 125.85, 126.28, 126.30, 126.59, 127.72, 128.67,
129.19,130.92,134.42,146.36,154.99,155.34,170.79, 175.87, 207.48;
HRMS (ESI) m/z: calcd for
683.3215 (
¼ 0.59 ppm).
4H), 1.80 (d, J ¼ 12.7 Hz, 1H), 1.85e2.02 (m, 2H), 2.15e2.33 (m, 1H),
2.64e2.81 (m, 1H), 3.52 (t, J ¼ 8.4 Hz, 1H), 3.50e3.77 (m, 2H), 4.48
(dd, J ¼ 16.4, 9.8 Hz, 1H), 4.75e4.89 (m, 1H), 5.16 (dt, J ¼ 17.2, 7.6 Hz,
1H), 6.78e6.96 (m, 2H), 6.99 (t, J ¼ 7.5 Hz, 1H), 7.02e7.11 (m, 3H),
C
40H50O6Si2Hþ, 683.3219, found,
D
7.13e7.31 (m, 7H). 13C NMR (125 MHz, CDCl3)
d 24.9, 25.9, 32.2, 41.5,
4.4.3.35. 1-Naphthyl
cyclobutane-1-carboxylate (4d). To solution of 4d-i (94 mg,
a
-3-hydroxycarbonyl-2,4-di(4-hydroxylphenyl)
49.7, 49.9, 126.2, 126.6, 127.5, 127.7, 128.0, 128.5, 129.7, 134.7, 143.0,
a
171.0. HRMS (ESI) m/z: calcd for C30H29Cl2O4Hþ, 523.1437, found,
0.14 mmol) in THF (10 mL) and acetic acid (0.5 mL) was added 1 M
TBAF (0.5 mL) and the mixture was stirred overnight at room
temperature. The reaction mixture was concentrated and purified
by flash chromatography 25% ethyl acetate and 0.1% acetic acid in
hexanes as eluent to give 4d (48 mg, 76%) as white solid; m. p.
523.1443 (
D
¼ 0.10 ppm).
4.4.3.40. (1R,2S)-2-phenylcyclohexyl
a
-3-hydroxycarbonyl-2,4-
(4g). White
1.40
di(2,6-dichlorophenyl)-cyclobutane-1-carboxylate
solid; 30% yield; m. p. 209e210 ꢁC; 1H NMR (700 MHz, CDCl3)
d
>230 ꢁC; 1H NMR (300 MHz, acetone-d6)
d
4.00 (dd, J ¼ 10.0, 7.5 Hz,
(dd, J ¼ 14.4, 11.3 Hz, 1H), 1.44e1.61 (m, 3H), 1.79 (d, J ¼ 13.2 Hz, 1H),
1.88 (d, J ¼ 10.1 Hz, 1H), 1.97 (d, J ¼ 13.4 Hz, 1H), 2.18e2.26 (m, 1H),
2.76 (td, J ¼ 12.1, 3.6 Hz, 1H), 4.45e4.52 (m, 1H), 4.70e4.78 (m, 1H),
1H), 4.40 (dd, J ¼ 10.0, 7.5 Hz, 1H), 4.62e4.51 (m, 2H), 6.61 (d,
J ¼ 7.5 Hz, 1H), 6.88 (q, J ¼ 7.6 Hz, 4H), 7.15 (d, J ¼ 8.5 Hz, 1H),

250
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
5.04 (dd, J ¼ 11.6, 9.3 Hz, 1H), 5.21 (td, J ¼ 10.4, 4.2 Hz, 1H), 5.27 (dd,
J ¼ 11.8, 8.1 Hz, 1H), 6.95 (t, J ¼ 8.0 Hz, 1H), 6.99 (t, J ¼ 8.0 Hz, 1H),
7.04 (d, J ¼ 8.0 Hz, 2H), 7.12 (d, J ¼ 7.5 Hz, 2H), 7.20 (t, J ¼ 7.1 Hz, 1H),
1H), 4.41 (dd, J ¼ 10.4, 8.0 Hz,1H), 4.65 (dd, J ¼ 10.4, 8.0 Hz,1H), 4.74
(dd, J ¼ 10.4, 8.0 Hz, 1H), 6.53 (d, J ¼ 7.5 Hz, 1H), 7.00e7.03 (m, 2H),
7.06 (t, J ¼ 7.5 Hz, 1H), 7.10 (d, J ¼ 8.4 Hz,1H), 7.28 (m, 1H), 7.35e7.44
(m, 3H), 7.49 (t, J ¼ 7.0 Hz, 2H), 7.61 (t, J ¼ 8.3 Hz, 1H), 7.85 (d,
J ¼ 8.3 Hz,1H), 8.85 (d, J ¼ 5.0,1H), 12.06 (s, 1H); 13C NMR (175 MHz,
7.30e7.25 (m, 5H). 13C NMR (175 MHz, CDCl3)
d 25.0, 25.9, 32.1, 42.3,
49.8, 76.7, 126.6, 127.8, 128.1, 128.5, 129.0, 134.1, 134.3, 143.3, 171.5,
177.5. HRMS (ESI) m/z: calcd for C30H27Cl4O4Hþ 591.0658, found
acetone-d6) d 36.64, 36.72, 44.87, 45.34, 55.89, 56.08, 111.12, 111.42,
591.0661 (
D
¼ 0.50 ppm).
