The design, synthesis and evaluation of 2-aminobenzoxazole analogues as potent and orally efficacious ChemR23 inhibitors

Article abstract of DOI:10.1016/j.bmc.2020.115622

We previously reported 2-aminobenzoxazole analogue 1 as a potent ChemR23 inhibitor. The compound showed inhibitory activity against chemerin-induced calcium signaling through ChemR23 internalization in CAL-1 cells, which are cell lines of plasmacytoid dendric cells (pDCs). Furthermore, compound 2 inhibited chemotaxis of CAL-1 triggered by chemerin in vitro. However, we noted a difference in the ChemR23 response to our inhibitor between rodents and non-rodents in a previous study. To address this issue, we performed optimization of ChemR23 inhibitors using CAL-1 cells endogenously expressing human ChemR23 and conducted a pharmacokinetics study in cynomolgus monkeys. Various substituents at the 4-position of the benzoxazole ring exhibited potent in vitro bioactivity, while those at the 6-position were not tolerated. Among substituents, a carboxyl group was identified as key for improving the oral bioavailability in cynomolgus monkeys. Compound 38a with the acidic part changed from a tetrazole group to a 1,2,4-oxadiazol-5-one group to improve bioactivity and pharmacokinetic parameters exhibited inhibitory activity against chemerin-induced chemotaxis in vitro. In addition, we confirmed the ChemR23 internalization of pDCs by compound 38a orally administered to cynomolgus monkeys. These 2-aminobenzoxazole-based ChemR23 inhibitors may be useful as novel immunotherapeutic agents capable of suppressing the migration of pDCs, which are known to be major producers of type I interferons in the lesion area of certain autoimmune diseases, such as systemic lupus erythematosus and psoriasis.

Full text of DOI:10.1016/j.bmc.2020.115622

Journal Pre-proofs
The design, synthesis and evaluation of 2-aminobenzoxazole analogues as po‐
tent and orally efficacious ChemR23 inhibitors
Takamichi Imaizumi, Shigeki Otsubo, Masato Komai, Hidenori Takada,
Michihiro Maemoto, Atsuko Kobayashi, Nobumasa Otsubo
PII:
DOI:
Reference:
S0968-0896(20)30452-1
https://doi.org/10.1016/j.bmc.2020.115622
BMC 115622
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Accepted Date:
9 June 2020
26 June 2020
Please cite this article as: Imaizumi, T., Otsubo, S., Komai, M., Takada, H., Maemoto, M., Kobayashi, A.,
Otsubo, N., The design, synthesis and evaluation of 2-aminobenzoxazole analogues as potent and orally
efficacious ChemR23 inhibitors, Bioorganic & Medicinal Chemistry (2020), doi: https://doi.org/10.1016/j.bmc.
2020.115622
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Elsevier Ltd. All rights reserved.

Title:
The design, synthesis and evaluation of 2-aminobenzoxazole analogues as
potent and orally efficacious ChemR23 inhibitors
Author:
*a
c
Takamichi Imaizumi , Shigeki Otsubo b, Masato Komai b, Hidenori Takada ,
Michihiro Maemoto a, Atsuko Kobayashi a Nobumasa Otsubo a
,
a Small Molecule Drug Research Laboratories, b Immunology & Allergy Research
c
Laboratories, Pharmacokinetics Research Laboratories, R&D Division, Kyowa
Kirin Co., Ltd., 1188 Shimotogari, Nagaizumi-cho, Sunto-gun, Shizuoka, 411-
8731, Japan
E-mail: takamichi.imaizumi.6p@kyowakirin.com
Keywords:
SLE, pDC, Interferon, ChemR23, Internalization
1

Abstract
We previously reported 2-aminobenzoxazole analogue 1 as a potent ChemR23
inhibitor. The compound showed inhibitory activity against chemerin-induced
calcium signaling through ChemR23 internalization in CAL-1 cells, which are cell
lines of plasmacytoid dendric cells (pDCs). Furthermore, compound 2 inhibited
chemotaxis of CAL-1 triggered by chemerin in vitro. However, we noted a
difference in the ChemR23 response to our inhibitor between rodents and non-
rodents in a previous study. To address this issue, we performed optimization of
ChemR23 inhibitors using CAL-1 cells endogenously expressing human
ChemR23 and conducted a pharmacokinetics study in cynomolgus monkeys.
Various substituents at the 4-position of the benzoxazole ring exhibited potent in
vitro bioactivity, while those at the 6-position were not tolerated. Among
substituents, a carboxyl group was identified as key for improving the oral
bioavailability in cynomolgus monkeys. Compound 38a with the acidic part
changed from a tetrazole group to a 1,2,4-oxadiazol-5-one group to improve
bioactivity and pharmacokinetic parameters exhibited inhibitory activity against
chemerin-induced chemotaxis in vitro. In addition, we confirmed the ChemR23
internalization of pDCs by compound 38a orally administered to cynomolgus
monkeys. These 2-aminobenzoxazole-based ChemR23 inhibitors may be useful
as novel immunotherapeutic agents capable of suppressing the migration of
pDCs, which are known to be major producers of type I interferons in the lesion
area of certain autoimmune diseases, such as systemic lupus erythematosus and
psoriasis.
2

