Inhibitory activities of anthraquinone and xanthone derivatives against transthyretin amyloidogenesis

Article abstract of DOI:10.1016/j.bmc.2021.116292

Transthyretin is a tetrameric protein which functions as a transporter of thyroxine and retinol-binding protein. Misfolding and amyloid aggregation of transthyretin are known to cause wild-type and hereditary transthyretin amyloidosis. Stabilization of the transthyretin tetramer by low molecular weight compounds is an efficacious strategy to inhibit the aggregation pathway in the amyloidosis. Here, we investigated the inhibitory activities of anthraquinone and xanthone derivatives against amyloid aggregation, and found that xanthone-2-carboxylic acid with one chlorine or methyl group has strong inhibitory activity comparable with that of diflunisal, which is one of the best known stabilizers of transthyretin. X-ray crystallographic structures of transthyretin in complex with the compounds revealed that the introduction of chlorine, which is buried in a hydrophobic region, is important for the strong inhibitory effect of the stabilizer against amyloidogenesis. An in vitro absorption, distribution, metabolism and elimination (ADME) study and in vivo pharmacokinetic study demonstrated that the compounds have drug-like features, suggesting that they have potential as therapeutic agents to stabilize transthyretin.

Full text of DOI:10.1016/j.bmc.2021.116292

Bioorg. Med. Chem. 44 (2021) 116292
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Inhibitory activities of anthraquinone and xanthone derivatives against
transthyretin amyloidogenesis
Ryota Kitakamia, Kishin Inuib, Yusuke Nakagawab, Yurika Sawaic, Wakana Katayamad,
,
Takeshi Yokoyamad, Takuya Okadac e, Kayoko Kanamitsuf, Shinsaku Nakagawag,
,
,*
Naoki Toyookac e, Mineyuki Mizuguchid
a Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama 930-0194, Japan
b Graduate School of Science and Engineering, University of Toyama, Toyama 930-8555, Japan
c Faculty of Engineering, University of Toyama, Toyama 930-8555, Japan
d Faculty of Pharmaceutical Sciences, University of Toyama, Toyama 930-0914, Japan
e Graduate School of Innovative Life Science, University of Toyama, Toyama 930-8555, Japan
f Drug Discovery Initiative, the University of Tokyo, Tokyo 113-0033, Japan
g Graduate School of Pharmaceutical Sciences, Osaka University, Osaka 565-0871, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Transthyretin is a tetrameric protein which functions as a transporter of thyroxine and retinol-binding protein.
Misfolding and amyloid aggregation of transthyretin are known to cause wild-type and hereditary transthyretin
amyloidosis. Stabilization of the transthyretin tetramer by low molecular weight compounds is an efficacious
strategy to inhibit the aggregation pathway in the amyloidosis. Here, we investigated the inhibitory activities of
anthraquinone and xanthone derivatives against amyloid aggregation, and found that xanthone-2-carboxylic acid
with one chlorine or methyl group has strong inhibitory activity comparable with that of diflunisal, which is one
of the best known stabilizers of transthyretin. X-ray crystallographic structures of transthyretin in complex with
the compounds revealed that the introduction of chlorine, which is buried in a hydrophobic region, is important
for the strong inhibitory effect of the stabilizer against amyloidogenesis. An in vitro absorption, distribution,
metabolism and elimination (ADME) study and in vivo pharmacokinetic study demonstrated that the compounds
have drug-like features, suggesting that they have potential as therapeutic agents to stabilize transthyretin.
Transthyretin
Stabilizer
X-ray crystallography
Anthraquinone
Xanthone
1. Introduction
TTR does not form amyloid fibrils when the tetramer in the native
state is maintained. However, TTR aggregates into amyloid fibrils when
Amyloidosis is a group of diseases which are caused by deposition of
abnormal protein aggregates known as amyloid fibrils in tissues or or-
gans. Amyloid fibrils are built up from amyloidogenic precursor proteins
and more than 30 amyloidogenic proteins are currently known.1
Transthyretin (TTR), previously called prealbumin, is one of the
amyloidogenic proteins. TTR forms a homotetramer with a molecular
weight of 55 kDa under physiological conditions and its subunit
monomer is composed of 127 amino acid residues. TTR is a serum
protein mainly synthesized in the liver, but it is also produced in the
choroid plexus of the brain, the pigmented epithelium of the retina, and
the islets of Langerhans. TTR circulates in serum at a concentration of
it is destabilized due to amyloidogenic amino acid mutations and/or in
vitro under acidic conditions. Wild-type TTR also possesses an inherent
amyloidogenic potential, which causes wild-type TTR amyloidosis, a
condition previously called senile systemic amyloidosis.6 In the amyloid
fibril formation of TTR, the tetramer dissociation occurs first, then the
monomer unfolds into an amyloidogenic intermediate, which self-
aggregates to form amyloid fibrils. The tetramer dissociation is a rate-
limiting step of the amyloid fibril formation of TTR.3,4 So far, more
than 130 mutations of TTR that cause hereditary TTR amyloidosis have
been reported. Among them, V30M is the best studied mutation, since it
is the most frequent amyloidogenic mutation worldwide and there are
large foci of hereditary V30M-TTR amyloidosis patients in Portugal,
Sweden, and Japan.7,8
3.1–7.6
μM, and plays a role in transporting thyroid hormone thyroxine
(T4) and retinol-loaded retinol-binding protein.2–5
* Corresponding author.
E-mail address: mineyuki@pha.u-toyama.ac.jp (M. Mizuguchi).
https://doi.org/10.1016/j.bmc.2021.116292
Received 14 December 2020; Received in revised form 17 June 2021; Accepted 21 June 2021
Available online 26 June 2021
0968-0896/© 2021 Elsevier Ltd. All rights reserved.