118.73, 120.65, 121.09, 122.12, 127.25, 127.50, 127.53, 127.84, 127.95,
128.60, 129.11, 129.35, 130.24, 146.35, 148.51, 151.40, 157.71, 157.93,
171.64, 173.52; HRMS (ESI) m/z calcd for C29H25NO6Hþ: 484.1755,
4.4.3.41. (1R,2S)-2-phenylcyclohexyl
a
-3-hydroxycarbonyl-2,4-di(2-
bromophenyl)-cyclobutane-1-carboxylate (4h). White solid; 29%
found 484.1763 (
D
¼ 1.66 ppm).
yield; m. p.136e138 ꢁC; 1H NMR (500 MHz, acetone-d6)
d
1.33e1.44
(m, 1H), 1.44e1.54 (m, 2H), 1.62 (dd, J ¼ 12.9, 3.1 Hz, 1H), 1.76 (d,
J ¼ 12.9 Hz, 1H), 1.86 (dd, J ¼ 28.6, 11.6 Hz, 2H), 2.09 (s, 1H), 2.16 (d,
J ¼ 9.6 Hz,1H), 2.71 (td, J ¼ 12.2, 3.6 Hz,1H), 3.61 (dd, J ¼ 10.1, 7.5 Hz,
1H), 3.74 (dd, J ¼ 10.0, 7.6 Hz, 1H), 4.44e4.54 (m, 1H), 4.82 (dd,
J ¼ 9.6, 7.7 Hz, 1H), 5.10 (td, J ¼ 10.2, 4.0 Hz, 1H), 7.01 (dd, J ¼ 12.9,
7.1 Hz, 3H), 7.11 (dt, J ¼ 15.3, 7.4 Hz, 4H), 7.21 (d, J ¼ 3.9 Hz, 4H), 7.34
(d, J ¼ 7.7 Hz, 1H), 7.41 (d, J ¼ 7.8 Hz, 1H); 13C NMR (125 MHz,
4.4.3.46. 1-[4-(5,6,7,8-tetrahydronaphth-2-yl)thiazol-2-yl]
hydroxycarbonyl-2,4-diphenylcyclobutane-1-carboxamide
a
-3-
(6a).
White solid; 43% yield; m. p. >230 ꢁC; 1H NMR (500 MHz, acetone-
d6)
d
1.31 (s, 1H), 1.79 (m, 4H), 2.79e2.72 (m, 4H), 4.12 (dd, J ¼ 10.5,
7.4 Hz, 1H), 4.37 (dd, J ¼ 10.5, 6.8 Hz, 1H), 4.59 (dd, J ¼ 10.5, 7.4 Hz,
1H), 4.66 (dd, J ¼ 10.5, 6.8 Hz, 1H), 7.04 (d, J ¼ 8.5 Hz, 1H), 7.13 (t,
J ¼ 7.4 Hz, 1H), 7.26 (m, 4H), 7.36 (m, 2H), 7.43 (m, 2H), 7.48 (m, 2H),
acetone-d6)
127.2, 127.9, 128.4, 129.1, 139.0, 144.1, 171.9, 173.2; HRMS (ESI) m/z:
calcd for C30H29Br2O4Hþ, 611.0427, found, 611.0424 (
¼ 0.49 ppm).
d
25.4, 26.5, 32.9, 34.7, 43.4, 50.4, 76.9, 125.7, 125.9,
7.54 (m, 2H), 10.98 (s, 1H); 13C NMR (126 MHz, acetone-d6)
d23.07,
23.09, 40.64, 42.32, 46.45, 47.46, 106.20, 123.02, 126.45, 126.77,
126.80, 127.69, 127.79, 128.17, 128.23, 129.16, 132.05, 136.48, 136.87,
139.02, 139.66, 149.67, 157.33, 169.46, 172.35; HRMS (ESI) m/z calcd
D
4.4.3.42. (1R,2S)-2-phenylcyclohexyl
a
-3-hydroxycarbonyl-2,4-di(2-
for C31H28N2O3SHþ, 509.1893, found 509.1092 (
D
¼ 1.61 ppm).