1. Introduction
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by
systemic inflammation and production of autoantibodies.1,2 Although belimumab,
a human monoclonal antibody that inhibits B-cell activating factor (BAFF), was
approved as a new drug for SLE in 20113,4, the treatment of SLE still largely
depends on steroids and immunosuppressive agents.5 There have been many
reports describing a relationship between SLE development and type-I interferon
(IFN). Indeed, the anti-IFN-α monoclonal antibody sifalimumab was developed
and met the primary endpoint at the phase II stage, although the development
was discontinued due to issues from a business strategic perspective.6 In addition,
anifrolumab, a monoclonal antibody against the IFN receptor, achieved the
primary endpoint in a phase III study for SLE patients recently.7 The positive
results with these biologics have driven the development of new drugs targeting
IFN.
Plasmacytoid dendritic cells (pDCs) are specialized DCs that can produce
large amounts of type I IFN through the activation of Toll-like receptors.8-12 The
molecular mechanism underlying the function and production of pDCs has been
discussed in recent reviews, with the inactivation of pDCs considered an
attractive approach to treating SLE.13-15 While the IFN-targeted antibodies
mentioned above would be useful for treating SLE the application of biologics
remains limited to certain patient populations due to the administration route or
for financial reasons. Therefore, a small-molecular drug with equivalent efficacy
to the abovementioned biologics is expected to enjoy widespread usage.
From this perspective, an orally available small-molecular agent that can
3

control IFN production or pDC activation would meet a number of unmet medical
needs. ChemR23 is a G-protein coupled receptor (GPCR) and is also known as
chemoattractant-like receptor 1 (CMKLR1). This receptor is expressed on
17
macrophages16, natural killer cells and pDCs.
Chemerin, which is an
inflammation-related protein, was discovered by Wittarmer at al. as a natural
ligand for ChemR23.18,19 Anti-ChemR23 antibody has been reported to suppress
pDC migration induced by chemerin from psoriatic skin fibroblasts. 20-22
In our previous paper23, compound 1 inhibited chemerin-induced
intracellular calcium fluctuation through the internalization of ChemR23 in the
pDC line CAL-1.24 In addition, compound 2, in which a methyl amide group was
introduced at the 5-position of the benzoxazole ring, showed an inhibitory effect
against CAL-1 migration induced by chemerin in a concentration-dependent
manner.24 However, differences in the inhibitory concentration between CAL-1
cells and Ba/F3 cells expressing murine ChemR23 were confirmed for several
compounds, and we faced difficulties optimizing our ChemR23 inhibitors for both
rodent and non-rodent models (Figure 1).23 The chemotaxis inhibition of CAL-1
by our ChemR23 inhibitors is presumably based on ChemR23 internalization,23
and we confirmed ChemR23 internalization in pDCs from cynomolgus monkey
by our compounds as well as chemerin. Therefore, we decided to carry out a
further structure activity relationship (SAR) study and pharmacokinetic screening
focusing on non-rodent species.
We herein report the synthesis and in vitro and in vivo activity of 2-
aminobenzoxazole analogues that are orally bioavailable ChemR23 inhibitors in
cynomolgus monkeys. Given our previously obtained SAR information showing
4

that modifications at the 5-position of benzoxazole did not markedly improve the
inhibitory activity against ChemR23 signaling24, we first performed further
optimization at the 4- and 6-positions of benzoxazole. We then focused on the
biphenyl tetrazole moiety to obtain good oral availability in cynomolgus monkeys.
5

2. Chemistry
Commercially available (R)-1-(4-bromophenyl)ethanamine (6) purchased from
Alfa Aesar was selected as the starting material for all compounds. For the
modifications of the 4- and 6-position on benzoxazole, the methyl esters 5a and
5b were prepared by a cyclization reaction of aminophenols 4a and 4b using
potassium ethylxthanthate. After reductive amination of 6 with 1-butanal, the
obtained secondary amine 7 was used for a coupling reaction with 2-
chlorobenzoxazoles prepared from the 2-mercaptobenzoxazoles 5a and 5b by
thionyl chloride to obtain compounds 8a and 8b, respectively. Basic hydrolysis of
the ester group in compounds 8a and 8b produced carboxylic acids 9a and 9b,
respectively. Amidation was then performed using 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide hydrochloride (EDCI) as a condensation reagent. Finally,
Suzuki coupling reaction of the intermediate 10a with 2-cyanophenyl borate
followed by tetrazole formation reaction in the presence of TMSN3 and dibutyltin
oxide provided the methylamide 11a. Commercially available 2-(tetrazol-5-
yl)phenylboronic acid was used to obtain amides 11a’ and 11b more easily than
11a and with good chemical yields (Scheme 1).
Further modifications of the 4-position of benzoxazole are outlined in
Scheme 2. The carboxylic acid 13 was synthesized from the intermediate 8a in
three steps (Suzuki coupling, hydrolysis and tetrazole formation), and
subsequent amidation with the appropriate amines afforded amides 14a and 14b.
The primary alcohol 16 was formed by the hydride reduction of the ester 8a and
following introduction of a phenyltetrazole moiety. The methylether 17 was
obtained via methylation of the alcohol 15 with iodomethane. Esterification of the
6