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Liver transplantation is a well-established treatment for hereditary
TTR amyloidosis, since it replaces the amyloidogenic TTR variant in
serum with wild-type TTR. However, there is a great need for develop-
ment of a less invasive treatment. One alternative treatment for liver
transplantation is the stabilization of the TTR tetramer by small
molecules, such as tafamidis and diflunisal. These compounds bind to
the two T4-binding channels located at the dimer-dimer interface of
TTR. The binding of the compounds stabilizes the TTR tetramer, which
results in the inhibition of the tetramer dissociation and the amyloid
fibril formation of TTR.3,4,9,10 Treatment with tafamidis has been shown
Fig. 1. Chemical structures of anthraquinone derivatives (1–27), xanthone derivatives (28–42), and positive controls (tafamidis and diflunisal). The chemical
structure of γ-mangostin is also shown for comparison.
2

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
to delay the neuropathic progression in patients with familial amyloid
polyneuropathy caused by the V30M mutated TTR.11,12 Tafamidis is the
first pharmacotherapy approved for the treatment of neurologic
impairment in familial amyloid polyneuropathy. It is also approved for
the treatment of cardiomyopathy caused by amyloid deposition of TTR
in the heart. Diflunisal has also been shown to reduce the rate of pro-
gression of neurological impairment and to preserve quality of life in
clinical trials.13 In addition, tolcapone and AG10 are currently under-
going clinical trials for the treatment of TTR amyloidosis.14–16
inhibiting amyloid aggregation of wild-type TTR.33 In addition, Maia
et al. showed that xanthone derivatives isolated from Calophyllum teys-
mannii var. inophylloide bind to the T4-binding channels of TTR.35 We
also showed that γ-mangostine, a xanthone derivative found in the
pericarp of mangosteen, inhibits the amyloid fibril formation of V30M-
TTR in vitro.37 To date, however, there has been no detailed analysis of
the relationship between the chemical structure of anthraquinone and
xanthone derivatives and their inhibitory activity against amyloid fibril
formation. Here, we report detailed comprehensive investigations on the
structure–activity relationship of anthraquinone and xanthone de-
rivatives (Fig. 1). We found that xanthone-2-carboxylic acid with a
chlorine or methyl group at position 4 has a strong amyloid inhibitory
activity comparable with that of the known inhibitor diflunisal
(Figure 2). High-resolution crystal structures of TTR in complex with
anthraquinone and xanthone derivatives revealed the details of how
these compounds bind to the T4-binding channels at the dimer-dimer
interface of TTR. In addition, we performed an in vitro ADME and in
vivo pharmacokinetic analysis of the compounds and demonstrated that
they have drug-like properties. Our results suggest that these com-
pounds have potential as therapeutic agents to stabilize TTR in the
blood.
Liver transplantation and orally administrable TTR kinetic stabilizers
(tafamidis and diflunisal) target the native TTR circulating in the blood,
which is mainly synthesized in the liver. In addition to the liver, how-
ever, smaller amounts of TTR are synthesized in the retina of the eye, the
choroid plexus of the brain, and the
α cells of the pancreatic islets. In
fact, ocular amyloidosis and cerebral amyloid angiopathy accompanied
by leptomeningeal amyloidosis are observed in patients with several
specific TTR gene mutations.17–21 Moreover, variant TTR deposition in
the eye does not stop after liver transplantation due to TTR biosynthesis
in the retinal pigment epithelium.18,22,23 Central nervous system
symptoms due to meningeal and/or cerebrovascular amyloidosis can
occur in patients who have survived for a long time after liver trans-
plantation.22,24 In addition, tafamidis is unable to halt the progression of
the disease in patients with advanced polyneuropathy caused by V30M-
TTR.25 Therefore, there is a great need for a new drug for TTR
amyloidosis, and the development of such a drug is still being actively
pursued.15,26–30 In order to promote this effort, it may be useful to
develop small molecule inhibitors with scaffolds different from tafami-
dis and diflunisal.
2. Results and discussion
2.1. Inhibitory activities of anthraquinone derivatives
We examined the inhibitory effect of commercially available
anthraquinone derivatives (1–17) on the amyloid fibril formation of
V30M-TTR. The V30M variant was used in this study, since it is the most
frequent amyloidogenic mutation causing hereditary TTR amyloid-
osis.7,8 Amyloid fibril formation was initiated by a pH-jump from 7.0 to
Anthraquinone and xanthone derivatives are a class of tricyclic
organic compounds that are present in various natural sources,
including plants, bacteria and fungi. Anthraquinone has two keto groups
at the central ring, while only one keto group exists in xanthone. These
compounds have been widely applied in industry and are expected to be
novel therapeutic agents against various diseases such as cancer, infec-
tious diseases, and amyloidosis.31–38 In a previous study, Baures et al.
reported that anthraquinone-2-carboxylic acid is modestly effective in
4.6 at a tetramer concentration of 10
μM in the presence of 0–100 μM
compounds. After incubation for 96 h at 37 ◦C, the amyloid aggregation
was quantified by thioflavin T fluorescence.37,39,40 Amyloid inhibitory
activity was assessed by determining the compound concentration for
50% inhibition of amyloid aggregation (IC50
) (Table 1 and
Fig. 2. Structure of compound 40. (a) Chemical structure of compound 40. (b) The overall structure of WT-TTR in complex with 40. (c) Omit difference Fourier map
for 40. Contoured at 3.3 (orange mesh) and 2.5 (blue mesh).
σ
σ
3

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Table 1
Inhibitory activities of anthraquinone and xanthone derivatives. The IC50 values were calculated from the inhibition ratio at 0–100
μM concentration.
Compd No.
