Structure-Activity Relationship Studies of Coumarin-like Diacid Derivatives as Human G Protein-Coupled Receptor-35 (hGPR35) Agonists and a Consequent New Design Principle

Article abstract of DOI:10.1021/acs.jmedchem.0c01624

A series of coumarin-like diacid derivatives were designed and synthesized as novel agonists of human G-protein-coupled receptor 35 (hGPR35). Active compounds were characterized to possess one acidic group on both sides of a fused tricyclic aromatic scaff

Full text of DOI:10.1021/acs.jmedchem.0c01624

pubs.acs.org/jmc
Article
Structure−Activity Relationship Studies of Coumarin-like Diacid
Derivatives as Human G Protein-Coupled Receptor-35 (hGPR35)
Agonists and a Consequent New Design Principle
Lai Wei,⊥ Tao Hou,⊥ Jiaqi Li, Xiuli Zhang, Han Zhou, Zhenyu Wang, Junxiang Cheng, Kaijing Xiang,
Jixia Wang, Yaopeng Zhao,* and Xinmiao Liang*
Cite This: J. Med. Chem. 2021, 64, 2634−2647
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A series of coumarin-like diacid derivatives were designed and synthesized as novel agonists of human G-protein-
coupled receptor 35 (hGPR35). Active compounds were characterized to possess one acidic group on both sides of a fused tricyclic
aromatic scaffold. Most of them functioned as full agonists selective to hGPR35 and exhibited excellent potency at low nanomolar
concentrations. Substitution on the middle ring of the scaffold could effectively regulate compound potency. Structure−activity
relationship studies and docking simulation indicated that compounds that carried two acidic groups with a proper special distance
and attached to a rigid aromatic scaffold would most likely show a potent agonistic activity on hGPR35. Following this principle, we
screened a list of known compounds and some were found to be potent GPR35 agonists, and compound 24 even had an EC50 of 8
nM. Particularly, a dietary supplement pyrroloquinoline quinone (PQQ) was identified as a potent agonist (EC50 = 71.4 nM). To
some extent, this principle provides a general strategy to design and recognize GPR35 agonists.
However, its weak potency in a high micromolar range
challenged its position as an endogenous ligand. Several other
ligands were also postulated to be the natural agonists for
GPR35, including 2-acyl lysophosphatidic acid,10 guanosine-
3′,5′-cyclic monophosphate,11 multiple tyrosine metabolites,12
and the mucosal chemokine CXCL17,13 but which is a true
natural agonist for GPR35 remains controversial.14 Therefore,
GPR35 has been considered to be an orphan receptor until now.
Despite these limitations, more and more synthetic agonists
have been identified, including zaprinast that has become the
most widely used reference agonist for GPR35.15
INTRODUCTION
■
G protein-coupled receptors (GPCRs) represent the largest
target of therapeutic agents, mostly because of their substantial
involvement in human pathophysiology and their pharmaco-
logical tractability.1 It has been reported that drugs targeting
GPCRs account for ∼27% of the global market share of
therapeutic drugs, with aggregated sales during 2011−2015 at
∼$890 billion.2 Given the history of GPCRs successfully being
drug targets, there is great passion, as well as a requirement, for
discovering lead compounds for orphan GPCRs, natural
agonists of which yet remain unidentified.3 The human orphan
G protein-coupled receptor-35 (GPR35) was first discovered in
1998 and expressed in the immune and gastrointestinal systems,
dorsal root ganglia, lungs, cerebellum, and brain.4−6 GPR35 is
implicated in a number of diseases such as pain, cancer, coronary
artery disease, and hypertension;5−9 thus, great efforts have been
made to identify compounds that can regulate GPR35.
Received: September 17, 2020
Published: February 25, 2021
Kynurenic acid (1), the tryptophan metabolite, was reported
to be the first endogenously expressed GPR35 ligand in 2006.6
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
© 2021 American Chemical Society
J. Med. Chem. 2021, 64, 2634−2647
2634

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 1. Selected GPR35 agonists with potencies (1−8) and GPR35 antagonist (9).
Figure 2. Conceptual development of the novel GPR35 agonists starting from “compound 30” via 6-bromo-8-formyl-7-hydroxy-2-oxo-2H-chromene-
3-carboxylic acid to evolve to the general structure of the novel coumarin-like dicarboxylic derivatives reported herein.
The GPR35 sequence contains numerous positively charged
amino acids that face the binding pocket.16 This is the possible
reason for what most GPR35 agonists comprise one or two
acidic groups (Figure 1), for example, YE-210 (7) and PSB-
13253 (6) individually having one carboxyl group;17,18
“compound 50” (4) with one tetrazole group;19 bufrolin (2),
cromolyn (3), and pamoic acid (8) individually containing two
carboxyl groups.20,21 As in Figure 1, ML145 (9), a known potent
GPR35 antagonist, has a carboxyl group also.22 Several docking
models also confirmed a significant interaction between the
acidic group and some amino acid residues on GPR35.20
Although a number of compounds have been screened to look
for GPR35 agonists and thus a few detailed structure−activity
relationship (SAR) studies were generated, these works and
related drug designs were relatively random.17−19 When
analyzing structure characterization of active compounds, we
noticed an interesting phenomenon: bearing two carboxyl
groups on aromatic moieties, bufrolin, cromolyn, and pamoic
acid exhibited much better agonistic activity than their partial
analogues kynurenic acid, chromocarb, and 3-hydroxy-2-
naphthoic acid (Figure S1), respectively. Although such a
phenomenon is less reported, it inspired us to look into more
diacid compounds to explore whether the diacid structure is
more advantageous to GPR35 agonists. We previously reported
a series of coumarin derivatives that were able to act as powerful
GPR35 agonists. The coumarin scaffold easily derivatives and is
usually seen in drug designs.23,24 Thus, we designed and
synthesized a series of novel diacid compounds based on
coumarin-like scaffolds and studied their SAR and binding
interaction with GPR35 through docking simulation, which lead
to an empirical principle for a GPR35 agonist design. More
significantly, we found, for the first time, that PQQ
(Pyrroloquinoline quinone) is a potent GPR35 agonist based
on this principle.
2635
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Scheme 1. Syntheses of Coumarin 3,9-Dicarboxylic Acid Derivatives 13a−13l and Coumarin 3-Carboxylic Acid Derivatives 15a−
a
15j
a
Reagents and conditions: (a) Br2, AcOH, rt, 1 h. (b) Diethyl malonate, Piperidine, 80 °C, 6 h. (c) i: HMTA, TFA, 80 °C, overnight; ii: HCl (2
M), 80 °C, 1 h. (d) Diethyl malonate, Piperidine, 80 °C, 6 h. (e) NaOH, 1 h, rt. (g, h) i: arylboronic acids, PdCl2(PPh)2, K2CO3, dioxane/H2O =
5:1, 80 °C, overnight; ii: HCl (2 M), 80 °C, 2 h.
a
Scheme 2. Syntheses of Coumarin 3-Tetrazyl-9-carboxylic Derivatives 13m and 13n
a
Reagents and conditions: (a) CH3COONH4, H2O, rt, 6 h. (b) HCl, H2O, 75 °C, 2 h. (c) MOMCl, Et3N, DCM, rt, 3 h. (d) NaN3, AlCl3, THF, 80
°C, overnight. (e) TFA, 80 °C, overnight. (f) HCl (2 M), 80 °C, 1 h. (g) Meldrum’s acid, CH3COONH4, EtOH, rt, 2 h. (h) malononitrile,
CH3COONH4, rt, 4 h. (i) HCl (2 M), H2O, 75 °C, 2 h.
synthesized (Scheme 1). The coumarin skeletons 10a and 10b
were created using Knoevenagel condensation of diethyl
malonate and 9a and 9b, respectively. An 8-aldehyde group
was introduced to construct compounds 11a and 11b through
Duff reaction by refluxing 10 with urotropin in trifluoroacetic
acid. Another Knoevenagel condensation of 11 and diethyl
malonate was carried out with piperidine as a base to produce
12a and 12b. By a hydrolysis reaction, dicarboxylic products 13a
and 13b were obtained with excellent yields. Further derivatives
13c−13l could be prepared by Suzuki−Miyaura cross-coupling
of 12b with suitable aromatic boronic acids. As coumarin-like
mono-acid control compounds, 15a−15k could be synthesized
by similar cross-coupling reactions using 14 as a common
intermediate.
RESULTS AND DISCUSSION
■
Structure Consideration. Our previous work indicated
that the acidic group on the 3-position of coumarin derivatives is
crucial for compounds to maintain their GPR35 activity. In
addition, the 7-hydroxy and 8-nitro groups also played
important roles in the high agonistic activity for several
compounds, such as “compound 30”, which is possibly related
to the cyclic structure formed by an intramolecular hydrogen
bond between the hydroxyl and nitro. Due to this consideration,
we designed a series of coumarin-like tricyclic derivatives and
introduced an acidic group on the additional ring. Figure 2
presents some representative structures.
Chemistry. A series of coumarin-like 3,9-dicarboxylic acids
with various aromatic substituents at the 6-position were first
2636
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Considering that the tetrazyl group was often used as an
isostere of carboxyl, we also tried to conduct such a replacement
on compounds 13a and 13b. As shown in Scheme 2, 3-cyano-
coumarin 16 could be prepared in a one-pot method as reported
in literature.25 Via a [3+2] cycloaddition reaction of nitriles with
sodium azide catalyzed by AlCl3 in refluxing THF, intermediate
17 directly became the 3-(1H-tetrazol-5-yl)-2H-chromen-2-one
derivative 18 along with losing an MOM protecting group. We
also performed this reaction with a naked phenolic hydroxyl
group but obtained no product. Then, the aldehyde group was
introduced at the 8-position by Duff reaction to create
compound 19. Subsequent products 13m and 13n could be
obtained through Knoevenagel condensation with Meldrum’s
acid and malononitrile, respectively.26 However, preparation of
compound 13o was beyond the conventional [3+2] cyclo-
addition reaction of nitriles with sodium azide.