carboxylic acid 13 followed by nucleophilic addition using methylmagnesium
bromide gave the gem-dimethyl alcohol 18.
The syntheses of target compounds possessing amine, amide, urea,
sulfonamide or acylsulfonamide groups are summarized in Scheme 3. Curtius
rearrangement and subsequent hydrolysis reaction afforded amine 19. Amino
compound 20 was prepared by palladium coupling reaction with the
corresponding boronic acid. The N-acetyl compound 21, N-sulfonyl compound 22
and N,N-dimethyl urea 23 were prepared using the corresponding chloride or
anhydride. The acylsulfonamide 24 was prepared by the activation of the carboxyl
group as a carbonyl chloride followed by substitution with methanesulfonamide.
A series of aromatic ring-substituted analogues is described in Scheme 4.
In order to avoid conflicts between two bromo groups, the bromide 7 was
converted to the biphenyl analogue 27 by a coupling reaction prior to nucleophilic
substitution reaction with the bromo-containing benzoxazole 26. Replacement
from the bromo moiety to the various heteroaryl ring was performed as follows:
In brief, the 4-pyridine 29a and 3-pyridine 29b were synthesized via Suzuki-
Miyaura coupling with boronic acid. The 2-pyridine 29c, N-methylpyrrole 29d and
N-methylimidazole 29e were prepared through a Stille reaction. 1-pyrrazole was
introduced by an Ullmann type reaction, and these cyano analogues were
converted into the desired tetrazole analogues 30a–f.
Modification of the external benzene ring of the biphenyl moiety is shown
in Scheme 5. After the conversion of bromide 3123 into pinacolborate 32, various
arylbromides were treated in the presence of palladium catalyst. The desired
analogues 33a–o were prepared by tin-mediated tetrazole formation reaction.
7

The synthetic approach for the 1,2,4-oxadiazol-5-one analogues
possessing a carboxylic acid at the 4-position on benzoxazole is outlined in
Scheme 6. Along with n-butyl-branched compound 8a, cyclopropanemethyl-
branched benzoxazole 35 was prepared, since the cyclopropanemethyl moiety
was found to be a substituent showing equivalent in vitro activity with reduced
lipophilicity. The consecutive coupling reactions were easily accomplished and
afforded biphenyl compounds 36a–d in good yields. The introduction of a
hydroxyl amine unit to the cyano moiety was more difficult than expected because
the 50% hydroxylamine solution reacted with the ester moiety on benzoxazole.25
Eventually, hydroxylamine hydrochloride pretreated with sodium bicarbonate was
used for N-hydroxyamidine formation at the cyano group.26 The subsequent
cyclization reaction with 1,1'-carbonyldiimidazole (CDI) afforded 1,2,4-oxadiazol-
5-ones 37a–d. Finally, the desired carboxylic acids 38a–d were easily obtained
by the hydrolysis of the methyl ester moiety. Analogues with n-butyl side chains
41a and 41b were also prepared in the same way as above.
8

3. Result and discussion
In the calcium assay using cell lines expressing human or mouse ChemR23, it
was recognized that the half maximal inhibitory concentration (IC50) values of
some benzoxazole analogues were not correlated between human and mouse
receptors.23 For this reason, the in vitro activity was optimized using CAL-1 cells,
a ChemR23-expressing pDC-like line established from a patient with blastic
natural killer cell lymphoma.24 For the in vivo study, we planned to use
cynomolgus monkeys because of the confirmed high homology and no species
difference between their ChemR23 and human ChemR23.
First, we compared the inhibitory activities of a series of N-methylamide
analogues (2, 11a and 11b) to figure out the best position for the benzoxazole
modification. The data indicated that modification of the 4- or 5-position showed
better tolerability than that of the 6-position of the benzoxazole ring (2: 4-position,
IC50 = 110 nM; 11a: 5-position, IC50 = 38 nM; and 11b: 6-position, IC50 > 3000
nM). Since we had previously confirmed the low tolerability for conversion of the
5-position, we decided to start exploration at the 4-position.23
The results of a SAR study at the 4-position on the benzoxazole ring are
summarized in Table 1. The dimethylamide 11a’ (IC50 = 13 nM) showed an
improved potency (by approximately 8-fold) compared with the methylamide 11a
or unsubstituted compound 1. In addition, a terminal hydroxyl group in the amide
chain was well tolerated with regard to the in vitro activity (14b: IC50 = 26 nM)
compared to the des-hydroxy analogue (14a: IC50 = 62 nM). The primary hydroxyl
analogue 16 (IC50 = 24 nM) showed an improved in vitro potency, whereas the
gem-dimethylated tertiary alcohol 18 (IC50 = 340 nM) showed a decreased
9