IC50 (µM)
CLogP
1
>100
>100
3.74
4.24
3.62
4.19
2.70
3.53
3.12
3.28
2.66
2.92
0.69
0.78
3.74
2.67
3.74
2.13
0.84
3.26
3.56
2.73
3.76
3.04
3.76
5.76
3.99
3.76
3.83
3.48
3.48
2.85
2.85
3.57
3.57
3.57
3.57
3.57
3.57
3.57
3.57
3.60
4.31
3.34
5.69
4.39
5.00
2
3
>100
4
>100
5
>100
6
12 ± 0.15
12 ± 1.7
>100
7
8
9
12 ± 0.23
>100
10
11
>100
12
11 ± 0.15
>100
13
14
>100
15
>100
16
5.4 ± 0.28
>100
17
18
28 ± 2.6
50 ± 12
10 ± 0.09
45 ± 10
>100
19
20
21
22
23
13 ± 0.40
21 ± 2.2
12 ± 0.49
24 ± 1.9
22 ± 1.0
>100
24
25
26
27
28
29
>100
30
22 ± 0.24
19 ± 0.70
10 ± 0.19
10 ± 0.45
13 ± 0.09
7.7 ± 0.34
13 ± 0.51
>100
31
32
33
34
35
36
37
38
12 ± 0.27
13 ± 0.08
6.7 ± 0.26
9.1 ± 0.01
6.5 ± 0.20
10.0 ± 0.3
6.6 ± 0.14
5.3 ± 0.60
39
40
41
42
γ-Mangostin
Diflunisal
Tafamidis
Supplementary Fig. 1). Compounds 6, 7, 9, and 12 showed amyloid
inhibitory activity with IC50 values of 11–12 µM. Compound 16 showed
the strongest amyloid inhibitory activity among 1–17, and its IC50 value
was 5.4 ± 0.28 µM (Fig. 1 and Table 1). Compound 16 is diacerein,
which is used to treat the symptoms of osteoarthritis. Diacerein is a
prodrug and its active metabolite is rhein (compound 6). The IC50 value
of 6 is 12 ± 0.15 µM, and thus its amyloid inhibitory activity is not
much weaker than that of diacerein (Fig. 1 and Table 1). However, there
is a possibility that the use of diacerein may cause side effects such as
diarrhea and hepatobiliary disorders, although it has an acceptable
safety profile.41 Therefore, we explored the possibility of developing
other anthraquinone derivatives as potential inhibitors of TTR
amyloidosis.
Next, we synthesized the anthraquinone derivatives (18–27). Syn-
theses of anthraquinone derivatives 18 and 21–26 are shown in Scheme
1.
The anthraquinone skeleton (A1–A5) was constructed from the
appropriate phthalic anhydride and toluene derivative in two steps of
Friedel-Crafts reaction followed by the cyclization process.42 The
oxidation of the methyl group on the anthraquinone skeleton with
chromium trioxide in an acetic acid (Method A)43 or with manganese
dioxide in a concentrated sulfuric acid (Method B)44 afforded the
Scheme 1. Syntheses of anthraquinone derivatives 18 and 21–26.
4

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Scheme 2. Syntheses of anthraquinone derivatives 19–20 and 27.
Scheme 3. Syntheses of xanthone derivatives 28–42.
anthraquinone derivatives 18 and 21–26. On the other hand, oxygen-
ated anthraquinone derivatives 19 and 27 were synthesized by the
known anthraquinone 20,45 which was prepared from 1-acetoxy-6-
methylanthraquinone as shown in Scheme 2. The acetyl group of 20
was hydrolyzed with sodium carbonate solution to give the anthraqui-
none 19. Moreover, anthraquinone 20 was transformed into the corre-
sponding ethyl ether in two steps, followed by hydrolysis of the resulting
ethyl ester with sodium hydroxide to obtain the anthraquinone 27.
The IC50 values of 18–27 are summarized in Table 1. A comparison of
the IC50 values of 16, 18, and 20 indicates that acetoxy groups at both
the 4 and 5 positions are important for the amyloid inhibitory activity of
16 (Fig. 1 and Table 1). From the IC50 values for compounds 18, 19, 20,
and 27, the compound with an acetoxy group at position 5 (20) has
higher inhibitory activity than the compounds with hydrogen (18),
hydroxyl (19), and ethoxy (27) groups at position 5 (Fig. 1 and Table 1).
Anthraquinone-2-carboxylic acid (compound 18) showed an IC50
value of 28 ± 2.6 µM, which was significantly higher than that of the
positive controls (diflunisal and tafamidis) (Table 1). This is consistent
with a previous study which showed that anthraquinone-2-carboxylic
acid is only modestly effective in inhibiting amyloid fibril formation of
wild-type TTR in vitro.33 A comparison of 21 and 23 shows that the
carboxy group at position 2 is important for the inhibitory activity.
Compound 22, which has carboxy groups at both positions 2 and 4,
showed little inhibitory activity (Fig. 1 and Table 1).
xanthone, both show IC50 values of approximately 20 µM, while com-
pounds 28 and 29, in which methyl groups are introduced at the same
position, do not inhibit amyloid fibril formation (Fig. 1 and Table 1).
These results indicate that the carboxy group at position 2 or 3 in
xanthone is important for the amyloid inhibitory activity. Compound 40
is an analogue of 25, and the comparison of their IC50 values suggests
that the xanthone derivatives have stronger inhibitory activity than the
corresponding anthraquinone derivatives (Fig. 1 and Table 1). Similarly,
compound 42 is a xanthone derivative of 23, and its inhibitory activity is
higher than that of 23. Among the xanthone derivatives in this study, 40
and 42, in which a carboxyl group is introduced at position 2 and a
chlorine or a methyl group is introduced at position 4, exhibited the
strongest inhibitory activities, which were comparable with that of
diflunisal (Fig. 1 and Table 1). We confirmed the binding of 40 and 42 to
V30M-TTR using isothermal titration calorimetry (ITC): the binding
affinities of 40 and 42 were lower than that of tafamidis, but were
comparable with that of diflunisal (Supplementary Fig. 2).
2.3. Binding modes of anthraquinone and xanthone derivatives
To elucidate the detailed binding mechanisms of the anthraquinone
and xanthone derivatives, we determined the crystal structures of TTR
(wild-type and/or V30M) in complex with 7, 25, 32 and 40 at
1.15–1.45 Å resolutions (Supplementary Tables 1 and 2). The root mean
square deviation (RMSD) for the backbone
α atoms when superposing
2.2. Inhibitory activities of xanthone derivatives
32-wt to 32-V30M was 0.21 Å, and the RMSD between 40-wt and 40-
V30M was 0.25 Å. The superposition also indicated that the positions of
the compounds in the T4-binding channels are completely identical with
each other. Therefore, the V30M mutation does not affect the binding
mode of the compounds in the T4-binding channels.