Pharmacological Evaluation. Dynamic mass redistribu-
tion (DMR) detected by a label-free resonant waveguide grating
biosensor was applied to profile compound activity at GPR35
endogenously expressed in human colorectal adenocarcinoma
cell line HT-29.27 In this work, zaprinast was used as a full
GPR35 agonist and tool molecule in DMR activation and
desensitization experiments. It triggered a robust DMR signal in
HT-29 cells with an EC50 of 0.34 0.03 μM. All coumarin-like
diacid derivatives not only gave rise to concentration-dependent
DMR signals in HT-29 but also, after 1 h incubation with the
cells, desensitized DMR responses induced by 1 μM zaprinast
that was added later (Tables 1−3). For all derivatives, the
potency to trigger DMR was found to be almost equivalent to
that to desensitize the zaprinast response, suggesting that these
compounds acted as GPR35 agonists.
A DMR antagonist assay using the known GPR35 antagonist
compound 9 (ML-145) confirmed that compound 9 concen-
tration dependently and completely blocked the DMR
responses generated by the diacid derivatives that were added
at concentrations equal to individual EC80−EC100. Together
with the desensitization assay results, it suggested that the diacid
compounds generated DMR responses in the HT-29 cells
specifically through the activation of GPR35. Therefore, these
compounds were GPR35 agonists.
Coumarin 3,10-dicarboxylic acid derivative 20 could be
readily prepared in an excellent yield by Knoevenagel
condensation of 2,3-dihydroxynaphthalene-1,4-dicarbaldehyde
with Meldrum’s acid in water with a catalytic amount of
ammonium acetate (Scheme 3).
Scheme 3. Synthesis of Coumarin 3,10-Dicarboxylic Acid
a
Derivative 20
Structure−Activity Relationship Analysis. Based on the
molecular scaffold (Figure S6), the coumarin-like diacid
compounds were roughly classified into a U-shaped type
(Table 1) and a linear type (Table 3), and the coumarin
mono-acid compounds 15a−15j (Table 2) were used as control
compounds.
Compound 13a, the simplest diacid derivative, showed a
moderate agonist property (EC50 0.31 μM). When a 6-position
bromine was introduced, the resulting compound 13b showed a
7-fold increase in potency (EC50 0.045 μM). This result was in
accord with an early report that such an introduction to 3-
carboxylic-coumarin could improve the agonistic activity.19 To
expand molecular diversity, the bromine was replaced with
various substituted phenyl groups. The 6-phenyl derivative 13c
(EC50 0.048 μM) showed almost the same activity to 13b, but
the potency of 13j with a bigger substituent 6-(para-isopropyl)-
phenyl increased 3.5-fold (EC50 0.013 μM) compared to that of
13b, suggesting a possible lipophilic pocket opposite to the 6-
position of the coumarin scaffold.18 The 6-aryl derivatives with a
a
Reagents and conditions: (a) CH3COONH4, H2O, rt, 6 h.
Besides the U-shaped molecular structure presented in
compounds 13 and 20, linear coumarin 3,7-dicarboxylic
derivatives 23a−23f were also designed and synthesized. As
shown in Scheme 4, salicylaldehyde derivatives 21a−21c were
treated with piperidine as a base in diethyl malonate at 80 °C to
construct intermediates 22a−22c. Target compounds 23a−23c
bearing two symmetric carboxyl groups were yielded by
hydrolysis of the ethyl esters with NaOH at room temperature.
Biphenyl analogues 23d−23f could be prepared from 22a by
Suzuki−Miyaura cross-coupling with suitable aromatic boronic
acids as mentioned above.
a
Scheme 4. Syntheses of [3,2-g]Chromene Derivatives 23a−23f
a
Reagents and conditions: (a) Piperidine, 80 °C, 6 h. (b) NaOH, H2O, 1 h, rt. (c) i: arylboronic acids, PdCl2(PPh)2, K2CO3, dioxane/H2O = 5:1,
80 °C, overnight; ii: HCl (2 M), 80 °C, 2 h.
2637
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 1. Potencies of the U-Shaped Coumarin-like Tricyclic Derivatives 13a−13n and 20 in DMR Assays on HT-29
a
b
c
EC50 to trigger DMR. IC50 to desensitize cells upon repeated stimulation with 1 μM zaprinast. IC50 of the known GPR35 antagonist 9 to block
the agonist-induced DMR.
variety of substituents at the ortho-, para-, or meta-position of
phenyl were further investigated. Most compounds among
13d−13l exhibited a considerable increase in potency compared
to the unsubstituted 13c, which might be due to a better filling of
the lipophilic pocket by the substituted phenyls. The only
exception was 13d, and the introduction of ortho-methoxyl
halved its potency (EC50 0.083 μM). Because the para-methoxy
and ortho-fluoro substitutions appeared more favorable (13f,
EC50 0.020 μM; 13g, EC50 0.022 μM), we would like to
investigate whether a combination of the two substitutions could
further improve potency. However, the resulting compound 13l
bearing o-fluoro and p-methoxy showed a significantly reduced
potency (EC50 0.092 μM) compared with its monosubstituted
counterparts. More substituents led to an undesired effect
probably due to the reason that those groups were too big to
properly fit in the lipophilic pocket of GPR35. The replacement
of phenyl (13c) with thiophene (13k) brought in no apparent
effect to the potency (13k, EC50 0.037 μM). Noticeably, with a
different scaffold arrangement, analogue 20 also showed a
competitive GPR35 agonist potency (EC50 0.017 μM).
We previously demonstrated that tetrazolyl was a more
effective active group than carboxyl in similar molecules in terms
of GPR35 potency,19 based on which compound 13m with a
tetrazolyl on the 3-position of scaffold 13 was designed. With the
replacement of the carboxyl with tetrazolyl, compound 13m
showed a 2-fold increase in potency (EC50 0.018 μM) compared
to compound 13b. Compound 13n with a cyano unit in the 9-
position showed a sharp reduction (EC50 2.9 μM). It indicated
that the presence of two acidic groups in the molecule was
extremely important to maintain a high activity.
2638
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
7-Hydroxy-2-oxo-2H-chromene-3-carboxylic acid derivatives
15a−15j were used as the mono-acid control compounds to
further examine the effect of two acidic substituents on the
scaffold 13. It was found that all derivatives presented a relatively
poor potency (EC50 > 1 μM, Table 2), about 100-fold less active
than the corresponding compound 13. But when comparing
results in Tables 1 and 2, we noticed an intriguing phenomenon
that the results seemed to have the same trend of activity among
the variations with substituents on the 6-position. For instance,
for methoxy substituents, the activity pattern of the compounds
13d−13f and 15c−15e (Figure 3A) was the same: meta ≈ para
> ortho. So was the activity pattern of the fluoro compounds
13g−13i and 15f−15h (Figure S3B). The above results were
noteworthy and unambiguously indicated that the two acidic
substituents on the coumarin-like scaffold played an important
role in improving GPR35 agonist potency.
Table 2. Potency of the Coumarin Mono-Acid Derivatives
15a−15j in DMR Assays on HT-29
Furthermore, a series of linear tricyclic compounds 23a−23f
were investigated. 3,7-Dicarboxyl derivative 23 could be
considered as an isomer of the 3,9-dicarboxyl scaffold 13.
Results in Table 3 showed that 23a−23f also had GPR35 agonist
activities at a submicromolar level. For compounds 23a−23c,
the bromine substitution (23a, EC50 0.025 μM) maintained
much better potency than methyl (23b, EC50 0.140 μM) and
hydroxyl (23c, EC50 0.390 μM) on the 10-position of the
scaffold (Figure 4A). The introduction of ortho-, meta-, and
para-methoxyl substituents to phenyls also endowed the
corresponding compounds with excellent agonistic potency,
especially the para-methoxyl. The rank order of potency was as
follows: 23f (EC50 0.017 μM) > 23d (EC50 0.065 μM) > 23e
(EC50 0.092 μM) (Figure 4B). These results indicated that the
10-position on scaffold 23 had a good tolerance to larger
lipophilic substituents.
a
b
EC50 to trigger DMR. IC50 to desensitize cells upon repeated
c
stimulation with 1 μM zaprinast. IC50 of the known GPR35
antagonist 9 to block the agonist-induced DMR.
For the three types of diacid compounds 13, 20, and 23, we
believed that the spatial distance between two acidic groups
should be a key structural parameter as it was related to whether
Table 3. Potency of the Linear Coumarin-like Tricyclic Derivatives 23a−23f in DMR Assays on HT-29
a
b
c
EC50 to trigger DMR. IC50 to desensitize cells upon repeated stimulation with 1 μM zaprinast. IC50 of the known GPR35 antagonist 9 to block
the agonist-induced DMR.
2639
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 3. (A, B) Amplitude comparisons of the DMR induced by coumarin-like 3,9-dicarboxyl derivatives 13d−13i and 3-carboxylic acid coumarin
derivatives 15c−15h as a function of concentrations. The data represented mean sd from two independent measurements, each with four replicates
(n = 8).
Figure 4. (A, B) Amplitudes of the DMR induced by coumarin-like 3,7-dicarboxyl derivatives 23a−23f as a function of concentrations. The data
represented mean sd from two independent measurements, each with four replicates (n = 8).
the two acidic groups could synergistically interact with cationic
residues of the receptor. The spatial distance between the two
carboxyl carbons of 13b was 8.568 Å, 20 was 9.539 Å, and 23a
was 9.382 Å according to ChemBio 3D measurement. However,
the distance difference did not seem to to result in a distinct
division in potency among the compounds. We speculated an
appropriate distance ranging from 8.5 to 10 Å.