activity. Since the methoxymethyl group retained strong activity (17: IC50 = 29
nM), the loss of potency of compound 18 was attributed to the bulkiness of the
gem-dimethyl moiety. Acidic functional groups, such as carboxylic acid and its
bioisostere acylsulfonamide, were found to be substituents that increased the
potency by 6- to 8-fold compared with the unsubstituted compound 1 (13: IC50 =
17 nM, 24: IC50 = 14 nM). Considering the strong activity of amino analogue 20
(IC50 = 27 nM), N-acetyl 21 (IC50 = 22 nM), N,N-dimethylurea 22 (IC50 = 32 nM)
and N-methanesulfonyl 23 (IC50 = 14 nM), the 4-position of benzoxazole was
expected to have broad tolerability. Among the 6-membered heteroaryl groups,
the 2-pyridyl 30c (IC50 = 110 nM) showed favorable potency over 4-pyridyl 30a
(IC50 = 300 nM) and 3-pyridyl 30b (IC50 > 1000 nM). Among the 5-membered
heteroaryl groups, N-methylpyrrole 30d (IC50 = 130 nM) and pyrazole 30f (IC50 =
330 nM) showed no improvement in the activity, but the N-methylimidazole 30e
(IC50 = 10 nM) exhibited the highest in vitro inhibitory activity among the
modifications of the 4-position on the benzoxazole ring.
The results of the modifications at the external benzene ring are shown in
Table 2. Regarding the conversion of the 3-position, fluorine 33a retained
potency, whereas the potency of chloro 33b was attenuated 3-fold. It was
revealed that the activity was significantly impaired by modification at the 4-
position by trifluoromethyl 33c, chloro 33d, methyl 33d or methoxy 33e, although
fluoro 33c (IC50 = 57 nM) increased the potency slightly. No substituent was
identified at the 6-position that could increase the activity. At the 5-position,
although trifluoromethyl 33i and methoxy 33l were not tolerated, the potencies of
fluoro 33h (IC50 = 75 nM), chloro 33j (IC50 = 75 nM) and methyl 33k (IC50 = 53
10

nM) were increased (Table 2). These results were combined with those of other
SAR studies as described below.
Pharmacokinetic studies after the oral administration of our ChemR23 inhibitors
were conducted by cassette administration to cynomolgus monkeys, and the
results are summarized in Table 3. Although most compounds, including the
initial lead compound 1, showed a poor oral absorption in cynomolgus monkeys,
the carboxylic acid 13 and N-methylimidazolyl compound 30e gave better results
than any other compounds regarding the Cmax value and bioavailability (F), which
motivated us to embark on further modifications. We also paid special attention
to carboxylic acid 13 because the structural features closely resembled those of
Candesartan, a well-known drug for hypertension. Candesartan possesses the
same two acidic moieties (carboxylic acid moiety and tetrazole moiety) and shows
low bioavailability. Researchers in Takeda's group modified its carboxylic acid
part to an ester part by utilizing a prodrug strategy and then developed
Candesartan cilexetil which is capable of improving the oral absorption of
Candesartan. 27-29 In addition, Azilsartan, which has a similar chemical structure
to Candesartan, has been launched for the same therapeutic indication without a
prodrug formation.30 The oral absorption of Azilsartan was improved over that of
Candesartan by the conversion of the tetrazole ring into a 1,2,4-oxadiazol-5-one
ring (Figure 2). Although each of these strategies was attractive for application
to our molecules to improve the bioavailability, we selected the methodology used
for Azilsaltan for our optimization study because the in vivo pharmcokinetic (PK)-
pharamcodynamics (PD) results were expected to be easier to understand than
with the prodrug strategy.
11

The in vitro potency and PK parameters of 1,2,4-oxadiazol-5-one
analogues are summarized in Table 4. The PK parameters were obtained by oral
cassette administration to cynomolgus monkeys at a dose of 1.0 mg/kg for each
compound. To our delight, all of the prepared 1,2,4-oxadiazol-5-one analogues
exhibited an improved oral availability. In addition, the effect of substituents on
the external phenyl ring was also observed, as had been confirmed in the
tetrazole analogs, and excellent potency was noted (IC50 = 1.2 nM and IC50 = 1.3
nM for chloro-substituted analogs 41b and 38c). Among these 1,2,4-oxadiazol-
5-one analogues, compounds 38a and 41b, which possessed a good balance
between in vitro potency and bioavailability, were selected for the additional PK
study after oral administration alone with a higher dose (10 mg/kg). As a result,
both compounds showed good oral absorption (38a: Cmax = 6950 ng/mL, 41b:
Cmax = 2110 ng/mL), although the half-life still had room for improvement (38a:
T1/2 = 1.0 hr, 41b: T1/2 = 2.0 hr). (Table 5). We also evaluated these two
compounds using a chemotaxis assay with CAL-1 cells and confirmed that both
compounds inhibited the migration of CAL-1 triggered by chemerin from a low
concentration (38a: chemotaxis IC50 = 63 nM, 41b: chemotaxis IC50 = 235 nM,
Figure 3). The slight decrease in the potency against the chemotaxis assay from
the calcium signaling assay was attributed to a difference in the free fraction ratio
of the compounds, as this chemotaxis assay contains 0.1% (w/v) bovine serum
albumin in the assay buffer. We determined that compound 38a was preferable
to 41b for further ex vivo and in vivo studies in monkeys in terms of both the in
vitro potency versus free plasma concentrations (38a: (fu.p・Cmax)/IC50 = 1.0, 41b:
(fu.p・Cmax)/IC50 = 0.5, Table 5) and chemotaxis inhibition potency.
12