We synthesized the xanthone derivatives (28–42) and investigated
their amyloid inhibitory activity (Fig. 1 and Table 1). Xanthone de-
rivatives 28–42 were synthesized from the appropriate o-nitro-
benzaldehyde and a phenol derivative in the presence of CuCl2, Ph3P,
and K3PO4 in toluene at 100 ◦C, as shown in Scheme 3.46,47
In all complexes, the electron densities of the compounds were
clearly observed in the T4-binding channels (Fig. 2), as has been
observed for typical stabilizers in previous studies.3,4,9,10 The carboxyl
Compounds 30 and 31, in which carboxyl groups are introduced into
5

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Fig. 3. Close-up view of the T4-binding channel of wild-type TTR in complex with 25 (left) and 40 (right). Oxygen, nitrogen, and chlorine atoms are colored red, blue
and dark grey, respectively. The Leu17 side-chain in subunit C is omitted to show the compounds.
groups of 25, 32 and 40 were close to Lys15 at the outer cavity,
corroborating the importance of the electrostatic interaction with Lys15
(Figs. 2–4). This is consistent with the results of the amyloid inhibitory
activity of 28 and 30. Compound 30, which has a carboxyl group at
position 2, showed amyloid inhibitory activity, while compound 28, in
which the 2-COOH is replaced with a methyl group, showed no amyloid
inhibitory activity (Fig. 1 and Table 1). A similar interaction has been
observed in the crystal structures of TTR in complex with flufenamic
acid, diflunisal, and tafamidis.48–50 This favorable electrostatic inter-
action is important for the inhibition of the tetramer dissociation that is
the rate-limiting step in the amyloid fibril formation of TTR.50
inhibitory activity: compound 40 with 4-Cl has stronger amyloid
inhibitory activity than 30, which has a hydrogen atom at position 4
(Fig. 1 and Table 1). The CLogP values of 30 and 40 are 2.85 and 3.60,
respectively (Table 1). This indicates that the hydrophobically driven
interaction between TTR and 40 mainly contributes to the strong
inhibitory activity of 40. Compounds 36, 38, and 39, which are isomers
of 40, have similar amyloid inhibitory activity (Table 1). This indicates
that a chlorine atom can be accommodated at positions 5, 7, and 8
(Figs. 1 and 3). On the other hand, the chlorine atom at position 4 in
compound 25 interacts with a region formed by Lys15, Leu17, Ala108,
Leu110 and T119 (Fig. 3). This interaction is important for the amyloid
inhibitory activity of 25. The amyloid inhibitory activity of 25 with 4-Cl
is stronger than that of 18 without 4-Cl (Fig. 1 and Table 1). Compound
22, in which 4-Cl is replaced with COOH, has no inhibitory activity,
probably because the charged group cannot be accommodated in this
hydrophobic space (Fig. 3). From the IC50 values of 18, 22, 23, 25, and
26, the compounds with a methyl group (23) or chlorine (25) at position
4 have higher inhibitory activity than those with a hydrogen (18), car-
boxy group (22), or fluorine (26) at position 4 (Fig. 1 and Table 1).
Comparing the CLogP values among 18, 22, 23, 25, and 26, the com-
pound with a relatively large CLogP value tends to have a strong
inhibitory activity. This corroborates that the hydrophobic interaction at
position 4 is important for the amyloid inhibitory activity. Comparing
the IC50 values of 24 and 25, compound 25, which has one chlorine at
Compound 25 is an anthraquinone derivative, while 40 is a xanthone
derivative, i.e., the only difference between 25 and 40 is located at
position 10 (Fig. 1 and Table 1). However, the binding modes in the
crystal structures of these two compounds are different (Fig. 3). Com-
pound 25 in the 25-TTR complex cannot adopt the binding mode
observed in the 40-TTR complex, probably due to the steric hindrance of
the carbonyl oxygen at position 10 in the 25-TTR complex. The ether
oxygen at position 10 of 40 in the 40-TTR complex is located near the
T119 side chain (Fig. 3). The chlorine atom at position 4 of 40 is buried
in a hydrophobic region formed by Leu17, Thr106, Ala108, Thr119, and
Val121. There is no water molecule in the region around the chlorine
atom and these side chains (Fig. 3). The interaction between 4-Cl of 40
and the surrounding hydrophobic region is important for the amyloid
Fig. 4. Close-up view of the T4-binding channel of V30M-TTR in complex with 7 (left) and wild-type TTR in complex with 32 (right). Oxygen, nitrogen, and chlorine
atoms are colored red, blue, and dark grey, respectively. The hydrogen-bond and salt-bridge are indicated by dashed lines. The red sphere is a water molecule.
6

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
the carboxyl group of 32 (Fig. 4). Moreover, 8-Cl of 32 is surrounded by
the amino acid residues Leu110, Ser117, and Thr119, which are
included in a region called halogen-binding pocket 3.51 The 8-Cl of 32
contributes to the amyloid inhibitory activity, since the IC50 values of 31
and 32 are 19
μM and 10 μM, respectively (Fig. 1 and Table 1). On the
other hand, compounds 32, 33, 34, and 35 have similar levels of amy-
loid inhibitory activity (Table 1). This indicates that a chlorine atom can
be accommodated at positions 5–8 in the ring without a carboxyl group
(Figs. 1 and 4). Interestingly, the effect of the introduction of a chlorine
atom at position 6 is different between 30 and 31: compound 34 has
amyloid inhibitory activity, while 37 has no inhibitory activity (Fig. 1
and Table 1). This indicates that the carboxyl group of 34 cannot be
replaced from position 3 to position 2 (Figs. 1 and 4).
A previous study has revealed that iododiflunisal binds strongly to
the T4-binding site of TTR and stabilizes the TTR tetramer better than
diflunisal.51 Previous crystallographic analysis has revealed that the
iodine atom of iododiflunisal is located at the hydrophobic region,
which is composed of Leu17, Thr106, Ala108, Thr119, and Val121.51
The position of the chlorine atom in the complex of 40-TTR was
consistent with the position of the iodine atom in the complex of iodo-
diflunisal and TTR (PDB code: 1Y1D): the distance between the iodine
and chlorine atoms after alignment was only 0.4 Å (Fig. 5). These data
suggest that the introduction of halogen substituents such as chlorine or
iodine, which can be oriented to this hydrophobic space, significantly
enhances the inhibitory activity of the stabilizers against amyloid
aggregation.