The above results indicated that the introduction of two acidic
groups in coumarin-like tricyclic scaffolds generally resulted in
good GPR35 agonist activity. Therefore, we speculated a general
phenomenon with caution that, carrying two acidic groups (such
as carboxyl or tetrazolyl) on a rigid aromatic scaffold, compound
would produce a GPR35 agonistic activity. The coumarin-like
tricyclic structure provided a rigid scaffold to maintain a
relatively favorable spatial distance between the acidic groups.
Other rigid scaffolds may also work for the GPR35 activity.
Although the electronic effect of different scaffolds might also
have influence on compound activity, the scaffold geometry
should be more fundamental. For instance, bufrolin is a potent
GPR35 agonist and has a phenanthroline-like tricyclic structure,
geometrically similar to compound 13. Following these
empirical principles, we tried to search and prepare more
compounds with those structural characters and test their
GPR35 activity.
10-Methyl-4,6-dioxo-1,4,6,9-tetrahydropyrido[3,2-g]-
quinoline-2,8-dicarboxylic acid 24 and its analogue 25 were
reported as potential antiallergy agents in the 1970s.28 Their
synthetic procedures are the same as described in literature
(Scheme 5).28,34 Compounds 24 and 25 have a quinoline-like
tricyclic scaffold, and the spatial distance between two carboxyl
groups is about 9.356 Å, and thus, they meet the above principle.
Indeed, the DMR assays showed that both were GPR35 agonists
(Table 4), especially compound 24 that exhibited rather
outstanding agonistic activity with an EC50 of 0.008 μM.
Compound 25 is the methylation product of 24. Its GPR35
agonistic potency decreased about 12-fold, but it was still
desirable (EC50 0.1 μM) (Figure S2). Compound 26 with 4,6-
dichloro substituents is a novel derivative of compound 24 and
exhibited an increased GPR35 agonistic activity (EC50 0.054
μM) compared to 25. It indicated that hydrogen bond receptors
at 4- and 6-positions of the quinoline-like tricyclic scaffold may
be more favorable to activity.
Dithieno[3,2-b:2′,3′-d]thiophene-2,6-dicarboxylic acid 27
was another diacid compound bearing a much different fused
2640
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
a
Scheme 5. Syntheses of 24−26
a
Reagents and conditions: (a) MeOH, 0 °C to rt, overnight. (b) Diphenyl ether, 260 °C, 2 h. (c) NaOH (2 M), MeOH, rt, 1 h. (d) MeI, K2CO3,
DMF, 60 °C, 12 h. (e) POCl3, 90 °C.
Table 4. Potency of the Predicted Compounds 24−27 and PQQ in DMR assays on HT-29
a
b
c
compd
EC50 (μM)
desensitization IC50 (μM)
antagonist IC50 (μM)
24
25
26
27
0.008 0.001
0.100 0.015
0.054 0.006
0.043 0.003
0.071 0.010
0.002 0.002
0.065 0.007
0.015 0.002
0.020 0.002
0.030 0.009
0.789 0.123
0.282 0.033
2.529 0.847
0.494 0.062
0.582 0.087
PQQ
a
b
c
EC50 to trigger DMR. IC50 to desensitize cells upon repeated stimulation with 1 μM zaprinast. IC50 of the known GPR35 antagonist 9 to block
the agonist-induced DMR.
Figure 5. (A) Structure of PQQ. (B) Real-time kinetic responses of PQQ at different concentrations in HT-29 cells. (C) DMR amplitudes of PQQ, the
concentration-dependent desensitization of 400 nM zaprinast DMR by PQQ and the concentration-dependent inhibition of the DMR of 1 μM PQQ
by ML145. The data represented mean sd from two independent measurements, each with four replicates (n = 8).
tricyclic scaffold and a spatial distance of 8.679 Å between two
carboxyl groups. It was initially reported as an intermediate
during the preparation of optoelectronic materials.29 Based on
our principle, compound 27 effectively activated the GPR35
receptor with an EC50 of 0.043 μM (Figure S2).
By searching natural compound libraries, we found another
example meeting the principle. Pyrroloquinoline quinone
(PQQ), an oxidoreductase co-factor widely present in
methylotrophic bacteria,30 is a natural product of great interest.
It not only takes part in the redox reaction in vivo but also has
some special biological activities and physiological functions,
such as regulation of the nervous system, improvement of
immune function, and inhibition of tumor growth and
metastasis.31 PQQ has been used as a dietary supplement to
reduce the ischemia−reperfusion injury caused by heart attack
or stroke.32,33 PQQ has a rigid pyrrolo[2,3-f ]quinoline scaffold
2641
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
and three carboxyl groups, two of which are located at the ends
of the scaffold with a distance of 8.959 Å, and the third one is at
the middle. As expected, the DMR assays showed that PQQ had
a nanomolar GPR35 agonistic activity (EC50 0.071 μM) and
could be antagonized by the specific antagonist ML-145 (Figure
5). To our knowledge, this should be the first time that PQQ was
discovered to activate GPR35; the current pharmacological
investigations of PQQ are controversial, and we hope our
finding of its agonistic activity on GPR35 may shed lights on
them.
Receptor Selectivity. To assess the receptor selectivity of
these novel GPR35 agonists, some compounds were selected for
further DMR assays. In CHO-K1 host cells, the selected
compounds showed no signals at all (Figure S3). However, on
CHO-K1-hGPR35 cells, all selected compounds exhibited
outstanding agonist activity, especially compounds 20 and 24
with EC50 values of 0.312 and 1.08 nM, respectively (Table 5).
In this work, we included acid groups, carboxyl and tetrazyl,
and the fused tricyclic aromatic scaffolds in the compound
design. Despite the limitation of experiments and simulations,
we still could propose that a combination of two acidic groups
and a suitable rigid scaffold would allow compounds to activate
GPR35, and this proposal might serve as a principle for future
GPR35 agonist designs.
CONCLUSIONS
■
In conclusion, a series of coumarin-like diacid derivatives with
U-shaped and linear fused tricyclic structures were synthesized
and evaluated to be agonists of an orphan receptor GPR35. Most
compounds behaved as selective and full GPR35 agonists and
exhibited excellent potency at low nanomolar concentrations.
Compounds carrying two acid groups, such as carboxyl or
tetrazyl, had significantly increased potency on GPR35
compared to mono acid analogues. Combining information of
some known agonists bearing different scaffolds, we believe that
such an increase is associated with the fact that the bonding
pocket of GPR35 is rich in cation residues to create a favorable
condition to form multiple electrostatic interactions between the
acid groups of ligand and the receptor. Based on the structural
features, more known compounds, such as 24−27 and PQQ,
were identified as potent GPR35 agonists. Particularly the
natural product PQQ, it is commercially used as a dietary
supplement to reduce the ischemia−reperfusion injury caused
by heart attack or stroke, and here, it was found to have powerful
agonistic activity on GPR35 for the first time, which would
provide more theoretical basis for its controversial pharmaco-
logical research. The docking simulation of some representative
agonists further rationalized the advantages of those diacid
compounds bearing a tricyclic aromatic scaffold. These results
promoted us to propose that suitable combination of two acidic
groups and a rigid scaffold would possibly endow the compound
with a potent agonistic activity on GPR35. We believe, to some
extent, this principle will provide a general method for designing
and recognize GPR35 agonists.
Table 5. Potency of the Selected Compounds on CHO-K1-
hGPR35 Cells in DMR Assays
a
b
compd
EC50 (nM)
desensitization IC50 (nM)
zaprinast
13f
13m
20
23a
23f
24
62.83 1.81
6.93 0.29
25.92 1.11
0.31 0.02
12.72 0.49
62.83 1.81
1.08 0.16
8.13 0.70
48.15 1.62
2.45 0.12
15.90 0.71
0.47 0.05
9.31 0.25
48.15 1.62
0.74 0.06
6.11 0.54
3.52 0.30
25
26
5.08 0.44
a
b
EC50 to trigger DMR. IC50 to desensitize cells upon repeated
stimulation with 400 nM zaprinast.
In other cell lines separately expressing M3, D2, μ-opioid, FFA4,
and H1 receptors, no apparent signals were observed when
compounds were added at a concentration of 1 μM (Figure S4),
suggesting no activity on these receptors. Therefore, these
selected fused tricyclic diacid compounds had a selective
agonistic activity on GPR35.
EXPERIMENTAL SECTION
■
General. All experimental reagents and solvents were obtained from
various providers and used without any additional purification or drying
except for tetrahydrofuran, which was distilled over calcium. The purity
of all coumarin-like compounds was ≥95%. The reactions were
monitored by thin-layer chromatography (TLC). If necessary, the
products were purified with column chromatography. NMR data were
collected on a Bruker Ascend 600 MHz NMR spectrometer at 600
MHz (1H) or 151 MHz (13C). The chemical shifts were reported in
parts per million (ppm) relative to the deuterated solvent DMSO-d6;
that is, δ 1H, 2.49 ppm; 13C, 39.7 ppm. High-resolution mass spectral
(HRMS) analyses were performed on an Agilent 1290 Infinity LC
instrument (Agilent, USA) coupled to an Agilent 6540 series QTOF-
MS (Agilent, USA) equipped with an ESI source, a diode-array detector
(DAD), an automatic sample injector, a degasser, and a column
thermostat.
Molecular Docking Studies. To further investigate the
binding mode of the diacid agonists with suggested structural
characters on the GPR35 protein, preliminary molecular
docking studies had been attempted using a homology model
and several representative agonists, shown in Figure S5. The
results showed that, for each agonist, there were strong
electrostatic and hydrogen binding interactions between its
acidic groups and cationic Arg residues in the binding pocket.
We noticed that Arg100/151/164 seemed to be more inclined
to such interactions, possibly because they were located in a
spatial range that was right approachable to both acidic groups
(about 8.5−10 Å between the two acidic groups). It has also
been reported that the GPR35 binding pocket contains
abundant cationic residues,16 and thus, it provided favorable
conditions for the diacid compounds to have dual interactions
with the receptor. The simulation results also indicated that,
apart from providing a suitable molecular geometry and spatial
distance between two acidic groups, the aromatic scaffold played
an important supporting role in the formation of hydrophobic
interactions with surrounding residues, such as Leu237/258 and
Pro176, and such interactions were seen in almost all docking
models.