nitrophenyl)-cyclobutane-1-carboxylate (4i). White solid; 36%
yield; m. p. 200e202 ꢁC; 1H NMR (500 MHz, acetone-d6)
4.4.3.47. 1-N-(Biphenyl-4-yl)
a-3-hydroxycarbonyl-2,4-
d
1.35e1.45 (m, 1H), 1.49 (dd, J ¼ 21.7, 11.8 Hz, 2H), 1.64 (dd, J ¼ 12.9,
diphenylcyclobutane-1-carboxamide (6b). White solid; 40% yield;
2.9 Hz, 1H), 1.77 (d, J ¼ 12.9 Hz, 1H), 1.86 (dd, J ¼ 25.5, 11.5 Hz, 2H),
1.96 (s, 1H), 2.70 (td, J ¼ 12.2, 3.4 Hz, 1H), 3.73 (dd, J ¼ 10.4, 6.0 Hz,
1H), 3.92 (t, J ¼ 9.6 Hz, 1H), 4.76 (t, J ¼ 9.4 Hz, 1H), 4.85 (dd, J ¼ 10.0,
6.0 Hz, 1H), 5.01e5.13 (m, 1H), 7.14 (d, J ¼ 7.6 Hz, 2H), 7.20 (d,
J ¼ 4.1 Hz, 4H), 7.34e7.25 (m, 2H), 7.37 (t, J ¼ 7.4 Hz, 2H), 7.52 (t,
J ¼ 7.5 Hz, 1H), 7.68 (d, J ¼ 7.8 Hz, 1H), 7.72 (d, J ¼ 8.0 Hz, 1H); 13C
m. p. > 230 ꢁC; 1H NMR (700 MHz, acetone-d6)
d
3.84 (t, J ¼ 10.4 Hz,
1H), 3.88 (t, J ¼ 10.4 Hz, 1H), 4.42 (t, J ¼ 10.4 Hz, 1H), 4.92 (t,
J ¼ 10.4 Hz, 1H), 7.14 (t, J ¼ 7.5 Hz, 1H), 7.24 (t, J ¼ 7.5 Hz, 3H), 7.30 (t,
J ¼ 7.3 Hz, 1H), 7.36 (t, J ¼ 7.7 Hz, 2H), 7.41 (t, J ¼ 7.7 Hz, 3H),
7.48e7.43 (m, 9H), 7.50 (s, 1H), 7.58 (d, J ¼ 7.0 Hz, 2H), 10.67 (s, 1H);
13C NMR (175 MHz, acetone-d6)
d40.66, 44.77, 45.65, 48.27, 119.66,
NMR (125 MHz, acetone-d6)
d
25.4, 26.5, 32.8, 34.4, 40.9, 50.5, 77.3,
119.76, 119.81, 126.30, 126.41, 126.60, 126.83, 127.87, 128.31, 128.74,
128.94, 135.70, 138.06, 138.46, 140.50, 143.17, 168.72, 172.04; HRMS
125.2, 125.5, 127.2, 128.5, 128.8, 129.1, 130.1, 130.2, 134.0, 144.0,
149.7, 151.0, 171.4; HRMS (ESI) m/z: calcd for C30H29N2O8Hþ,
(ESI) m/z calcd for
¼ 0.22 ppm).
C
30H25NO3Hþ: 448.1913, found 448.1914
545.1918, found, 545.1921 (
D
¼ 0.55 ppm).
(D
4.4.3.43. 9-Fluorenylmethyl -3-hydroxycarbonyl-2,4-di(2-
a
4.4.3.48. 1-[4-(5,6,7,8-tetrahydronaphth-2-yl)thiazol-2-yl]
hydroxycarbonyl-2,4-diphenylcyclobutane-1-carboxamide (6a-
To a solution of -truxillic anhydride [11] (100 mg, 0.36 mmol) and
g
-3-
methoxylphenyl)cyclobutane-1-carboxylate (4j). White solid; 28%
g).
yield; m. p. 160e162 ꢁC; 1H NMR (500 MHz, CDCl3)
d
3.76 (d,
g
J ¼ 11.6 Hz, 4H), 3.80 (s, 3H), 3.96 (dd, J ¼ 10.6, 8.0 Hz, 1H), 4.06 (dd,
J ¼ 10.4, 7.8 Hz, 1H), 4.17e4.10 (m, 2H), 4.66 (dd, J ¼ 10.5, 7.1 Hz, 1H),
4.72 (dd, J ¼ 10.4, 7.8 Hz, 1H), 6.82 (d, J ¼ 8.1 Hz, 1H), 6.86 (d,
J ¼ 8.1 Hz, 1H), 7.00 (q, J ¼ 7.1 Hz, 2H), 7.24 (t, J ¼ 7.2 Hz, 1H),
7.36e7.27 (m, 5H), 7.41 (t, J ¼ 6.9 Hz, 3H), 7.51 (d, J ¼ 7.5 Hz,1H), 7.77
6-acetamido-1-naphthol (80 mg, 0.40 mmol) in THF (2.0 ml), was
added diisoproplethylamine (DIPEA) (0.06 ml, 0.40 mmol) drop-
wise. The reaction mixture was stirred at room temperature over-
night. Upon completion, the reaction mixture was diluted with
water (2 mL) and extracted thrice with dichloromethane (10 mL x
3). The crude mixture was purified by flash chromatography on
(d, J ¼ 7.5 Hz, 2H). 13C NMR (125 MHz, CDCl3)
d 37.1, 37.7, 44.6, 46.9,
55.2, 55.3, 66.4, 110.1, 110.3, 120.0, 125.1, 127.1, 127.7, 141.2,
silica gel (CH2Cl2/MeOH, 95/5) to give 3o-
g (36 mg, 82% yield) as a
141.3144.0,144.4,157.5,157.6,172.6,178.6. HRMS (ESI) m/z: calcd for
white solid: m. p. 165 ꢁC (decomp.); 1H NMR (500 MHz, acetone-d6)
C
34H30O6Hþ, 535.2115; found, 535.2115 (
D
¼ 0.05 ppm).