Comparison of the binding mode between 40 and tafamidis revealed
that the chlorine atom in 40 does not fit with those of tafamidis (Fig. 5).
The chlorine of 40 is located near the entrance of the T4-binding
channel. On the other hand, the two chlorine atoms of tafamidis are
buried in the two symmetrical hydrophobic regions called halogen-
binding pockets 3 and 3′, which are located in the inner channel.50,51
The chlorine atom of 32 is also located in halogen-binding pocket 3
(Fig. 4). The removal of two chlorine atoms from tafamidis resulted in a
substantial decrease of amyloid inhibitory activity.50,52 Our results also
indicated that the chlorine atom is important for the inhibitory activity
of 32 (Fig. 1 and Table 1).
We have previously reported the X-ray crystal structure of TTR in
complex with γ-mangostin, which is one of the xanthone derivatives.37
The 2-dimethylallyl group of γ-mangostin is located near the entrance of
the T4-binding channel (Fig. 1), and it interacts with the side chain of
V121 via hydrophobic interaction.37 The 8-dimethylallyl group is
located in the channel, and it interacts with the side chains of A108,
L110, S117 and T119. The 6-hydroxyl group undergoes water-mediated
interaction with the side chains of Thr116 and Val121 (Fig. 1).37
A
chloride ion-mediated interaction takes place between the 7-hydroxyl
group and the side chains of Ser117 and Thr119 in the structure of
TTR in complex with γ-mangostin.37 Because of the lack of these
dimethylallyl and hydroxyl groups in 40, the binding direction of 40 is
quite different from that of γ-mangostin (Fig. 5).
Fig. 5. A comparison of the binding direction of 40 with the known TTR sta-
bilizers. The compound 40 is colored grey. Iododiflunisal (1Y1D), tafamidis
(6E72), and γ-mangostin (4Y9E) are colored yellow. The chloride atoms are
colored black. Oxygen, nitrogen, fluorine, and iodine are colored red, blue, light
blue, and purple, respectively.
2.4. In vitro ADME assay and in vivo pharmacokinetic study
An in vitro distribution, metabolism and elimination (ADME) assay is
a standard method to evaluate the key physicochemical properties that
are associated with the efficacy profile of a candidate compound. We
performed a standard ADME assay of compounds 40 and 42, which have
the highest amyloid inhibitory activity among the xanthone derivatives
(Table 2). The results were compared with those for tafamidis, which is
the first approved drug for the treatment of familial amyloidotic
polyneuropathy.50
position 4, has higher inhibitory activity than 24, which has multiple
chlorines at positions 5, 6, 7, and 8 (Figs. 1 and 3 and Table 1).
We also determined the structures of TTR in complex with com-
pounds 7 and 32. In the 7-TTR complex, the distance between the ox-
ygen of the 2-OH of 7 and Lys15-Nζ near the entrance of the T4-binding
channel is 2.4 Å, suggesting the formation of a hydrogen bond. The
importance of the 2-OH of 7 was also shown by the amyloid inhibitory
activity of 7 and 13. Compound 13, which lacks the 2-OH of 7, showed
no amyloid inhibitory activity (Fig. 1 and Table 1). In the 32-TTR
complex, Lys15-Nζ is close to the carboxyl group of 32, which also
suggests the formation of a salt bridge. One water molecule is trapped by
The solubility was evaluated using Japanese Pharmacopoeia (JP) 1st
test fluid and JP 2nd test fluid, and the results showed that the three
compounds—40, 42 and tafamidis—were comparable. A parallel arti-
ficial membrane permeability assay (PAMPA) showed that 40 and 42
have good permeability properties. A PAMPA permeability assessment
7

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
has also been used to evaluate antipyrine and metoprolol, revealing that
the two drugs have 100% and 95% gastrointestinal absorption in
humans, respectively,53 at values of 13 and 1.0 × 10ꢀ 6 cm/sec. Since
40, 42 and tafamidis are more permeable than antipyrine, it was
inferred that all three agents were 100% absorbed in the gastrointestinal
tract. In the metabolic stability test using rat liver microsomes, the re-
sidual rate after 30 min at 0.2 mg protein/mL was stable at around 88%
for all three compounds, with and without the addition of NADPH,
which is a coenzyme of CYP. The plasma protein-binding ratio of 40 is
lower than that of tafamidis, while 42 showed a plasma protein-binding
ratio similar to that of tafamidis. These results indicate that 40 and 42
have drug-like physicochemical properties that are similar to those of
tafamidis (Table 2).
Tokyo Chemical Industry (Tokyo), and Kanto Chemical (Tokyo) and
used without further purification. Column chromatography was done on
Cica silica gel 60 N (spherical, neutral; particle size, 63–210 nm; Kanto
Chemical), while thin-layer chromatography was performed using
Merck silica gel 60F254 plates. Melting points were taken on a Yanaco-
micro melting point apparatus and are uncorrected. The nuclear mag-
netic resonance (NMR) spectra were acquired in the specified solvent by
means of a JEOL JNM-A400 (400 and 100 MHz for 1H and 13C,
respectively) or JEOL JNM-ECX500 (500 and 125 MHz for 1H and 13C,
respectively). The chemical shifts (δ) are reported in ppm downfield
from TMS, and coupling constants (J) are expressed in Hertz. IR spectra
were measured with a JASCO FT/IR-460 Plus spectrophotometer
(JASCO, Tokyo). The low-resolution mass spectra and high-resolution
mass spectra were obtained with a Shimadzu GCMS-QP 500 (Shi-
madzu, Kyoto, Japan), JEOL D-200, or JEOL AX505 (JEOL, Tokyo) mass
spectrometer in the electron impact mode at the ionization potential of
70 eV.