The purity of all final compounds analyzed by high-pressure liquid
chromatography (HPLC) was >95%. The determination of purity was
conducted on a Waters ACQUITY UPLC system (Waters Corp.,
Milford, MA, USA) with an ACQUITY UPLC HSS T3 column (2.1 ×
100 mm,1.8 μm). Elution was performed with a gradient of water/
acetonitrile (containing 0.1% formic acid) from 95/5 to 5/95 for 15
min and maintained 5/95 for another 15 min. The flow rate was 200
μL/min. Peaks were detected at 290 or 254 nm.
The synthesis and structural characterization data of the active
compounds and their intermediates are described as below. For the
2642
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
detailed synthetic and structural characterization data of other
compounds see the Supporting Information.
Screening for Pan-Assay Interference Compounds (PAINS).
Screening of all target compounds for PAINS via the public tool http://
zinc15.docking.org/patterns/home74 yielded no hits.
3.75 (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ 164.21, 163.83,
157.24, 155.85, 155.60, 155.53, 151.98, 148.77, 141.35, 136.30, 131.60,
130.86, 123.74, 123.33, 121.08, 118.74, 117.81, 114.16, 112.25, 107.25,
56.11. Purity: 96.92%. HRMS for C21H12O9: calcd, [M + H]+ 409.0554,
found 409.0545.
General Procedure of Method A for the Syntheses of Compounds
15a−15j. A mixture of the 14 (1.0 equiv) and arylboronic acids (1.5
equiv) in 30 mL degassed (by sonication followed by stream of
nitrogen) dioxane/H2O (5:1) was evacuated and flushed with nitrogen.
K2CO3 (2.2 equiv) and PdCl2(PPh3)2 (0.1 equiv) were added, and the
reaction mixture was stirred under a nitrogen atmosphere overnight.
Hydrochloric acid solution (2 M) was added to adjust pH to 4−5, and
the resulting solution was stirred for 2 h more. Dioxane was evaporated
under reduced pressure. The mixture was diluted with water (30 mL)
and extracted three times with EtOAc (3 × 20 mL). The combined
organic layers were dried over anhydrous Na2SO4 and concentrated
under reduced pressure. The crude product was washed with a
modicum of methyl alcohol (MeOH) (3 × 2 mL) and dried under
vacuum at 50 °C.
General Procedure of Method B for the Syntheses of Compounds
13a, 13b, and 23a−23c. The appropriate ethyl carboxylate derivatives
(1.0 equiv) were added into a solution of NaOH (2.5 equiv) in water
(10 mL). The reaction mixture was stirred at rt until a clear solution was
obtained. The mixture was acidified to pH 4−5 with hydrochloric acid
solution (2 M). The obtained precipitate was filtered off, washed with a
modicum of MeOH (3 × 2 mL), and dried under vacuum at 50 °C.
General Procedure of Method C for the Syntheses of Compounds
13c−13l and 23d−23f. A mixture of the 12b or 22a (1.0 equiv) and
aryboronic acids (1.5 equiv) in 30 mL degassed (by sonication followed
by stream of nitrogen) dioxane/H2O (5:1) was evacuated and flushed
with nitrogen. K2CO3 (2.2 equiv) and PdCl2(PPh3)2 (0.1 equiv) were
added, and the reaction mixture was stirred under a nitrogen
atmosphere overnight. Hydrochloric acid solution (2 M) was added
to adjust pH to 4−5, and the resulting solution was stirred for 2 h more.
Dioxane was evaporated under reduced pressure. The mixture was
diluted with water (30 mL) and extracted three times with EtOAc (3 ×
20 mL). The combined organic layers were dried over anhydrous
Na2SO4, concentrated under reduced pressure. The crude product was
purified by column chromatography (9:1 DCM/MeOH).
6-(3-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13e). As described in method C,
compound 13e was obtained from 12a and (3-methoxyphenyl)boronic
acid with 67% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.86 (s, 1H), 8.82 (s, 1H), 8.32 (s, 1H), 7.47 (t, J = 7.9 Hz, 1H), 7.21 (d,
J = 7.1 Hz, 2H), 7.09−7.05 (m, 1H), 3.83 (s, 3H). 13C NMR (151 MHz,
DMSO-d6) δ 164.18, 163.86, 159.69, 155.78, 155.50, 155.16, 151.91,
148.92, 141.19, 135.76, 135.53, 130.19, 125.94, 122.09, 118.93, 117.81,
115.70, 114.29, 107.62, 55.71. Purity: 95.57%. HRMS for C21H12O9:
calcd, [M + H]+ 409.0554, found 409.0548.
6-(4-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13f). As described in method C,
compound 13f was obtained from 12a and (4-methoxyphenyl)boronic
acid with 55% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.81 (s, 1H), 8.76 (s, 1H), 8.23 (s, 1H), 7.58 (d, J = 8.7 Hz, 2H), 7.12
(d, J = 8.8 Hz, 2H), 3.84 (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ
164.74, 161.42, 159.02, 157.97, 156.08, 150.01, 132.10, 130.77, 129.15,
126.95, 114.05, 113.22, 111.37, 102.41, 55.59. Purity: 98.67%. HRMS
for C21H12O9: calcd, [M + H]+ 409.0554, found 409.0548.
6-(2-Fluorophenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13g). As described in method C,
compound 13g was obtained from 12a and (2-fluorophenyl)boronic
acid with 42% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.93 (s, 1H), 8.89 (s, 1H), 8.36 (s, 1H), 7.64 (dtd, J = 9.2, 7.4, 1.7 Hz,
2H), 7.48 (ddd, J = 11.6, 8.5, 5.0 Hz, 2H). 13C NMR (151 MHz,
DMSO-d6) δ 164.14 (s), 163.76 (s), 159.80 (d, J = 247.0 Hz), 155.69
(s), 155.34 (s), 155.20 (s), 152.46 (s), 148.62 (s), 141.16 (s), 136.30
(s), 132.48 (d, J = 2.5 Hz), 131.65 (d, J = 8.4 Hz), 125.33 (d, J = 3.3
Hz), 122.06 (d, J = 15.4 Hz), 120.46 (s), 119.07 (s), 118.08 (s), 116.35
(d, J = 21.7 Hz), 114.37 (s), 107.55 (s). Purity: 98.90%. HRMS for
C20H9FO8: calcd, [M + H]+ 397.0354, found 397.0366.
6-(3-Fluorophenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13h). As described in method C,
compound 13h was obtained from 12a and (3-fluorophenyl)boronic
acid with 58% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.83 (s, 1H), 8.81 (s, 1H), 8.34 (s, 1H), 7.61 (td, J = 8.0, 6.3 Hz, 1H),
7.55−7.52 (m, 1H), 7.51−7.48 (m, 1H), 7.37−7.32 (m, 1H). 13C NMR
(151 MHz, DMSO-d6) δ 164.10 (s), 163.79 (s), 162.48 (d, J = 243.5
Hz), 155.66 (s), 155.36 (s), 155.08 (s), 152.12 (s), 148.75 (s), 141.07
(s), 136.70 (d, J = 8.4 Hz), 135.62 (s), 131.14 (d, J = 8.5 Hz), 126.04 (d,
J = 2.5 Hz), 124.67 (s), 119.04 (s), 117.91 (s), 116.85 (s), 115.79 (d, J =
20.9 Hz), 114.29 (s), 107.65 (s). Purity: 96.12%. HRMS for C20H9FO8:
calcd, [M + H]+ 397.0354, found 397.0358.
6-(4-Fluorophenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13i). As described in method C,
compound 13i was obtained from 12a and (4-fluorophenyl)boronic
acid with 55% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.86 (s, 1H), 8.82 (s, 1H), 8.30 (s, 1H), 7.70 (dd, J = 8.0, 5.8 Hz, 2H),
7.42 (t, J = 8.8 Hz, 2H). 13C NMR (151 MHz, DMSO-d6) δ 164.16 (s),
163.83 (s), 162.67 (d, J = 246.3 Hz), 155.75 (s), 155.44 (s), 155.13 (s),
151.91 (s), 148.82 (s), 141.13 (s), 135.47 (s), 132.02 (d, J = 8.4 Hz),
130.86 (d, J = 3.4 Hz), 125.08 (s), 119.00 (s), 117.90 (s), 116.08 (d, J =
21.5 Hz), 114.30 (s), 107.64 (s). Purity: 95.81%. HRMS for C20H9FO8:
calcd, [M + H]+ 397.0354, found 397.0367.
6-(4-Isopropylphenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13j). As described in method C,
compound 13j was obtained from 12a and (4-isopropylphenyl)boronic
acid with 63% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.86 (s, 1H), 8.82 (s, 1H), 8.28 (s, 1H), 7.57 (d, J = 8.2 Hz, 2H), 7.43
(d, J = 8.2 Hz, 2H), 3.02−2.96 (m, 1H), 1.28 (d, J = 6.9 Hz, 6H). 13C
NMR (151 MHz, DMSO-d6) δ 163.12, 162.78, 154.75, 154.50, 154.06,
150.65, 148.17, 147.76, 143.25, 140.08, 134.25, 130.90, 128.75, 125.98,
125.07, 116.81, 113.27, 106.56, 32.68, 23.23. Purity: 97.91%. HRMS for
C23H16O8: calcd, [M + H]+ 421.0918, found 421.0930.