d
2.04 (s, 3H), 3.90 (t, J ¼ 9.8 Hz, 1H), 4.28 (t, J ¼ 9.8 Hz, 1H), 4.52 (t,
J ¼ 9.8 Hz,1H), 4.74 (t, J ¼ 9.8 Hz, 1H), 6.30 (d, J ¼ 7.6 Hz, 1H), 6.89 (d,
J ¼ 9.0 Hz, 1H), 7.21 (dd, J ¼ 14.6, 7.0 Hz, 2H), 7.37e7.26 (m, 6H),
4.4.3.44. 9-Fluorenylmethyl
a-3-hydroxycarbonyl-2,4-di(2-
chlorophenyl)cyclobutane-1-carboxylate (4k). White solid; 55%
7.43e7.55 (m, 5H), 8.27 (s, 1H), d
9.24 (s, 1H); 13C NMR (176 MHz,
yield; m. p. 187e188 ꢁC; 1H NMR (500 MHz, acetone-d6),
d
4.01 (dd,
acetone-d6) d23.46, 42.27, 44.65, 46.32, 46.43, 114.76, 116.40,
J ¼ 9.7, 7.0 Hz, 1H), 4.10e4.20 (m, 1H), 4.23 (t, J ¼ 7.0 Hz, 1H), 4.38 (p,
J ¼ 10.6 Hz, 2H), 4.60e4.70 (m, 1H), 4.70e4.82 (m, 1H), 7.04e7.20
(m, 4H), 7.26 (dt, J ¼ 24.3, 8.7 Hz, 4H), 7.34 (d, J ¼ 6.2 Hz, 1H),
7.37e7.46 (m, 3H), 7.62 (t, J ¼ 7.5 Hz, 2H), 7.87 (dd, J ¼ 7.1, 3.2 Hz,
120.03, 121.92, 123.36, 125.18, 125.80, 126.70, 126.72, 126.73, 127.43,
128.48, 128.59, 129.06, 135.17, 137.56, 138.60, 142.18, 146.71, 168.24,
170.20, 171.70; HRMS (ESI) m/z calcd for C30H26NO5Hþ: 480.1811,
found 480.1809 (
D
¼ 0.45 ppm).
2H), 12.74 (s, 1H); 13C NMR (125 MHz, acetone-d6)
d 41.2, 41.6, 42.5,
46.2, 66.2, 120.2, 125.2, 133.5, 140.7, 143.4, 143.6, 172.1, 173.7. HRMS
4.4.4. Computational study
(ESI) m/z: calcd for C32H25Cl2O4Hþ, 543.1124, found, 543.1123
In order to optimize the activity of 3-analogs, the docking pro-
gram AutoDock 4.2 was used. Analogs were drawn in the Perkin
Elmer program ChemDraw and then had their internal energies
optimized by the program Avogadro [43]. The force field chosen for
this was the Merck Molecular Force Field (MMFF94) [44]. Docking
was performed with the default parameters based on the
(D
¼ 0.18 ppm).
4.4.3.45. Quinolin-5-yl
a
-3-hydroxycarbonyl-2,4-di(2-
methoxylphenyl)cyclobutane-1-carboxylate (4l). White solid; m. p.
>230 ꢁC; 1H NMR (500 MHz, acetone-d6)
3.95 (dd, J ¼ 10.4, 8.0 Hz,
d

S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
251
Lamarckian genetic algorithm of AutoDock 4.2. Analysis of pre-
dicted binding poses and their associated scores was performed in
the program UCSF Chimera [45]. Analogs with high similarity in
predicted binding mode to 3 were prioritized for synthesis. Analysis
of the selectivity of truxillic acid congeners for FABP isoforms
additionally employed the local search algorithm of AutoDock 4.2
[46,47].