The effective concentration range of the selected compound 38a for in vivo
ChemR23 internalization in cynomolgus monkeys was predicted beforehand
using cynomolgus monkey whole blood. The internalization rate of ChemR23 on
pDCs was calculated by a flow cytometry analysis using fluorescence-conjugated
ChemR23 antibody after compound addition to the blood. The receptor
internalization rate of 53% (N = 2, average) was observed at 30 μM of the
compound, and a rate of 82% (N = 2, average) was observed at 100 μM (Figure
4). Based on these results, along with the monkey PK results of 10 mg/kg (oral
administration [p.o.]), we estimated that the oral administration of 50 mg/kg could
cause sufficient receptor internalization of ChemR23 by pDCs at the time of
reaching Cmax. The results of the in vivo test are shown in Figure 5, and the PK
(plasma concentration)-PD (receptor internalization rate) correlation is
summarized in Table 6. The internalization rate of ChemR23 on pDCs was
measured at 2, 4 and 8 h after the oral administration of the test compound. The
maximum pharmacological effect was confirmed at 4 h, and the internalization
rate was 90% and 74% in each monkey (Figure 5). Plasma concentrations of
compound 38a were also measured at each time point, and it was confirmed that
each animal had the maximum compound concentration at 4 h (Table 6). Based
on in vitro experiments with CAL-1, we assumed that the internalization of
ChemR23 would indicate the inhibition of pDC migration. Therefore, the reduction
of pDCs in the tissue is expected to inhibit IFN-α production at the inflamed site,
leading to the improvement of autoimmune diseases, particularly SLE or psoriasis.
However, further structural optimization is desired to not only improve the short
13

half-life in vivo, but also to reduce the required dosage Optimization studies
aimed at confirming pharmacological effects are ongoing.
14

4. Conclusion
In this study, we identified a benzoxazole analogue capable of reducing the
ChemR23 expression on monkey pDCs in vivo. We initially modified the 4-
position on the benzoxazole ring as a promising position to improve the potency
and identified the carboxylic acid analogue 13, which showed potent activity and
slightly improved oral bioavailability in cynomolgus monkeys. To further improve
the potency and pharmacokinetic parameters, we embarked on the conversion
of the tetrazole group as well as the modification of the external benzene ring and
ultimately discovered the promising 1,2,4-oxadiazole-5-one analogue 38a, which
was designed based on a previous study regarding Azilsaltan given its structural
similarity to compound 13. Compound 38a showed high potency in both the
calcium assay and chemotaxis assay and preferable pharmacokinetics and ex
vivo effects in cynomolgus monkeys. Furthermore, compound 38a showed an in
vivo ChemR23 internalization effect on pDCs after its oral administration to
cynomolgus monkeys. We expect that our ChemR23 inhibitors will suppress pDC
migration into inflamed tissues and IFN production in vivo. A further study to
elucidate the chemerin/ChemR23 axis in animal models using our compound will
be reported in due course.
15

N
N
NH
N
O
Me
N
2
Compound
N
Me
N
O
N
N
H
NH
human ChemR23 (CAL-1) IC50 (nM) = 38
mouse ChemR23 (Ba/F3) IC50 (nM) = 150
N
Me
N
Me
N
O
N
N
N
NH
Me
Me
N
H
N
3
Compound
1
Compound
N
O
Me
human ChemR23 (CAL-1) IC50 (nM) = 120
mouse ChemR23 (Ba/F3) IC50 (nM) > 10000
O
human ChemR23 (CAL-1) IC50 (nM) = 110
mouse ChemR23 (Ba/F3) IC50 (nM) = 150
Me
Figure 1. Summary of previous report.
O
O
4
3
MeO
MeO
NH2
OH
N
O
b
a
SH
5
6
5a :
5b :
4-CO2Me : 95%
6-CO2Me : 88%
4a
4b
: 3-CO2Me
: 5-CO2Me
R
R
Me
N
Me
4
Me
HN
Br
Br
N
O
N
O
Me
Br
c
d
e
N
Br
6
H2N
85%
Me
Me
Me
6
7
8a
8b
9a
9b
: R = 4-CO2Me : 46%
: R = 6-CO2Me : 55%
: R = 4-CO2H : quant.
: R = 6-CO2H : 75%
N
N
NH
N
R
R
Me
N
Me
N
11a
11a', 11b ;
For
For
; g, h
Br
N
O
N
O
i
f
Me
Me
10a
10a'
10b
11a
11a'
11b
: R = 4-CONHMe : 99%
: R = 4-CONMe2 : 95%
:R = 6-CONHMe : 92%
: R = 4-CONHMe : 14%
: R = 4-CONMe2 : 73%
: R = 6-CONHMe : 96%
Scheme 1. Reagents and conditions: (a) potassium ethylxanthate, pyridine,
90 °C; (b) DMF, SOCl2, reflux; (c) 1-butanal, NaBH(OAc)3, MeOH, rt; (d) DIPEA,
DMF, 50 °C; (e) LiOH·H2O, MeOH, 60 °C; (f) amines, THF, DIPEA, EDCI·HCl,
HOBt·H2O, DMF, rt; (g) 2-cyanophenylboronic acid pinacol ester, Pd(PPh3)4,
KOAc, DME, 85 °C; (h) TMSN3, dibutyltin oxide, toluene, reflux; (i) 2-(tetrazol-5-
yl)phenylboronic acid, Pd(PPh3)4, Cs2CO3, DMF, H2O, 100 °C.
16