2,8-Dioxo-2,8-dihydropyrano[2,3-f ]chromene-3,9-dicarboxylic
Acid (13a). As described in method B, compound 13a was obtained
from 12a with 81% yield as a yellow solid. 1H NMR (600 MHz, DMSO-
d6) δ 8.83 (s, 1H), 8.74 (s, 1H), 8.20 (d, J = 8.8 Hz, 1H), 7.45 (d, J = 8.8
Hz, 1H). 13C NMR (151 MHz, DMSO-d6) δ 164.16, 163.83, 158.22,
155.77, 155.74, 152.61, 148.87, 140.88, 135.52, 119.09, 117.49, 114.44,
113.59, 107.30. Purity: 97.74%. HRMS for C14H6O8: calcd, [M + H]+
303.0135, [M + H2O + H]+ 321.0241, found 303.0135, 321.0245.
6-Bromo-2,8-dioxo-2,8-dihydropyrano[2,3-f ]chromene-3,9-di-
carboxylic Acid (13b). As described in method B, compound 13b was
1
obtained from 12b with 83% yield as a yellow solid. H NMR (600
MHz, DMSO-d6) δ 8.77 (s, 1H), 8.71 (s, 1H), 8.56 (s, 1H). 13C NMR
(151 MHz, DMSO-d6) δ 163.95, 163.54, 155.34, 154.99, 154.36,
151.82, 147.83, 140.77, 137.18, 119.54, 118.32, 115.51, 108.66, 104.77.
Purity: 97.87%. HRMS for C14O8H5Br: calcd, [M − H]− 334.9197 and
336.9176, found 334.9115 and 336.9100.
2,8-Dioxo-6-phenyl-2,8-dihydropyrano[2,3-f ]chromene-3,9-di-
carboxylic Acid (13c). As described in method C, compound 13c was
obtained from 12b and phenylboronic acid with 42% yield as a yellow
1
solid. H NMR (600 MHz, DMSO-d6) δ 8.91 (s, 1H), 8.86 (s, 1H),
8.35 (s, 1H), 7.70 (d, J = 7.2 Hz, 2H), 7.62 (t, J = 7.6 Hz, 2H), 7.55 (t, J
= 7.4 Hz, 1H). 13C NMR (151 MHz, DMSO-d6) δ 164.19, 163.87,
155.80, 155.54, 155.10, 151.84, 148.77, 141.05, 135.45, 134.52, 129.88,
129.09, 128.90, 126.14, 119.12, 117.96, 114.35, 107.65. Purity: 95.66%.
HRMS for C20H10O8: calcd, [M + H]+ 379.0448, found 379.0448.
6-(2-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[2,3-f ]-
chromene-3,9-dicarboxylic Acid (13d). As described in method C,
compound 13d was obtained from 12a and (2-methoxyphenyl)boronic
acid with 51% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.85 (s, 1H), 8.82 (s, 1H), 8.17 (s, 1H), 7.52−7.47 (m, 1H), 7.33 (dd, J
= 7.4, 1.4 Hz, 1H), 7.21 (d, J = 8.3 Hz, 1H), 7.11 (t, J = 7.4 Hz, 1H),
2643
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
2,8-Dioxo-6-(thiophen-2-yl)-2,8-dihydropyrano[2,3-f ]chromene-
3,9-dicarboxylic Acid (13k). As described in method C, compound 13k
was obtained from 12a and naphthalen-2-ylboronic acid with 48% yield
as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 8.89 (s, 1H), 8.81
(s, 1H), 8.61 (s, 1H), 7.78 (dd, J = 5.1, 1.1 Hz, 1H), 7.74 (dd, J = 3.7, 1.1
Hz, 1H), 7.28 (dd, J = 5.1, 3.7 Hz, 1H). 13C NMR (151 MHz, DMSO-
d6) δ 163.03, 162.66, 154.58, 154.09, 152.76, 150.41, 147.68, 140.19,
133.82, 132.23, 127.56, 127.20, 127.10, 118.09, 117.80, 116.99, 113.44,
106.75. Purity: 97.83%. HRMS for C18H8O8S: calcd, [M + H]+
385.0013, found 385.0004.
109.73, 100.87, 54.75. Purity: 97.44%. HRMS for C17H12O6: calcd, [M
+ H]+ 313.0707, found 313.0710.
7-Hydroxy-6-(3-methoxyphenyl)-2-oxo-2H-chromene-3-carbox-
ylic Acid (15d). As described in method A, compound 15d was
obtained from 14 and (3-methoxyphenyl)boronic acid with 78% yield
as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 8.73 (s, 1H), 7.87
(d, J = 4.8 Hz, 1H), 7.36 (dd, J = 15.0, 7.1 Hz, 1H), 7.12 (d, J = 7.8 Hz,
1H), 7.10−7.08 (m, 1H), 6.95−6.92 (m, 1H), 6.89 (s, 1H), 3.79 (s,
3H). 13C NMR (151 MHz, DMSO-d6) δ 163.63, 160.33, 158.35,
156.76, 155.29, 148.89, 137.21, 131.55, 128.55, 125.93, 120.92, 114.32,
112.31, 112.05, 110.26, 101.42, 54.46. Purity: 95.38%. HRMS for
C17H12O6: calcd, [M + H]+ 313.0707, found 313.0706.
6-(2-Fluoro-4-methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano-
[2,3-f]chromene-3,9-dicarboxylic Acid (13l). As described in method
C, compound 13l was obtained from 12a and (2-fluoro-4-
7-Hydroxy-6-(4-methoxyphenyl)-2-oxo-2H-chromene-3-carbox-
ylic Acid (15e). As described in method A, compound 15e was obtained
from 14 and (4-methoxyphenyl)boronic acid with 82% yield as a yellow
1
methoxyphenyl)boronic acid with 38% yield as a yellow solid. H
NMR (600 MHz, DMSO-d6) δ 8.84 (s, 1H), 8.82 (s, 1H), 8.17 (s, 1H),
7.37 (t, J = 7.6 Hz, 1H), 7.14 (d, J = 11.5 Hz, 1H), 6.96 (t, J = 8.1 Hz,
1H), 3.76 (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ 164.21 (s),
163.96 (d, J = 246.13 Hz), 163.83 (s), 158.69 (d, J = 10.6 Hz), 155.84
(s), 155.57 (s), 155.52 (s), 152.01 (s), 148.63 (s), 141.12 (s), 136.35
(s), 132.75 (d, J = 10.2 Hz), 122.74 (s), 119.57 (d, J = 3.1 Hz), 117.94
(s), 114.16 (s), 107.57 (s), 107.43 (s), 107.31 (s), 100.67 (d, J = 26.1
Hz), 56.65 (s). Purity: 95.24%. HRMS for C21H11FO9: calcd, [M + H]+
427.0460, [M + H2O + H]+ 445.0566, found 427.0467, 445.0569.
Preparation of 6-Bromo-2,8-dioxo-3-(1H-tetrazol-5-yl)-2, 8-
dihydropyrano[2,3-f]chromene-9-carboxylic Acid (13m). A mixture
of 19b (1.0 equiv) and Meldrum’s acid (1.5 equiv) were mixed in 10 mL
of ethanol at room temperature. A catalytic amount of ammonium
acetate was added, and the reaction mixture was stirred at the room
temperature for 2 h. Hydrochloric acid solution (2 M) was added to
adjust pH to 4−5. Ethanol was evaporated under reduced pressure. The
resulting solid was filtered, washed three times with 2 mL of methanol,
and dried under vacuum at 50 °C. The target product was obtained with
83% yield as a gray solid. 1H NMR (600 MHz, DMSO-d6) δ 9.08 (s,
1H), 8.79 (s, 1H), 8.69 (s, 1H). 13C NMR (151 MHz, DMSO-d6) δ
163.60, 157.26, 155.02, 154.09, 150.68, 143.28, 140.53, 136.89, 119.90,
116.03, 112.97, 108.89, 105.32. Purity: 95.31%. HRMS for
C14H5BrN4O6: calcd, [M + H]+ 404.9465, found [M + H]+
404.9471, [M + H + 2]+ 406.9463.
Preparation of 2,8-Dioxo-3-(1H-tetrazol-5-yl)-2,8-
dihydropyrano[2,3-f]chromene-9-carbonitrile (13n). A mixture of
19a (1.0 equiv) and malononitrile (1.5 equiv) were mixed in 20 mL of
water at room temperature. A catalytic amount of ammonium acetate
was added, and the reaction mixture was stirred at room temperature for
4 h. Hydrochloric acid solution (2 M) was added to adjust pH to 4−5.
The resulting solid was filtered, washed three times with 2 mL of
methanol, and dried under vacuum at 50 °C. The target product was
obtained with 78% yield as a gray solid. 1H NMR (600 MHz, DMSO-
d6) δ 9.28 (s, 1H), 9.14 (s, 1H), 8.37 (d, J = 8.8 Hz, 1H), 7.60 (d, J = 8.8
Hz, 1H). 13C NMR (151 MHz, DMSO-d6) δ 156.31, 156.08, 155.52,
150.03, 145.33, 143.03, 135.14, 114.41, 113.47, 113.40, 111.48, 106.42,
102.69. Purity: 97.4%. HRMS for C14H5N5O4: calcd, [M − H]−
306.0269, found 306.0283.
1
solid. H NMR (600 MHz, DMSO-d6) δ 8.73 (s, 1H), 7.81 (s, 1H),
7.51−7.45 (m, 2H), 7.01 (t, J = 10.8 Hz, 2H), 6.88 (s, 1H), 3.80 (s,
3H). 13C NMR (151 MHz, DMSO-d6) δ 164.73, 161.43, 159.02,
157.99, 156.07, 149.99, 132.10, 130.76, 129.16, 126.96, 114.05, 113.23,
111.37, 102.42, 55.59. Purity: 96.79%. HRMS for C17H12O6: calcd, [M
+ H]+ 313.0707, found 313.0708.