Crystal structures (PDB ID: 5UR9 for FABP5/3; 5URA for FABP7/
3; 6AQ1 for FABP3 apoprotein) were processed by first adding polar
hydrogens to reflect a pH of 7.4 via AutoDock Tools. The Gasteiger
function was applied to assign partial atomic charges to a rigid
receptor. Additionally, all water molecules and ligands were
deleted from the binding site model. The portion of the binding site
chosen for sampling was based on determining a region that would
result in low RMSD to the crystal structure binding pose, or in the
case of FABP3 apoprotein, the region that produced results with the
canonical interaction as the top score. The AutoDock program
autogrid was then used to generate an energy potential map of
various atoms in the binding site. The grid files were used for
docking calculations.
4.4.5.5. CFA-induced paw edema and thermal hyperalgesia. All
procedures involving vertebrate animals were approved by the
Institutional Animal Care and Use Committee (IACUC) at Stony
Brook University. The experiments were conducted on male wild-
type C57BL/6 mice or FABP5-KO mice (bred on a C57BL/6 back-
ground). Mice were acclimated to the testing room, equipment, and
experimenter for at least two days before behavioral testing. Paw
edema was induced by injecting complete Freund's adjuvant (CFA)
(20 ml, in sterile saline) into the plantar surface of the right hind
paw using a 30-gauge needle attached to a tuberculin syringe
(Becton, Dickinson and Company). On day 5, thermal hyperalgesia
was measured using the Hargreaves plantar apparatus (Ugo Basile)
as previously reported [11,33]. Measurements were taken prior to
the administration of drug, and at 90 min post-administration. 3,
3a, 4a, and 3l-A were dissolved in 1:1:8 DMSO: Cremophor-
EL:saline. Compound 5f was dissolved in 1:2:7 DMSO:Solutol:sa-
line with gentle heating. Drugs were administered via the intra-
peritoneal route at a dose of 40 mg/kg in a volume of 10 ml/g, with
the exception of 3l-A which was administered at 20 mg/kg.
Acknowledgments
4.4.5. Biological evaluations
4.4.5.1. Protein purification. Recombinant human FABP3, FABP4,
FABP5, and FABP7 were expressed as N-terminal His-tagged pro-
teins using a pET-28a vector (Novagen, Madison, WI, USA). The
protein purification and delipidation was performed as previously
described [1].
This work was supported by a grant (DA042545) from the Na-
tional Institutes of Health. The authors thank Dr. Bela Ruzsicska,
Institute of Chemical Biology and Drug Discovery, and Robert
Rieger, Proteomics Center, of the Stony Brook University for mass
spectrometry analyses. The authors thank Mr. Glenn Yu and Ms.
Sarah Lee and Mr. Alby Joseph for their technical assistance for
docking studies as well as chemical synthesis. The authors also
4.4.5.2. Direct ligand binding assay. Measurement of direct binding
of the fluorescent probes to purified FABPs was performed as
described previously [11,48]. NBD-stearate affinity towards FABP3,
€
thank Dr. Gokhan Hotamisligil for graciously providing the FABP5-
KO mice.
FABP5, and FABP7 was determined in
a previous study
(Kd ¼ .018
m
M, 0.16 M, and 0.22 M, respectively) [11]. Incubation
m
m
Abbreviations
of NBD-stearate with a subset of tested compounds (3k, 3p, 3q, 4j
and 4k) was found to yield unexpectedly high non-specific fluo-
rescence in the absence of protein, therefore the alternative fluo-
FABP
FABP3
FABP5
FABP7
FAAH
NAE
fatty acid binding protein
heart fatty acid binding protein
epidermal fatty acid binding protein
brain fatty acid binding protein
fatty acid amide hydrolase
rophores DAUDA (for FABP3 and FABP5; Kd ¼ 2.04
mM and 2.56 mM,
respectively) and ANS (FABP7; Kd ¼ 1.01 M) were utilized.
m
4.4.5.3. Fluorescence displacement binding assay. Purified FABPs
(3 M) were incubated with fluorescent probe (500 nM) in 30 mM
Tris, 100 mM NaCl buffer (pH 7.5). Compounds to be tested were
then added to the wells (0.01e50 M) and the system was allowed
N-acyl ethanolamines
m
AEA
arachidonoyl ethanolamide;
cannabinoid receptor type 1
tetra-n-butylammonium fluoride;
N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide
hydrochloride;
CB1R
TBAF
EDC
m
to reach equilibrium by incubating in the dark at room temperature
for 20 min. Each independent assay included wells containing a
strong competitive ligand (arachidonic acid, 10 mM) as a positive
CFA
complete Freund's adjuvant
control for probe displacement. Loss of fluorescence intensity was
monitored with a F5 Filtermax Multi-Mode Microplate Reader
(Molecular Devices, Sunnyvale, CA, USA) using excitation and
emission wavelengths appropriate for each respective probe (NBD-
stearate ex./em. ¼ 465/535 nm, DAUDA ex./em. ¼ 345/535 nm, ANS
ex./em. ¼ 370/465 nm). Following background subtraction, the
fluorescence intensity values were normalized and fit to a one-site
binding analysis using the GraphPad Prism software (Prism version
7.0 for Mac OS, Graphpad Software Inc., La Jolla, CA, USA) to
determine the Ki of the tested compounds from the equation
Ki ¼ IC50/(1 þ ([Probe]/Kd).