N
N
N
NH
N
R
NH
NC
N
HO
O
HO
O
N
N
O
Me
N
Me
N
Me
Me
N
c
a, b
d
N
O
N
O
N
O
8a
31%
97%
Me
Me
Me
: R = H : 50%
12
13
14a
14b
: R = OH : 30%
N
g, h
9.8%
N
N
NH
N
NH
HO
N
HO
Me
N
HO
Me
Me
N
e
Me Me
Br
N
O
a, c
N
O
N
N
O
N
73%
10%
Me
Me
N
Me
15
16
N
18
NH
N
MeO
f, a, c
27%
Me
N
N
O
Me
17
Scheme 2. Reagents and conditions: (a) 2-cyanophenylboronic acid pinacol
ester, Pd(PPh3)4 (for 12, 16), Pd(dppf)Cl2 (for 17), K3PO4·7H2O (for 12, 17), KOAc
(for 16), DME, 85 °C; (b) LiOH·H2O, MeOH, 60 °C; (c) TMSN3, dibutyltin oxide,
toluene, reflux; (d) amines, HATU, DIPEA, DMF, rt; (e) LiAlH4, THF, 0 °C; (f) MeI,
NaH, DMF, 0 °C; (g) H2SO4, MeOH, reflux; (h) MeMgBr, THF, rt.
N
N
NH
N
NH2
NH2
Me
N
Me
N
b
a
N
O
Br
N
O
9a
13%
75%
Me
Me
19
20
N
N
Me
NH
f, g, b
10%
N
O
NH
Me
c, b
N
N
26%
O
Me
S
N
Me
N
O
O
NH
21
N
HN
O
Me
N
N
Me
O
Me
NH
N
N
N
O
N
NH
d, b
Me
N
Me
N
O
24
20%
Me
22
N
N
N
Me
S
NH
O
e, d
NH
Me
N
O
N
O
23%
Me
23
17

Scheme 3. Reagents and conditions: (a) DPPA, Et3N, DMF, H2O, 100 °C; (b) 2-
(tetrazol-5-yl)phenylboronic acid, Pd(PPh3)4, Cs2CO3, DMF, H2O, 100 °C; (c)
Ac2O, Et3N, CH2Cl2, 0 °C; (d) Me2NC(O)Cl, Et3N, CH2Cl2, reflux; (e) MsCl, Et3N,
CH2Cl2, 0 °C; (f) (COCl)2, DMF, CH2Cl2, rt; (g) MeSO2NH2, DIPEA, DMF, 100 °C.
Br
Br
NH2
OH
N
O
b
a
SH
83%
26
25
NC
NC
NC
R
Br
Me
N
Me
N
Me
HN
Me
HN
N
O
d
N
O
e or f or g
c
Br
59%
21%
Me
Me
Me
27
Me
28
29a-f
29a
29b
29c
29d
29e
29e
: 62%
: 82%
: quant.
: 85%
: 20%
: 25%
7
N
N
N
NH
N
N
R
N
N
N
N
N
Me
N
Me
Me
R =
N
h
N
O
30a
63%
30b
93%
30c
55%
30d
67%
30e
29%
30f
90%
Me
30a-f
Scheme 4. Reagents and conditions: (a) potassium ethylxanthate, MeOH, reflux;
(b) DMF, SOCl2, reflux; (c) 2-cyanophenylboronic acid, Pd(PPh3)4, Cs2CO3, DMF,
H2O, 85 °C; (d) DIPEA, 1,4-dioxane, reflux; (e) 3-pyridylboronic acid or 4-
pyridylboronic acid, Pd(PPh3)4, Cs2CO3, DMF, H2O, 100 °C (for 29a and 29b), (f)
2-(tributylstannyl)pyridine or 1-methyl-2-(tributylstannyl)pyrrole or 1-methyl-2-
(tributylstannyl)imidazole, Pd(PPh3)4, toluene, 100 °C (for 29c–e), (g) pyrazole
CuI, Cs2CO3, DMF, 130 °C (for 29f); (h) TMSN3, dibutyltin oxide, toluene, reflux.
18