7-Hydroxy-6-(2-fluorophenyl)-7-hydroxy-2-oxo-2H-chromene-3-
carboxylic Acid (15f). As described in method A, compound 15f was
obtained from 14 and (2-fluorophenyl)boronic acid with 64% yield as a
1
yellow solid. H NMR (600 MHz, DMSO-d6) δ 8.73 (d, J = 8.3 Hz,
1H), 7.80 (s, 1H), 7.47−7.43 (m, 1H), 7.40 (td, J = 7.5, 1.5 Hz, 1H),
7.28 (t, J = 8.1 Hz, 2H), 6.90 (s, 1H). 13C NMR (151 MHz, DMSO-d6)
δ 164.68 (s), 161.71 (s), 159.96 (d, J = 245.9 Hz), 157.77 (s), 156.91
(s), 149.75 (s), 133.30 (s), 132.40 (d, J = 3.3 Hz), 130.33 (d, J = 8.1
Hz), 124.82 (d, J = 7.8 Hz), 124.75 (d, J = 4.8 Hz), 121.96 (s), 115.92
(d, J = 22.0 Hz), 113.56 (s), 111.09 (s), 102.19 (s). Purity: 98.76%.
HRMS for C16H9FO5: calcd, [M + H]+ 301.0507, found 301.0512.
7-Hydroxy-6-(3-fluorophenyl)-7-hydroxy-2-oxo-2H-chromene-3-
carboxylic Acid (15g). As described in method A, compound 15g was
obtained from 14 and (3-fluorophenyl)boronic acid with 75% yield as a
yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 8.70 (s, 1H), 8.26 (s,
1H), 7.26 (s, 1H), 5.46 (s, 2H), 4.29 (q, J = 7.1 Hz, 2H), 3.44 (s, 3H),
1.30 (t, J = 7.1 Hz, 3H). 13C NMR (151 MHz, DMSO-d6) δ 164.66 (s),
162.36 (d, J = 242.8 Hz), 161.30 (s), 157.71 (s), 156.60 (s), 149.81 (s),
139.29 (d, J = 8.3 Hz), 132.75 (s), 130.56 (d, J = 8.6 Hz), 125.75 (d, J =
2.2 Hz), 125.70 (s), 116.35 (d, J = 22.0 Hz), 114.53 (d, J = 20.6 Hz),
113.63 (s), 111.40 (s), 102.61 (s). Purity: 97.20%. HRMS for
C16H9FO5: calcd, [M + H]+ 301.0507, found 301.0510.
7-Hydroxy-6-(4-fluorophenyl)-7-hydroxy-2-oxo-2H-chromene-3-
carboxylic Acid (15h). As described in method A, compound 15h was
obtained from 14 and (4-fluorophenyl)boronic acid with 79% yield as a
yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 8.72 (s, 1H), 7.85 (s,
1H), 7.61−7.55 (m, 2H), 7.28 (dd, J = 12.4, 5.4 Hz, 2H), 6.89 (s, 1H).
13C NMR (151 MHz, DMSO-d6) δ 164.70 (s), 161.96 (d, J = 244.62
Hz), 161.42 (s), 157.87 (s), 156.42 (s), 149.83 (s), 133.32 (d, J = 3.1
Hz), 132.51 (s), 131.62 (d, J = 8.2 Hz), 126.17 (s), 115.46 (d, J = 21.3
Hz), 113.39 (s), 111.35 (s), 102.52 (s). Purity: 98.67%. HRMS for
C16H9FO5: calcd, [M + H]+ 301.0507, found 301.0538.
7-Hydroxy-2-oxo-6-phenyl-2H-chromene-3-carboxylic Acid
(15b). As described in method A, compound 15b was obtained from
14 and phenylboronic acid with 80% yield as a yellow solid. 1H NMR
(600 MHz, DMSO-d6) δ 8.73 (s, 1H), 7.85 (s, 1H), 7.57−7.53 (m,
2H), 7.44 (t, J = 7.7 Hz, 2H), 7.36 (t, J = 7.4 Hz, 1H), 6.90 (s, 1H). 13C
NMR (151 MHz, DMSO-d6) δ 164.73, 161.47, 157.91, 156.38, 149.92,
137.02, 132.57, 129.65, 128.61, 127.73, 127.26, 113.36, 111.38, 102.50.
Purity: 96.76%. HRMS for C16H10O5: calcd, [M + H]+ 283.0601, found
283.0600.
7-Hydroxy-6-(4-isopropylphenyl)-2-oxo-2H-chromene-3-carbox-
ylic Acid (15i). As described in method A, compound 15i was obtained
from 14 and (4-isopropylphenyl)boronic acid with 83% yield as a
yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 8.72 (s, 1H), 7.83 (s,
1H), 7.46 (d, J = 8.1 Hz, 2H), 7.31 (d, J = 8.2 Hz, 2H), 6.89 (s, 1H),
2.93 (dp, J = 13.7, 6.8 Hz, 2H), 1.24 (d, J = 6.9 Hz, 6H). 13C NMR (151
MHz, DMSO-d6) δ 164.74, 161.51, 157.92, 156.26, 149.97, 147.92,
134.48, 132.39, 129.59, 127.26, 126.54, 113.27, 111.37, 102.44, 33.68,
24.37. Purity: 96.37%. HRMS for C19H16O5: calcd, [M + H]+ 325.1071,
found 325.1132.
7-Hydroxy-2-oxo-6-(thiophen-2-yl)-2H-chromene-3-carboxylic
Acid (15j). As described in method A, compound 15j was obtained
from 14 and thiophen-2-ylboronic acid with 68% yield as a yellow solid.
1H NMR (600 MHz, DMSO-d6) δ 8.80 (s, 1H), 8.29 (s, 1H), 7.67 (d, J
= 3.6 Hz, 1H), 7.64 (d, J = 5.1 Hz, 1H), 7.21 (dd, J = 6.3, 2.3 Hz, 1H),
6.95 (s, 1H). 13C NMR (151 MHz, DMSO-d6) δ 163.64, 160.43,
156.84, 155.24, 148.83, 136.53, 135.89, 131.46, 129.14, 127.40, 127.27,
7-Hydroxy-6-(2-methoxyphenyl)-2-oxo-2H-chromene-3-carbox-
ylic Acid (15c). As described in method A, compound 15c was obtained
from 14 and (2-methoxyphenyl)boronic acid with 82% yield as a yellow
1
solid. H NMR (600 MHz, DMSO-d6) δ 8.70 (s, 1H), 7.66 (s, 1H),
7.39−7.34 (m, 1H), 7.17 (dd, J = 7.4, 1.7 Hz, 1H), 7.08 (d, J = 8.1 Hz,
1H), 7.00 (td, J = 7.4, 0.8 Hz, 1H), 6.84 (s, 1H), 3.71 (s, 3H). 13C NMR
(151 MHz, DMSO-d6) δ 163.69, 161.09, 156.99, 156.21, 155.40,
148.89, 132.13, 130.48, 128.54, 125.01, 124.02, 119.51, 111.88, 110.63,
2644
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
126.33, 125.73, 112.22, 110.26, 101.38, 20.50. Purity: 95.93%. HRMS
for C14H8O5S: calcd, [M + H]+ 289.0165, found 289.0153.
MeOH (3 × 2 mL), and dried under vacuum at 50 °C. 1H NMR (600
MHz, DMSO-d6) δ 8.75 (s, 1H), 8.21 (s, 2H), 3.29 (s, 3H). 13C NMR
(151 MHz, DMSO-d6) δ 165.77, 150.86, 144.74, 143.52, 141.59,
125.74, 121.58, 117.29, 13.68. Purity: 98.12%. HRMS for
C15H8Cl2N2O4: calcd, [M + H]+ 350.9934 and 352.9904, found
350.9919 and 352.9892.
Materials and Cell Culture. Zaprinast, PQQ and compound 27
were obtained from Sigma-Aldrich. ML-145 was obtained from Tocris.
Epic 384-well biosensor microplates were obtained from Corning
Incorporated (Corning, NY). HT-29 and CHO-K1 cells were cultured
using McCoy’s 5A and Ham’s F12K. The medium was supplemented
with 10% FBS and 1% penicillin/streptomycin, and all cells were
maintained in a 5% CO2 incubator at 37 °C.
Transfection of hGPR35 Cell Line. CHO-K1 cells were
transfected with 8 μg of pcDNA3.1-hGPR35 plasmid mixed with 24
μL of lipofectamine 2000 reagent (Invitrogen). After 24 h post-
transfection, clones were selected using a complete medium containing
400 μg/mL zeocin (TransGen Biotech Co., Ltd., Beijing, China). Stable
clones were selected with zeocin treatment for 3−4 weeks to obtain
successfully transfected cell line CHO-K1-hGPR35. After cultured for
3−4 months, the stably transfected cell line CHO-K1-hGPR35 was
obtained. The function expression of hGPR35 was detected every 2
weeks using zaprinast as the probe in the DMR assay.
DMR Assays Using an Epic BT System. All DMR assays were
performed using an Epic BT system (Corning Incorporated). Epic is a
swept wavelength interrogation reader system tailored for resonant
waveguide grating biosensors in microtiter plates. Cells were directly
seeded in Epic plates and cultured overnight to form a confluent
monolayer in the cell culture medium. After being washed, the cells
were maintained with Hank’s balanced salt solution and further
incubated inside the system for 1 h. For agonist profiling, a 2 min
baseline was then established. After the compound addition, the cellular
responses were recorded immediately. For desensitization assays, cells
were initially treated with compounds for 1 h, followed by stimulation
with zaprinast at 1 μM for HT29 or 400 nM for CHO-K1-hGPR35. The
cellular responses were recorded throughout the assays. All EC50 or IC50
described in the main text were calculated based on the amplitudes of
DMR signals at 8 min post-stimulation. All GPR35 agonists led to a
sustained positive-DMR signal. The data represents mean sd from
two independent measurement, each with four replicates (n = 8).
Modeling of GPR35 and Docking. Homology modeling of
hGPR35 was performed using a MODELER protocol in Discovery
Studio 2019 based on the X-ray structure of envelope protein US28.