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.ejmech.2018.04.050.
References
[1] M. Kaczocha, S. Vivieca, J. Sun, S.T. Glaser, D.G. Deutsch, Fatty acid-binding
proteins transport N-acylethanolamines to nuclear receptors and are targets
of endocannabinoid transport inhibitors, J. Biol. Chem. 287 (2012) 3415e3424.
[2] M.W. Elmes, M. Kaczocha, W.T. Berger, K. Leung, B.P. Ralph, L. Wang,
J.M. Sweeney, J.T. Miyauchi, S.E. Tsirka, I. Ojima, D.G. Deutsch, Fatty acid-
binding proteins (FABPs) are intracellular carriers for Delta9-
tetrahydrocannabinol (THC) and cannabidiol (CBD), J. Biol. Chem. 290 (2015)
8711e8721.
[3] D.G. Deutsch, A personal retrospective: elevating anandamide (AEA) by tar-
geting fatty acid amide hydrolase (FAAH) and the fatty acid binding proteins
(FABPs), Front. Pharmacology (Basel) 7 (2016) 370.
[4] M. Kaczocha, S.T. Glaser, D.G. Deutsch, Identification of intracellular carriers
for the endocannabinoid anandamide, Proc. Natl. Acad. Sci. Unit. States Am.
106 (2009) 6375e6380.
4.4.5.4. Dynamic light scattering. Small molecule colloidal aggre-
gation was measured by Protein Solutions DynaPro 99 dynamic
light-scattering instrument (Wyatt Technology, Santa Barbara, CA)
at 24 ꢁC. Compounds were dissolved to 10 mM in DMSO and diluted
with filtered PBS to a final concentration of 20 mM. Data were
analyzed with the Dynamics V5.25.44 program supplied by Wyatt
Technology.

252
S. Yan et al. / European Journal of Medicinal Chemistry 154 (2018) 233e252
[5] A.H. Lichtman, C.C. Shelton, T. Advani, B.F. Cravatt, Mice lacking fatty acid
(2008) 3755e3764.
amide hydrolase exhibit a cannabinoid receptor-mediated phenotypic hypo-
algesia, Pain 109 (2004) 319e327.
[28] Y.-M. Chi, M. Nakamura, X.-Y. Zhao, T. Yoshizawa, W.-M. Yan, F. Hashimoto,
J. Kinjo, T. Nohara, S. Sakurada, Anti-inflammatory activity of 4,4'-Dihydroxy-
a-truxillic acid, Biol. Pharm. Bull. 29 (2006) 489e493.
[29] Y.-M. Chi, M. Nakamura, X.-Y. Zhao, T. Yoshizawa, W.-M. Yan, F. Hashimoto,
J. Kinjo, T. Nohara, S. Sakurada, Antinociceptive activities of a-truxillic acid and
b-Truxinic acid derivatives, Biol. Pharm. Bull. 29 (2006) 580e584.
[30] M. Furuhashi, G.S. Hotamisligil, Fatty acid-binding proteins: role in metabolic
[6] M. Kaczocha, M.J. Rebecchi, B.P. Ralph, Y.H. Teng, W.T. Berger, W. Galbavy,
M.W. Elmes, S.T. Glaser, L. Wang, R.C. Rizzo, D.G. Deutsch, I. Ojima, Inhibition
of fatty acid binding proteins elevates brain anandamide levels and produces
analgesia, PLoS One 9 (2014) e94200.
[7] J.E. Schlosburg, S.G. Kinsey, A.H. Lichtman, Targeting fatty acid amide hydro-
lase (FAAH) to treat pain and inflammation, AAPS J. 11 (2009) 39e44.
[8] G.L. Li, H. Winter, R. Arends, G.W. Jay, V. Le, T. Young, J.P. Huggins, Assessment
of the pharmacology and tolerability of PF-04457845, an irreversible inhibitor
of fatty acid amide hydrolase-1, in healthy subjects, Br. J. Clin. Pharmacol. 73
(2012) 706e716.
[9] K. Otrubova, C. Ezzili, D.L. Boger, The discovery and development of inhibitors
of fatty acid amide hydrolase (FAAH), Bioorg. Med. Chem. Lett. 21 (2011)
4674e4685.
[10] Y. Owada, T. Yoshimoto, H. Kondo, Spatio-temporally differential expression
of genes for three members of fatty acid binding proteins in developing and
mature rat brains, J. Chem. Neuroanat. 12 (1996) 113e122.
diseases and potential as drug targets, Nat. Rev. Drug Discov.
7 (2008)
489e503.