We also performed in vivo pharmacokinetics studies in rats using a
liquid chromatography- tandem mass spectrometry (LC-MS/MS)
method (Table 3). Compound solutions were injected intravenously (i.
v.) or orally (p.o.) to male SD rats (n = 2). Blood samples were collected
from the caudal vein at 5 (i.v.), 15, and 30 min and 1, 2, 4, and 8 h. The
compound concentrations were determined by LC-MS/MS analysis of
the supernatants of plasma samples. From the analysis of oral (p.o.)
administration, the observed maximum peak plasma concentration was
reached at approximately 2 h for 40. On the other hand, compound 42
showed a maximum peak plasma concentration at approximately
0.25 h. The data of intravenous (i.v.) administration showed that the
mean elimination half-lives were 1.7 h and 2.4 h for 40 and 42,
respectively, and both these half-lives were shorter than that of tafa-
midis (12 h). The estimated bioavailability (BA) of 40 was 69.8%,
which is comparable with that of tafamidis. On the other hand, the BA of
42 is 34.9%, which is lower than that of 40 (Table 3). The main meta-
bolic pathway of tafamidis is the formation of glucuronide conjugates
for carboxyl groups (https://pfizerpro.jp/documents/if/vyn/vyn01if.
pdf). Differences in metabolism other than CYP, such as differences in
glucuronidation, may have contributed to the differences in the in vivo
pharmacokinetic profiles of these three compounds.
4.1.2. General procedure for construction of the anthraquinone skeleton
To a stirred solution of phthalic anhydride (290 mg, 1.96 mmol) in
toluene (5 mL) was added anhydrous AlCl3 (522 mg, 3.92 mmol) at
0 ◦C. The reaction mixture was stirred at 50 ◦C for 2 h. The reaction was
quenched with water (5 mL), and the organic layer was separated. The
aqueous layer was extracted with EtOAc (3 mL × 3). The organic layer
and extracts were combined, dried over Na2SO4, and evaporated to give
a white solid, which was used directly in the next step. The solution of
the above solid in concentrated H2SO4 (2.0 mL) was heated to 100 ◦C
for 1 h. After cooling, the reaction mixture was diluted with EtOAc
(3 mL), and the organic layer was separated and washed with water
(1 mL × 3). The organic layer was then dried over Na2SO4 and
evaporated to give a pale yellow solid, which was recrystallized from
EtOH to afford the corresponding anthraquinone A1–A5 as a pale yellow
solid.42
4.1.2.1. 2-Methylanthracene-9,10-dione (A1). Yield: 71%; 1H NMR
(400 MHz, CDCl3) δ: 8.32–8.29 (2H, m), 8.21 (1H, d, J = 8.0 Hz), 8.11
(1H, s), 7.80–7.78 (2H, m), 7.60 (1H, dd, J = 8.0, 1.2 Hz), 2.54 (3H,
s).54
After the final collection of the blood samples, the compound con-
centrations in the eyeball and the brain were determined by LC-MS/MS
(Supplementary Table 3). Considering that the dose of tafamidis is one-
tenth of those of 40 and 42, the concentrations of 40 and 42 in the
eyeball and brain were lower than those of tafamidis. We therefore
concluded that it would be difficult to use 40 and 42 as therapeutics
targeting TTR in the eye or brain. However, compounds 40 and 42 have
drug-like properties similar to those of tafamidis (Tables 2 and 3), and
thus they may have potential as therapeutic agents to stabilize TTR in
the blood and thereby ameliorate TTR amyloidosis.
1
4.1.2.2. 1,3-Dimethylanthracene-9,10-dione (A2). Yield: 76%; H NMR
(400 MHz, CDCl3) δ: 8.26 (1H, dd, J = 7.2, 2.0 Hz), 8.26 (1H, dd,
J = 7.2, 2.0 Hz), 8.05 (1H, s), 7.78 (1H, td, J = 7.2, 2.0 Hz), 7.74
(1H, td, J = 7.2, 2.0 Hz), 7.39 (1H, s), 2.83 (3H, s), 2.48 (3H, s).55
4.1.2.3. 1,2,3,4-Tetrachloro-6-methylanthracene-9,10-dione (A3). Yield:
15%; 1H NMR (400 MHz, CDCl3) δ: 8.06 (1H, d, J = 8.0 Hz), 7.95 (1H,
s), 7.60 (1H, d, J = 8.0 Hz), 2.54 (3H, s).56
3. Conclusions
In this study, we investigated the amyloid inhibitory activities of
anthraquinone and xanthone derivatives, and found that xanthone-2-
carboxylic acid with a chlorine or methyl group at position 4 (40 or
42) has the highest amyloid inhibitory activity in vitro. X-ray crystallo-
graphic analysis of the 40-TTR complex indicated the importance of the
interaction of chlorine with the hydrophobic region formed by Leu17,
Thr106, Ala108, Thr119, and Val121. An in vitro ADME assay and an in
vivo pharmacokinetic analysis suggested that 40 and 42 have drug-like
properties, although their elimination half-lives are shorter than that
of tafamidis. Our results suggest that these compounds have potential as
therapeutic agents to stabilize TTR in the blood.
4.1.2.4. 1-Chloro-3-methylanthracene-9,10-dione (A4). Yield: 9%; 1H
NMR (400 MHz, CDCl3) δ: 8.30 (1H, dd, J = 8.0, 1.6 Hz), 8.26 (1H,
dd, J = 7.2, 2.0 Hz), 8.12 (1H, d, J = 1.2 Hz), 7.84–7.76 (2H, m),
7.63 (1H, s), 2.51 (3H, s); 13C NMR (100 MHz, CDCl3) δ: 182.41,
181.49, 145.13, 138.14, 135.61, 134.99, 134.44, 133.67, 132.34,
127.43, 127.34, 127.11, 126.69, 21.42; IR (KBr): 3952, 2363, 1676,
1591, 1306, 1275, 1246 cmꢀ 1; mp: 196–197 ◦C; MS (EI) m/z: 256 (M+);
HRMS (EI) m/z: 256.0282 (M+) [calcd for C15H9ClO2: 256.0291].