The PDB IDs of the template used for hGPR35 were 5WB1, 5WB2, and
4XT1. Homology sequence searching and alignment were conducted
using sequence analysis and multiple sequence alignment modules,
respectively. Energy minimization optimization was performed for the
selected agonist molecules, which were treated as dianions here since
they would be ionized at physiological pH. The optimized agonists were
docked into the active site of the hGPR35 model using a CDOCKER
protocol in Discovery Studio with default parameters. Ten binding
modes were suggested for each agonist; then the highest scoring ones
were selected for further analysis.
Preparation of 2,11-Dioxo-2,11-dihydrobenzo[f]pyrano[3,2-h]-
chromene-3,10-dicarboxylic Acid (20). A mixture of 13 (1.0 equiv)
and Meldrum’s acid (3.0 equiv) were mixed in 10 mL of water at room
temperature. A catalytic amount of ammonium acetate was added, and
the reaction mixture was stirred at the room temperature for 6 h.
Hydrochloric acid solution (2 M) was added to adjust pH to 4−5. The
resulting solid was filtered, washed three times with 2 mL of methanol,
and dried under vacuum at 50 °C. The target product was obtained with
91% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ 9.35 (s,
2H), 8.65 (dd, J = 6.0, 3.1 Hz, 2H), 7.79 (dd, J = 6.2, 3.0 Hz, 2H). 13C
NMR (151 MHz, DMSO-d6) δ 164.39, 155.59, 143.15, 142.93, 128.79,
125.74, 123.73, 121.13, 116.99. Purity: 95.22%. HRMS for C18H8O8:
calcd, [M + Na]+ 375.0111, found 375.0087.
10-Bromo-2,8-dioxo-2,8-dihydropyrano[3,2-g]chromene-3,7-di-
carboxylic Acid (23a). As described in method B, compound 23a was
1
obtained from 22a with 82% yield as a yellow solid. H NMR (600
MHz, DMSO-d6) δ 13.49 (s, 2H), 8.77 (s, 2H), 8.43 (s, 1H). 13C NMR
(151 MHz, DMSO-d6) δ 163.90, 155.52, 155.15, 148.14, 131.65,
118.64, 116.44, 97.25. Purity: 97.14%. HRMS for C14O8H5Br: calcd,
[M − H]− 334.9197 and 336.9176, found 334.9112 and 336.9102.
10-Methyl-2,8-dioxo-2,8-dihydropyrano[3,2-g]chromene-3,7-di-
carboxylic Acid (23b). As described in method B, compound 23b was
obtained from 22b with 91% yield as a white solid. 1H NMR (600 MHz,
DMSO-d6) δ 8.72 (s, 2H), 8.28 (s, 1H), 2.36 (s, 3H). 13C NMR (151
MHz, DMSO-d6) δ 163.06, 155.05, 155.00, 147.39, 129.16, 116.93,
114.39, 111.66, 7.12. HRMS for C15H8O8: calcd, [M + H]+ 317.0292,
found 317.0305.
10-Hydroxy-2,8-dioxo-2,8-dihydropyrano[3,2-g]chromene-3,7-
dicarboxylic Acid (23c). As described in method B, compound 23c was
obtained from 22c with 82% yield as a white solid. 1H NMR (600 MHz,
DMSO-d6) δ 8.65 (s, 2H), 7.87 (s, 1H). 13C NMR (151 MHz, DMSO-
d6) δ 164.26, 156.17, 148.46, 146.80, 132.04, 121.75, 118.47, 115.95.
Purity: 96.69%. HRMS for C14H6O9: calcd, [M + H]+ 319.0085, found
319.0094.
10-(2-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[3,2-g]-
chromene-3,7-dicarboxylic Acid (23d). As described in method C,
compound 23d was obtained from 22a and (2-methoxyphenyl)boronic
acid with 55% yield as a grey solid. Purity: 95.42%. 1H NMR (600 MHz,
DMSO-d6) δ 8.87 (s, 2H), 8.53 (s, 1H), 7.61−7.57 (m, 1H), 7.42 (dd, J
= 7.5, 1.7 Hz, 1H), 7.31 (d, J = 8.2 Hz, 1H), 7.19 (t, J = 7.4 Hz, 1H),
3.78 (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ 163.02, 156.45,
154.97, 154.40, 147.33, 131.17, 130.98, 130.02, 119.85, 117.13, 116.95,
114.66, 113.28, 111.24, 55.05. Purity: 96.63%. HRMS for C21H12O9:
calcd, [M + H]+ 409.0554, found 409.0544.
10-(3-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[3,2-g]-
chromene-3,7-dicarboxylic Acid (23e). As described in method C,
compound 23e was obtained from 22a and (3-methoxyphenyl)boronic
1
acid with 59% yield as a white solid. Purity: 96.19%. H NMR (600
MHz, DMSO-d6) δ 8.86 (s, 2H), 8.51 (s, 1H), 7.54 (t, J = 7.9 Hz, 1H),
7.19−7.17 (m, 1H), 7.15 (ddd, J = 5.7, 4.6, 2.5 Hz, 2H), 3.87 (s, 3H).
13C NMR (151 MHz, DMSO-d6) δ 163.05, 158.43, 154.96, 154.08,
147.26, 130.87, 129.50, 128.76, 122.27, 117.20, 115.99, 115.84, 114.82,
113.33, 54.60. Purity: 95.26%. HRMS for C21H12O9: calcd, [M + H]+
409.0554, found 409.0565.
10-(4-Methoxyphenyl)-2,8-dioxo-2,8-dihydropyrano[3,2-g]-
chromene-3,7-dicarboxylic Acid (23f). As described in method C,
compound 23f was obtained from 22a and (4-methoxyphenyl)boronic
acid with 67% yield as a yellow solid. 1H NMR (600 MHz, DMSO-d6) δ
8.79 (s, 2H), 8.42 (s, 1H), 7.51−7.45 (m, 2H), 7.15−7.11 (m, 2H),
3.86 (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ 164.17, 159.92,
156.13, 155.20, 148.33, 132.47, 131.49, 121.12, 118.32, 116.89, 115.94,
114.28, 55.68. Purity: 95.83%. HRMS for C21H12O9: calcd, [M + H]+
409.0554, found 409.0541.
4,6-Dichloro-10-methylpyrido[3,2-g]quinoline-2,8-dicarboxylic
Acid (26). The compound 33 (1.0 equiv) were added into a solution of
2 M NaOH (2.5 equiv) in 10 mL water/MeOH (1:1). The reaction
mixture was stirred at rt until a clear solution was obtained (1 h). The
mixture was acidified to pH 4−5 with hydrochloric acid solution (2 M).
The obtained precipitate was filtered off, washed with a modicum of
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c01624.
Molecular formula strings (CSV)
hGPR35 (PDB)
Synthetic procedures and 1H and 13C NMR spectral data
for compounds 9b, 10a, 10b, 11a, 11b, 12a, 12b, 14, 16a,
16b, 17a, 17b, 18a, 18b, 19a, 19b, 21a, 21b, 22a, 22b,
22c, and 33 were shown in supporting information (PDF)
2645
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
liquid chromatography; THF, tetrahydrofuran; HRMS, high-
resolution mass spectra
AUTHOR INFORMATION
■
Corresponding Authors
Yaopeng Zhao − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China; Jiangxi Chinese
Medicine Science Center of DICP, CAS, Nanchang 330000,
China; Phone: +86 411 84379519; Email: ypzhao@
dicp.ac.cn; Fax: +86 411 84379539
Xinmiao Liang − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China; Jiangxi Chinese
Medicine Science Center of DICP, CAS, Nanchang 330000,
China; orcid.org/0000-0001-5802-1961; Phone: +86
411 84379519; Email: liangxm@dicp.ac.cn
REFERENCES
■
(1) Chung, S.; Funakoshi, T.; Civelli, O. Orphan GPCR research. Brit.
J. Pharmacol. 2008, 153, S339−S346.
(2) Congreve, M.; de Graaf, C.; Swain, N. A.; Tate, C. G. Impact of
GPCR Structures on Drug Discovery. Cell 2020, 181, 81−91.
(3) Hauser, A. S.; Attwood, M. M.; Rask-Andersen, M.; Schiöth, H. B.;
Gloriam, D. E. Trends in GPCR drug discovery: new agents, targets and
indications. Nat. Rev. Drug. Discov. 2017, 16, 829−842.
(4) O’Dowd, B. F.; Nguyen, T.; Marchese, A.; Cheng, R.; Lynch, K. R.;
Heng, H. H.; Kolakowski, L. F., Jr.; George, S. R. Discovery of three
novel G-protein-coupled receptor genes. Genomics 1998, 47, 310−313.
(5) Okumura, S.; Baba, H.; Kumada, T.; Nanmoku, K.; Nakajima, H.;
Nakane, Y.; Hioki, K.; Ikenaka, K. Cloning of a G-protein-coupled
receptor that shows an activity to transform NIH3T3 cells and is
expressed in gastric cancer cells. Cancer Sci. 2004, 95, 131−135.
(6) Wang, J.; Simonavicius, N.; Wu, X.; Swaminath, G.; Reagan, J.;
Tian, H.; Ling, L. Kynurenic acid as a ligand for orphan G protein-
coupled receptor GPR35. J. Biol. Chem. 2006, 281, 22021−22028.
(7) Sun, Y. V.; Bielak, L. E.; Peyser, P. A.; Turner, S. T.; Sheedy, P. F.,
II; Boerwinkle, E.; Kardia, S. L. Application of machine learning
algorithms to predict coronary artery calcification with a sibship-based
design. Genet. Epidemiol. 2008, 32, 350−360.
Authors
Lai Wei − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
Tao Hou − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
Jiaqi Li − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
(8) Yang, Y.; Lu, J. Y. L.; Wu, X.; Summer, S.; Whoriskey, J.; Saris, C.;
Reagan, J. D. G-Protein-Coupled Receptor 35 Is a Target of the Asthma
Drugs Cromolyn Disodium and Nedocromil Sodium. Pharmacology
2010, 86, 1−5.