[31] H. Lan, C.C. Cheng, T.J. Kowalski, L. Pang, L. Shan, C.C. Chuang, J. Jackson,
A. Rojas-Triana, L. Bober, L. Liu, J. Voigt, P. Orth, X. Yang, G.W. Shipps Jr.,
J.A. Hedrick, Small-molecule inhibitors of FABP4/5 ameliorate dyslipidemia
but not insulin resistance in mice with diet-induced obesity, J. Lipid Res. 52
(2011) 646e656.
[32] M. Kaczocha, S.T. Glaser, T. Maher, B. Clavin, J. Hamilton, J. O'Rourke,
M. Rebecchi, M. Puopolo, Y. Owada, P.K. Thanos, Fatty acid binding protein
deletion suppresses inflammatory pain through endocannabinoid/N-
acylethanolamine-dependent mechanisms, Mol. Pain 11 (2015) 52.
[33] K. Hargreaves, R. Dubner, F. Brown, C. Flores, J. Joris, A new and sensitive
method for measuring thermal nociception in cutaneous hyperalgesia, Pain 32
(1988) 77e88.
[11] W.T. Berger, B.P. Ralph, M. Kaczocha, J. Sun, T.E. Balius, R.C. Rizzo, S. Haj-
Dahmane, I. Ojima, D.G. Deutsch, Targeting fatty acid binding protein (FABP)
anandamide transporters
- a novel strategy for development of anti-
inflammatory and anti-nociceptive drugs, PLoS One 7 (2012) e50968.
[12] B. Binas, H. Danneberg, J. McWhir, L. Mullins, A.J. Clark, Requirement for the
heart-type fatty acid binding protein in cardiac fatty acid utilization, Faseb. J.
13 (1999) 805e812.
[13] M. Nakamura, Y.M. Chi, W.M. Yan, Y. Nakasugi, T. Yoshizawa, N. Irino,
F. Hashimoto, J. Kinjo, T. Nohara, S. Sakurada, Strong antinociceptive effect of
incarvillateine, a novel monoterpene alkaloid from Incarvillea sinensis, J. Nat.
Prod. 62 (1999) 1293e1294.
[14] Y.M. Chi, M. Nakamura, T. Yoshizawa, X.Y. Zhao, W.M. Yan, F. Hashimoto,
J. Kinjo, T. Nohara, S. Sakurada, Anti-inflammatory activities of a-truxillic acid
derivatives and their monomer components, Biol. Pharm. Bull. 28 (2005)
1776e1778.
[15] M. Nakamura, Y.M. Chi, W.M. Yan, A. Yonezawa, Y. Nakasugi, T. Yoshizawa,
F. Hashimoto, J. Kinjo, T. Nohara, S. Sakurada, Structure-antinociceptive ac-
[34] X. Hui, H. Li, Z. Zhou, K.S. Lam, Y. Xiao, D. Wu, K. Ding, Y. Wang,
P.M. Vanhoutte, A. Xu, Adipocyte fatty acid-binding protein modulates in-
flammatory responses in macrophages through
a positive feedback loop
involving c-Jun NH2-terminal kinases and activator protein-1, J. Biol. Chem.
285 (2010) 10273e10280.
[35] D. Bogdan, J. Falcone, M.P. Kanjiya, S.H. Park, G. Carbonetti, K. Studholme,
M. Gomez, Y. Lu, M.W. Elmes, N. Smietalo, S. Yan, I. Ojima, M. Puopolo,
M. Kaczocha, Fatty acid binding protein 5 controls microsomal prostaglandin
e synthase 1 (MPGES-1) induction during inflammation, J. Biol. Chem. 293
(2018) 5295e5306.
[36] X. Peng, K. Studholme, M.P. Kanjiya, J. Luk, D. Bogdan, M.W. Elmes,
G. Carbonetti, S. Tong, Y.H. Gary Teng, R.C. Rizzo, H. Li, D.G. Deutsch, I. Ojima,
M.J. Rebecchi, M. Puopolo, M. Kaczocha, Fatty-acid-binding protein inhibition
produces analgesic effects through peripheral and central mechanisms, Mol.
Pain 13 (2017), 1744806917697007.
tivity studies of incarvillateine,
a monoterpene alkaloid from Incarvillea
sinensis, Planta Med. 67 (2001) 114e117.
[37] T. Sterling, J.J. Irwin, ZINC 15eligand discovery for Eeveryone, J. Chem. Inf.
Model. 55 (2015) 2324e2337.
[38] O. Diffraction, P. CrysAlis, Oxford Diffraction Ltd, 2009. Yarnton, England.
[39] L.J. Farrugia, WinGX and ORTEP for windows: an update, J. Appl. Crystallogr.
45 (2012) 849e854.
[16] M.L. Wang, G. Yu, S.P. Yi, F.Y. Zhang, Z.T. Wang, B. Huang, R.B. Su, Y.X. Jia,
Z.H. Gong, Antinociceptive effects of incarvillateine, a monoterpene alkaloid
from Incarvillea sinensis, and possible involvement of the adenosine system,
Sci. Rep. 5 (2015) 16107.