N
N
NH
N
3
5
Me
Me
Me
N
Me
N
O
O
Me
N
a
Br
N
O
b, c
B
N
O
N
O
Me
Me
R
95%
4.5-37%
Me
Me
Me
32
33a-o
31
33c : 4-F
33h : 5-F
33m: 6-F
33a : 3-F
33d : 4-CF3
33e : 4-Cl
33f : 4-Me
33g : 4-OMe
33i : 5-CF3
33j : 5-Cl
33k : 5-Me
33l : 5-OMe
33n : 6-CF3
33o : 6-Me
33b : 3-Cl
Scheme 5. Reagents and conditions: (a) bis(pinacolato)diboron, KOAc,
Pd(PPh3)4, DME, reflux; (b) ArBr, Pd(PPh3)4, K3PO4 ･ 7H2O, DME, reflux; (c)
TMSN3, dibutyltin oxide, toluene, reflux.
MeO
O
NC
MeO
O
Me
HN
Me
N
36a-c
For
For
; c, d
Me
N
Me
b
a
Br
Br
N
O
36d
; e
N
O
Br
36a
36b
36c
36d
: R = Me : 85%
: R = F : 85%
: R = Cl : 65%
: R = H : 81%
H2N
quant.
80%
R
34
6
35
36a-d
O
O
O
O
NH
NH
N
HO
O
N
MeO
O
Me
Me
N
h
f, g
N
O
N
O
N
R
R
38a
38b
38c
38d
: R = Me : 81%
: R = F : 30%
: R = Cl : 10%
: R = H : 62%
37a
37b
37c
37d
: R = Me : 40%
: R = F : 38%
: R = Cl : 43%
: R = H : 45%
38a-d
37a-d
O
O
O
O
NH
NH
N
NC
N
HO
O
MeO
O
MeO
O
Me
Me
N
c, d
Me
f, g
h
N
O
N
O
N
O
8a
N
N
R
R
R
39a 39b
40a 40b
41a 41b
,
,
,
Me
Me
Me
39a
39b
40a
40b
41a
41b
: R = F : 82%
: R = Cl : 66%
: R = F : 29%
: R = Cl : 48%
: R = Fl : 58%
: R = Cl : 68%
Scheme 6. Reagents and conditions: (a) c-PrCHO, NaBH4, MeOH rt; (b) 4-
methoxycabonyl-2-chlorobenzoxazole prepared from 5a, DIPEA, DMF, 60 °C; (c)
bis(pinacolato)diboron, Pd(dppf)Cl2, KOAc, 1,4-dioxane, 100 °C; (d) ArBr,
Pd(PPh3)4, Cs2CO3, 1,4-dioxane, H2O, 100 °C; (e) 2-cyanophenylboronic acid
pinacol ester, Pd(PPh3)4, K3PO4·7H2O, DME, 85 °C; (f) NH2OH･HCl, NaHCO3,
DMSO, 80 °C; (g) CDI, THF, 75 °C; (h) 3 M NaOH, EtOH, 60 °C.
19

Table 1. The structure–activity relationship at the 4-position of the benzoxazole
ring
N
N
NH
N
R
Me
N
N
O
Me
a
a
a
hChemR23 IC50
(nM)
hChemR23 IC50
hChemR23 IC50
(nM)
Compd.
R
Compd.
R
Compd.
R
(nM)
N
HO
H
1
110
110
13
16
30a
24
300
>1000
110
Me
HN
Me
O
O
HO
N
11a
11a'
14a
14b
13
Me
18
17
20
21
22
23
30b
30c
30d
30e
30f
340
29
27
22
32
14
Me
N
Me
MeO
N
N
Me
Me
O
O
N
N
NH2
NH
62
130
Me
HO
Me
Me
N
N
N
26
O
Me
10
O
HO
Me
O
Me
NH
N
17
330
N
Me
S
O
O
O
O
HN
Me
S
24
14
NH
O
ahuman ChemR23 IC50: 50% inhibitory concentration against chemerin-induced
calcium signaling (all values are the mean of four measurements).
20

Table 2. The structure–activity relationship at the external phenyl ring
N
N
NH
N
3
Me
N
4
N
O
R
6
5
Me
a
a
Compd.
R
hChemR23 IC50
(nM)
Compd.
R
hChemR23 IC50
(nM)
75
1
H
3-F
110
33h
33i
5-F
5-CF3
5-Cl
33a
33b
33c
33d
33e
33f
100
1320
65
3-Cl
350
33j
4-F
57
33k
33l
5-Me
5-OMe
6-F
53
4-CF3
4-Cl
>3000
>3000
2100
860
160
1300
210
33m
33n
33o
4-Me
4-OMe
6-CF3
6-Me
33g
3000
ahuman ChemR23 IC50: 50% inhibitory concentration against chemerin-induced
calcium signaling (all values are the mean of four measurements).
21