4.1.2.5. 1-Fluoro-3-methylanthracene-9,10-dione (A5). Yield: 2%; 1H
NMR (400 MHz, CDCl3) δ: 8.31–8.27 (2H, m), 7.98 (1H, s), 7.84–7.61
(2H, m), 7.30 (1H, d, J = 11.6 Hz), 2.54 (3H, s).57
4. Materials and methods
4.1. Chemistry
4.1.3. General procedure for the synthesis of anthraquinone derivatives 18
and 21–26
4.1.1. General
Chemicals were purchased from Sigma-Aldrich, Merck (Darmstadt,
Germany), FUJIFILM Wako Chemicals (Osaka, JAPAN), Nacalai Tesque,
4.1.3.1. Method A. To a stirred solution of the anthraquinone (A1–A5,
8

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Table 2
Summary of in vitro ADME properties of 40 and 42.
Compound
Aqueous Solubility (
μM)
PAMPA Permeability (×10ꢀ 6 cm/s)
Hepatic Microsome Stability (mL/min/kg)
Plasma Protein Binding (% unbound)
pH 1.2
pH 6.8
40
2.9
2.0
>100
>100
>100
17
18
37.9
35.7
38.7
1.7
42
0.73
0.51
tafamidis
0.088
>50
0.14 mmol) in AcOH (1 mL) was added CrO3 (104 mg, 1.04 mmol) at
0 ◦C, and the resulting mixture was refluxed for 18 h. After cooling, the
reaction mixture was filtered through a pad of Celite and washed with
EtOAc (1 mL × 3). The filtrate and washings were combined and
evaporated to give a black solid, which was chromatographed on SiO2
(1H, s), 8.28 (1H, d, J = 1.2 Hz), 8.19 (2H, m), 8.00–7.91 (2H, m); 13C
NMR (100 MHz, DMSO d6) δ: 181.16, 180.94, 164.87, 136.92, 136.12,
135.67, 134.98, 134.39, 134.20, 133.83, 132.04, 131.75, 127.05,
126.48, 126.45; IR (KBr): 3086, 3072, 2982, 1701, 1680, 1589, 1302,
1261, 1178 cmꢀ 1; mp: 249–250 ◦C; MS (EI) m/z: 286 (M+); HRMS (EI)
m/z: 286.0031 (M+) [Calcd for C15H7ClO4: 286.0033].
(5 g, CH2Cl2/MeOH
= 50/1–5/1) to afford the corresponding
anthraquinone derivatives as an orange solid.43
4.1.3.9. 4-Fluoro-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid
1
4.1.3.2. Method B. To a stirred solution of the anthraquinone (A1–A5,
0.52 mmol) in concentrated H2SO4 (3.0 mL) was added MnO2
(272 mg, 3.12 mmol) at 0 ◦C, and the resulting mixture was heated at
(26). Yield: 85% (Method A); H NMR (400 MHz, DMSO d6) δ: 8.50
(1H, s), 8.21–8.16 (2H, m), 8.05 (1H, d, J = 11.6 Hz), 7.96–7.90 (2H,
m), 2.49 (3H, s); 13C NMR (100 MHz, DMSO d6) δ: 181.04, 180.15,
164.80, 161.97, 137.19, 137.11, 135.14, 134.90, 134.46, 133.64,
132.11, 126.67, 123.21, 122.87, 122.63; IR (KBr): 3971, 3751, 3636,
3094, 2361, 2328, 1678, 1612, 1591, 1570, 1470, 1423, 1223, 1165,
1047, 1022 cmꢀ 1; mp: 224–225 ◦C; MS (EI) m/z: 270 (M+); HRMS (EI)
m/z: 270.0327 (M+) [Calcd for C15H7FO4: 270.0328].
◦
60 C for 20 h. After cooling, the reaction mixture was diluted with
EtOAc (3 mL), and the organic layer was separated and washed with
water (1 mL × 3). The organic layer was dried over Na2SO4 and
evaporated to give a black solid, which was chromatographed on SiO2
(5 g, CH2Cl2/MeOH
= 50/1–5/1) to afford the corresponding
anthraquinone derivatives as an orange solid.44
4.1.3.10. 5-Acetyloxy-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic
acid (20). Anthraquinone 20 was prepared from 1-acetoxy-6-methylan-
thraquinone according to the literature.45
4.1.3.3. 9,10-Dioxo-9,10-dihydroanthracene-2-carboxylic acid (18).
Yield: 41% (Method A); 1H NMR (400 MHz, DMSO d6) δ: 8.67 (1H, s),
8.40 (1H, dd, J = 8.0, 1.2 Hz), 8.31 (1H, d, J = 8.0 Hz), 8.26–8.22
(2H, m), 7.98–7.94 (2H, m).58
1H NMR (400 MHz, DMSO d6) δ: 8.63 (1H, d, J = 2.0 Hz), 8.38
(1H, dd, J = 8.0, 2.0 Hz), 8.22 (1H, d, J = 8.0 Hz), 8.21 (1H, dd,
J = 8.0, 1.2 Hz), 7.98 (1H, t, J = 8.0 Hz), 7.66 (1H, dd, J = 8.0,
1.2 Hz), 2.42 (3H, s).
4.1.3.4. 3-Methyl-9,10-dioxo-9,10-dihydroanthracene-1-carboxylic acid
(21). Yield: 54% (Method B); 1H NMR (400 MHz, DMSO d6) δ:
8.21–8.19 (1H, m), 8.16–8.14 (1H, m), 8.08 (1H, s), 7.93 (2H, m), 7.66
(1H, s), 2.52 (3H, s); 13C NMR (100 MHz, DMSO d6) δ: 182.15, 181.62,
170.19, 145.49, 136.28, 134.73, 134.56, 133.26, 133.05, 132.94,
132.66, 127.65, 127.09, 126.75, 21.14; IR (KBr): 3975, 2991, 2659,
2338, 1709, 1682, 1591, 1325, 1277, 1219 cmꢀ 1; mp: 285–286 ◦C; MS
(EI) m/z: 266 (M+); HRMS (EI) m/z: 266.0585 (M+) [Calcd for
C16H10O4: 266.0579].