[17] Y.M. Chi, M. Nakamura, T. Yoshizawa, X.Y. Zhao, W.M. Yan, F. Hashimoto,
J. Kinjo, T. Nohara, S. Sakurada, Pharmacological study on the novel anti-
nociceptive agent, a novel monoterpene alkaloid from Incarvillea sinensis,
Biol. Pharm. Bull. 28 (2005) 1989e1991.
[18] P.K. Thanos, B.H. Clavin, J. Hamilton, J.R. O'Rourke, T. Maher, C. Koumas,
E. Miao, J. Lankop, A. Elhage, S. Haj-Dahmane, D. Deutsch, M. Kaczocha, Ex-
amination of the addictive and behavioral properties of fatty acid-binding
protein inhibitor SBFI26, Front. Psychiatr. 7 (2016) 54.
[40] O.V. Dolomanov, L.J. Bourhis, R.J. Gildea, J.A. Howard, H. Puschmann, OLEX2: a
complete structure solution, refinement and analysis program, J. Appl. Crys-
tallogr. 42 (2009) 339e341.
[41] T. Gruene, H.W. Hahn, A.V. Luebben, F. Meilleur, G.M. Sheldrick, Refinement of
macromolecular structures against neutron data with SHELXL2013, J. Appl.
Crystallogr. 47 (2014) 462e466.
[42] S.X. Liu, H.Z. Jin, L. Shan, H.W. Zeng, B.Y. Chen, Q.Y. Sun, W.D. Zhang, Inhibitory
effect of 4, 40-dihydroxy-
a-truxillic acid derivatives on NO production in
[19] AutoDock 4.2, http://autodock.scripps.edu/.
lipopolysaccharide-induced RAW 264.7 macrophages and exploration of
structureeactivity relationships, Bioorg. Med. Chem. Lett 23 (2013)
2207e2211.
[20] H.-C. Hsu, S. Tong, Y. Zhou, M.W. Elmes, S. Yan, M. Kaczocha, D.G. Deutsch,
R.C. Rizzo, I. Ojima, H. Li, The antinociceptive agent SBFI-26 binds to anan-
damide transporters FABP5 and FABP7 at two different sites, Biochemistry 56
(2017) 3454e3462.
[21] M.P. Edwards, D.A. Price, Role of physicochemical properties and ligand lip-
ophilicity efficiency in addressing drug safety risks, Annu. Rep. Med. Chem. 45
(2010) 380e391.
[22] P.D. Leeson, B. Springthorpe, The influence of drug-like concepts on decision-
making in medicinal chemistry, Nat. Rev. Drug Discov. 6 (2007) 881.
[23] ChemDraw 15, http://www.cambridgesoft.com/news/details/?News¼191.
[24] M. Cohen, G. Schmidt, 383. Topochemistry. Part I. A survey, J. Chem. Soc
(1964) 1996e2000.
[43] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeersch, E. Zurek,
G.R. Hutchison, Avogadro: an advanced semantic chemical editor, visualiza-
tion, and analysis platform, J. Cheminf. 4 (2012) 17.
[44] T.A. Halgren, Merck molecular force field. I. Basis, form, scope, parameteri-
zation, and performance of MMFF94, J. Comput. Chem. 17 (1996) 490e519.
[45] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt,
E.C. Meng, T.E. Ferrin, UCSF Chimeraea visualization system for exploratory
research and analysis, J. Comput. Chem. 25 (2004) 1605e1612.
[46] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of
docking with
a new scoring function, efficient optimization, and multi-
[25] M. Ichikawa, M. Takahashi, S. Aoyagi, C. Kibayashi, Total synthesis of
(ꢀ)-incarvilline,(þ)-incarvine C, and (ꢀ)-incarvillateine, J. Am. Chem. Soc. 126
(2004) 16553e16558.
threading, J. Comput. Chem. 31 (2010) 455e461.
[47] G.M. Morris, R. Huey, A.J. Olson, Using AutoDock for ligand-receptor docking,
Chapter 8, Curr. Protoc. Bioinformatics (2008). Unit 8 14.
[26] ChemBio 3D Ultra 14.0, https://www.cambridgesoft.com/Ensemble_for_
Chemistry/details/Default.aspx?fid¼13&pid¼668.
[48] A.E. Thumser, C. Evans, A.F. Worrall, D.C. Wilton, Effect on ligand binding of
arginine mutations in recombinant rat liver fatty acid-binding protein, Bio-
chem. J. 297 (Pt 1) (1994) 103e107.
[27] S. Chuang, T. Velkov, J. Horne, C.J. Porter, M.J. Scanlon, Characterization of the
drug binding specificity of rat liver fatty acid binding protein, J. Med. Chem. 51