Table 3. Pharmacokinetic properties of tetrazole analogues in cynomolgus
monkeys
N
N
NH
N
R
Me
N
N
O
Me
a
a
a
b
Compd.
R
Cmax
T1/2
AUC0-∞
fu.p
Fa,c
(%)
(ng/mL・h)
(ng/mL) (h)
(%)
1
11a
11a’
16
-H
12.8
9.74
27.5
9.68
115
22.6
64
3.2
1.3
3.2
1.6
3.7
2.7
5.9
3.9
2.2
4.7
44.3
25.7
104
20
0.0713 1.8
0.0975 NTd
0.411 6.7
0.112 0.8
0.0285 8.2
0.103 NTd
0.129 NTd
0.299 7.5
0.102 NTd
0.0203 NTd
-CONHMe
-CONMe2
-CH2OH
13
-CO2H
429
22.2
248
152
41.6
73.8
24
-CONHSO2Me
-2-pyridyl
30c
30e
21
-2-Me-imidazole
-NHCOMe
-NHSO2Me
79.9
13.2
26.5
23
aThe values shown as the means of three determinations by cassette dosing (1.0
mg/kg, p.o.).
bPlasma unbound fraction obtained from an equilibrium dialysis method.
cCompounds were administered at doses of 1.0 mg/kg (p.o.) and 0.1 mg/kg (iv)
using a cassette dosing method.
dNot tested.
22

(A)
(B)
O
Me
O
O
O
O
N
N
HO
N
NH
HO
N
O
N
O
N
HO
N
N
O
O
N
N
Me
O
O
NH
Me
N
N
N
NH
N
NH
This work
N
Me
Candesartan cilexetil (Blopress)
O
N
O
N
N
O
Me
Me
Me
Candesartan
HO
compound 13
O
O
O
N
NH
N
N
O
Me
Azilsartan (Azilva)
Figure 2. (A) The chemical structures of Candesartan, Candesartan cilexetil and
Azilsartan. (B) Optimization strategy of compound 13 to improve the oral
bioavailability in this study.
23

Table 4. Pharmacokinetic properties in cynomolgus monkeys and ChemR23 in
vitro inhibitory activity of 1,2,4-oxadiazol-5-one analogues
O
O
NH
N
HO
O
Me
N
N
O
R1
R2
a
a
a
b
d
Compd.
R1
R2 Cmax
T1/2
AUC0-∞
fu.p
%
F a,c IC50
(ng/mL・h)
(ng/mL) (h)
(%) (nM)
41a
41b
38a
38b
38c
38d
n-Bu-
n-Bu-
F
610
361
2.2
2.0
2.8
4.1
0.8
2.6
1630
1040
4830
4410
1300
11600
0.00736 NTe 2.7
0.0135 42 1.2
0.0230 58 3.2
0.00977 82 2.3
0.0108 NTe 1.3
Cl
c-PrCH2- Me 1490
c-PrCH2-
F
1490
518
c-PrCH2- Cl
c-PrCH2- H
3830
0.00851 60
11
aThe values shown as the means of three determinations by cassette dosing (1.0
mg/kg, po).
bPlasma unbound fraction obtained from an equilibrium dialysis method.
cCompounds were administered at doses of 1.0 mg/kg (po) and 0.1 mg/kg (iv)
using a cassette dosing method.
dhuman ChemR23 IC50: 50% inhibitory concentration against chemerin-induced
calcium signaling (all values are the mean of four measurements).
eNot tested.
24

Table 5. Pharmacokinetic parameters of compounds 38a and 41b in cynomolgus
monkey plasma following 10 mg/kg oral administration
a–c
(fu.p・Cmax
)
MRT
a
a
a
a
Compd. Dose
Cmax
C8h
T1/2
AUC0-∞
a
/IC50
(mg/kg) (ng/mL) (ng/mL) (h) (ng/mL*h)
(h)
2.9
4.1
38a
41b
10
10
6950
2110
206
NDd
1.0
2.0
22200
7090
1.0
0.5
aThe values shown as the means of three determinations.
bPlasma unbound fraction obtained from an equilibrium dialysis method.
chuman ChemR23 IC50: 50% inhibitory concentration against chemerin-induced
calcium signaling (All values are the mean of four measurements).
dNot detected.
25

Figure 3. Effects of compound 38a and compound 41b against chemerin-
induced chemotaxis using CAL-1 cells. Each point represents the mean ± SD of
triplicate experiments.
26

Figure 4. Effects of compound 38a and chemerin on ChemR23 expression in
whole blood of cynomolgus monkeys (N=1 experiment, each individual).
27

Figure 5. Expression of ChemR23 on the surface of pDCs after oral dosing of
compound 38a (50 mg/kg) to cynomolgus monkeys.
28

Table 6. In vivo PK (plasma concentration)-PD (receptor internalization rate)
correlation after oral dosing of compound 38a (50 mg/kg) to cynomolgus monkeys
Time
Plasma Concentration
ChemR23 Expression
(hr)
2
4
8
2
(µmol/L)
178
(%)
55
10
44
30
26
60
Monkey No.1
Monkey No.2
204
179
231
259
4
8
30.4
29