Xiuli Zhang − College of Pharmaceutical Science, Soochow
University, Suzhou, Jiangsu 215123, China
(9) Zhao, P.; Sharir, H.; Kapur, A.; Cowan, A.; Geller, E. B.; Adler, M.
W.; Seltzman, H. H.; Reggio, P. H.; Heynen-Genel, S.; Sauer, M.;
Chung, T. D.; Bai, Y.; Chen, W.; Caron, M. G.; Barak, L. S.; Abood, M.
E. Targeting of the Orphan Receptor GPR35 by Pamoic Acid: A Potent
Activator of Extracellular Signal-Regulated Kinase and beta-Arrestin2
with Antinociceptive Activity. Mol. Pharmacol. 2010, 78, 560−568.
Han Zhou − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
Zhenyu Wang − National Engineering Research Center of
Seafood, School of Food Science and Technology, Dalian
Polytechnic University, Dalian 116034, China; orcid.org/
0000-0003-2839-5377
(10) Kaya, B.; Donas, C.; Wuggenig, P.; Diaz, O. E.; Morales, R. A.;
̃
Melhem, H.; Investigators, S. I. C.; Hernández, P. P.; Kaymak, T.; Das,
S.; Hruz, P.; Franc, Y.; Geier, F.; Ayata, C. K.; Villablanca, E. J.; Niess, J.
H. Lysophosphatidic Acid-Mediated GPR35 Signaling in CX3CR1(+)
Macrophages Regulates Intestinal Homeostasis. Cell Rep. 2020, 32,
107979.
Junxiang Cheng − Jiangxi Chinese Medicine Science Center of
DICP, CAS, Nanchang 330000, China
Kaijing Xiang − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
Jixia Wang − Key Lab of Separation Science for Analytical
Chemistry, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116034, China
(11) Southern, C.; Cook, J. M.; Neetoo-Isseljee, Z.; Taylor, D. L.;
Kettleborough, C. A.; Merritt, A.; Bassoni, D. L.; Raab, W. J.; Quinn, E.;
Wehrman, T. S.; Davenport, A. P.; Brown, A. J.; Green, A.;
Wigglesworth, M. J.; Rees, S. Screening β-Arrestin Recruitment for
the Identification of Natural Ligands for Orphan G-Protein−Coupled
Receptors. J. Biomol. Screen 2013, 18, 599−609.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jmedchem.0c01624
(12) Deng, H.; Hu, H.; Fang, Y. Multiple tyrosine metabolites are
GPR35 agonists. Sci. Rep. 2012, 2, 373.
(13) Maravillas-Montero, J. L.; Burkhardt, A. M.; Hevezi, P. A.;
Carnevale, C. D.; Smit, M. J.; Zlotnik, A. Cutting Edge: GPR35/
CXCR8 Is the Receptor of the Mucosal Chemokine CXCL17. J.
Immunol. 2015, 194, 29−33.
Author Contributions
⊥L.W. and T.H. contributed equally to this work.
Notes
(14) Park, S. J.; Lee, S. J.; Nam, S. Y.; Im, D. S. GPR35 mediates
lodoxamide-induced migration inhibitory response but not CXCL17-
induced migration stimulatory response in THP-1 cells; is GPR35 a
receptor for CXCL17? Brit. J. Pharmacol. 2018, 175, 154−161.
(15) Taniguchi, Y.; Tonai-Kachi, H.; Shinjo, K. Zaprinast, a well-
known cyclic guanosine monophosphate-specific phosphodiesterase
inhibitor, is an agonist for GPR35. FEBS Lett. 2006, 580, 5003−5008.
(16) Zhao, P.; Lane, T. R.; Gao, H. G.; Hurst, D. P.; Kotsikorou, E.;
Le, L.; Brailoiu, E.; Reggio, P. H.; Abood, M. E. Crucial positively
charged residues for ligand activation of the GPR35 receptor. J. Biol.
Chem. 2014, 289, 3625−3638.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by “National Science and Technology
Major Project (2018ZX09735-002)”; supported by the
innovation program of science and research from DICP, CAS
(DICP I201933); supported by the innovation program of
science and research from DICP, CAS (DICP ZZBS201803).
(17) Deng, H.; Hu, H.; He, M.; Hu, J.; Niu, W.; Ferrie, A. M.; Fang, Y.
Discovery of 2-(4-methylfuran-2(5H)-ylidene)malononitrile and
thieno[3,2-b]thiophene-2-carboxylic acid derivatives as G protein-
coupled receptor 35 (GPR35) agonists. J. Med. Chem. 2011, 54, 7385−
7396.
ABBREVIATIONS USED
■
hGPR35, human G protein-coupled receptor 35; GPCRG,
protein-coupled receptor; SARs, structure activity relationships;
DMR, dynamic mass redistribution; HPLC, high-performance
2646
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(18) Funke, M.; Thimm, D.; Schiedel, A. C.; Muller, C. E. 8-
̈
Benzamidochromen-4-one-2-carboxylic acids: potent and selective
agonists for the orphan G protein-coupled receptor GPR35. J. Med.
Chem. 2013, 56, 5182−5197.
(19) Wei, L.; Wang, J.; Zhang, X.; Wang, P.; Zhao, Y.; Li, J.; Hou, T.;
Qu, L.; Shi, L.; Liang, X.; Fang, Y. Discovery of 2H-Chromen-2-one
Derivatives as G Protein-Coupled Receptor-35 Agonists. J. Med. Chem.
2017, 60, 362−372.
(20) MacKenzie, A. E.; Caltabiano, G.; Kent, T. C.; Jenkins, L.;
McCallum, J. E.; Hudson, B. D.; Nicklin, S. A.; Fawcett, L.; Markwick,
R.; Charlton, S. J.; Milligan, G. The Antiallergic Mast Cell Stabilizers
Lodoxamide and Bufrolin as the First High and Equipotent Agonists of
Human and Rat GPR35. Mol. Pharmacol. 2014, 85, 91−104.
(21) Jenkins, L.; Brea, J.; Smith, N. J.; Hudson, B. D.; Reilly, G.;
Bryant, N. J.; Castro, M.; Loza, M. I.; Milligan, G. Identification of novel
species-selective agonists of the G-protein-coupled receptor GPR35
that promote recruitment of beta-arrestin-2 and activate G(alpha 13).
Biochem. J. 2010, 432, 451−459.
(22) Deng, H.; Fang, Y. Discovery of nitrophenols as GPR35 agonists.
Med. Chem. Comm. 2012, 3, 1270−1274.
(23) Zhao, H.; Neamati, N.; Hong, H.; Mazumder, A.; Wang, S.;
Sunder, S.; Milne, G. W.; Pommier, Y.; Burke, T. R. Coumarin-based
inhibitors of HIV integrase. J. Med. Chem. 1997, 40, 242−249.
(24) Sashidhara, K. V.; Kumar, M.; Khedgikar, V.; Kushwaha, P.;
Modukuri, R. K.; Kumar, A.; Gautam, J.; Singh, D.; Sridhar, B.; Trivedi,
R. Discovery of Coumarin-Dihydropyridine Hybrids as Bone Anabolic
Agents. J. Med. Chem. 2013, 56, 109−122.
(25) Fringuelli, F.; Piermatti, O.; Pizzo, F. One-pot synthesis of 3-
carboxycoumarins via consecutive knoevenagel and pinner reactions in
water. Synthesis-Stuttgart 2003, 2331−2334.
́
(26) Tena Perez, V.; Fuentes de Arriba, A. L.; Monleón, L. M.; Simón,
L.; Rubio, O. H.; Sanz, F.; Morán, J. R. A High Yield Procedure for the
Preparation of 2-Hydroxynitrostyrenes: Synthesis of Imines and
Tetracyclic 1,3-Benzoxazines. Eur. J. Org. Chem. 2014, 2014, 3242−
3248.
(27) Deng, H.; Hu, H.; Fang, Y. Tyrphostin analogs are GPR35
agonists. FEBS Lett. 2011, 585, 1957−1962.
(28) Hall, C. M.; Wright, J. B.; Johnson, H. G.; Taylor, A. J. Quinoline
Derivatives as Antiallergy Agents 2. Fused-Ring Quinaldic Acids. J. Med.
Chem. 1977, 20, 1337−1343.
(29) Frey, J.; Bond, A. D.; Holmes, A. B. Improved synthesis of
dithieno[3,2-b: 2 ’,3 ’-d] thiophene (DTT) and derivatives for cross
coupling. Chem. Commun. 2002, 2424−2425.
(30) Anthony, C. Pyrroloquinoline quinone (PQQ) and quinoprotein
enzymes. Antioxid. Redox. Sign. 2001, 3, 757−774.
(31) Kasahara, T.; Kato, T. A new redox-cofactor vitamin for
mammals. Nature 2003, 422, 832−832.
(32) He, K.; Nukada, H.; Urakami, T.; Murphy, M. P. Antioxidant and
pro-oxidant properties of pyrroloquinoline quinone (PQQ): implica-
tions for its function in biological systems. Biochem. Pharmacol. 2003,
65, 67−74.
(33) Klinman, J. P.; Bonnot, F. Intrigues and Intricacies of the
Biosynthetic Pathways for the Enzymatic Quinocofactors: PQQ, TTQ,
CTQ, TPQ, and LTQ. Chem. Rev. 2014, 114, 4343−4365.
(34) Berni, E.; Dolain, C.; Kauffmann, B.; Léger, J. M.; Zhan, C.; Huc,
I. Expanding the registry of aromatic amide foldamers: Folding,
photochemistry and assembly using diaza-anthracene units. J. Org.
Chem. 2008, 73, 2687−2694.
2647
https://dx.doi.org/10.1021/acs.jmedchem.0c01624
J. Med. Chem. 2021, 64, 2634−2647