4.1.3.11. Synthesis of 5-Hydroxy-9,10-dioxo-9,10-dihydroanthracene-2-
carboxylic acid (19). Anthraquinone 20 (20 mg, 0.06 mmol) was dis-
solved in 2.5% Na2CO3 aq. (1.0 mL). The resulting mixture was refluxed
for 2 h. After cooling, the reaction mixture was acidified with 10% HCl
aq. (2 mL). The aqueous mixture was extracted with CH2Cl2
(1 mL × 3). The organic extracts were combined, dried over Na2SO4,
and evaporated to give a pale yellow solid, which was chromatographed
on SiO2 (5 g, CH2Cl2/MeOH = 50/1) to afford 19 (17 mg, 0.06 mmol,
98%) as a pale yellow solid.
4.1.3.5. 9,10-Dioxo-9,10-dihydroanthracene-1,3-dicarboxylic acid (22).
Yield: 7% (Method B); 1H NMR (400 MHz, CDCl3) δ: 8.57 (1H, br),
8.21–8.19 (1H, m), 8.14–8.12 (1H, m), 8.02 (1H, br), 7.92–7.90 (2H,
m).59
1H NMR (400 MHz, DMSO d6) δ: 12.32 (1H, s), 8.64 (1H, d,
J = 2.0 Hz), 8.41 (1H, d, J = 8.0 Hz), 8.34 (1H, d, J = 8.0 Hz), 7.85
(1H, dd, J = 8.4, 7.6 Hz), 7.77 (1H, dd, J = 7.6, 1.2 Hz), 7.43 (1H,
dd, J = 8.4, 1.2 Hz).60
4.1.3.6. 4-Methyl-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid
1
4.1.3.12. Synthesis of 5-Ethoxy-9,10-dioxo-9,10-dihydroanthracene-2-
carboxylic acid (27). To a stirred solution of 20 (22 mg, 0.07 mmol)
in EtOH (2.0 mL) was added concentrated H2SO4 (1 drop) at 0 ◦C, and
the resulting mixture was refluxed for 18 h. After cooling, EtOH was
evaporated and the residue was extracted with CH2Cl2 (1 mL × 3). The
organic extracts were combined, dried over Na2SO4, and evaporated to
give a pale yellow solid, which was used directly in the next step. To a
stirred solution of the above ethyl ester in DMF (1 mL) were added
Cs2CO3 (26 mg, 0.08 mmol) and EtI (0.01 mL, 0.10 mmol) at 0 ◦C, and
the resulting mixture was heated at 60 ◦C for 20 h. After cooling, the
reaction was quenched with water (3 mL), and then the aqueous
mixture was extracted with Et2O (1 mL × 3). The organic extracts were
combined, dried over Na2SO4, and evaporated to give a yellow solid,
which was used directly in the next step. To a stirred solution of the
above ethyl ether in EtOH (1.5 mL) was added 10% NaOH aq. (1.5 mL)
at room temperature, and the resulting mixture was refluxed for 22 h.
(23). Yield: 19% (Method A); H NMR (400 MHz, DMSO d6) δ: 8.60
(1H, d, J = 1.6 Hz), 8.20–8.18 (3H, m), 7.96–7.89 (2H, m), 2.83 (3H,
s); 13C NMR (100 MHz, DMSO d6) δ: 183.95, 182.14, 165.84, 141.80,
137.81, 134.72, 134.60, 134.35, 134.17, 134.13 133.19, 132.10,
126.88, 126.28, 125.85, 22.77; IR (KBr): 1701, 1676, 1593, 1269 cmꢀ +1
;
mp: 272–273 ◦C; MS (EI) m/z: 266 (M+); HRMS (EI) m/z: 266.0572 (M )
[Calcd for C16H10O4: 266.0579].
4.1.3.7. 5,6,7,8-Tetrachloro-9,10-dioxo-9,10-dihydroanthracene-2-car-
boxylic acid (24). Yield: 76% (Method A); 1H NMR (400 MHz,
DMSO d6) δ: 8.51 (1H, d, J = 1.2 Hz), 8.36 (1H, dd, J = 8.0, 1.2 Hz),
8.18 (1H, d, J = 8.0 Hz).56
4.1.3.8. 4-Chloro-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid
1
(25). Yield: 61% (Method A); H NMR (400 MHz, DMSO d6) δ: 8.63
9

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
Table 3
Pharmacokinetic parameters of 40 and 42 from rat experiments.
Compound
Route
Dose (mg/kg)
C0 or Cmax (ng/mL)
Tmax (h)
t1/2 (h)
AUC∞ (ng/mL⋅h)
CLtot (mL/h/kg)
Vdss (mL/kg)
AUC0-8h (ng/mL⋅h)
BA (%)
69.8
40
i.v.
p.o.
i.v.
1
1
8650
1124
19229
476
–
2
1.7
5.0
2.4
4.4
12
13163
9189
7516
2625
4334
NC
76
–
164
–
42
1
–
133
–
148
–
34.9
p.o.
i.v.
1
0.25
–
tafamidis
0.1
0.1
537
23
–
375
–
1714
956
55.8*
p.o.
163
8
NC
NC: Not calculated.
BA: Bioavailability.
*
The bioavailability of tafamidis was estimated from the value of AUC0–8h
.
After cooling, EtOH was evaporated and the residue was acidified with
10% HCl aq. (2 mL). The aqueous mixture was extracted with CH2Cl2
(1 mL × 3). The organic extracts were combined, dried over Na2SO4,
and evaporated to give a pale yellow solid, which was chromatographed
on SiO2 (5 g, CH2Cl2/MeOH = 50/1) to afford 27 (12 mg, 0.04 mmol,
57% in 3 steps) as a yellow solid.
(400 MHz, CDCl3) δ: 8.27 (1H, d, J = 8.8 Hz), 8.21 (1H, d,
J = 8.4 Hz), 7.50 (1H, d, J = 2.0 Hz), 7.34 (1H, dd, J = 8.8, 2.0 Hz),
7.23 (1H, s), 7.22 (1H, dd, J = 8.4, 0.8 Hz), 2.52 (3H, s).63
4.1.4.6. 5-Chloro-3-methyl-9H-xanthen-9-one (X4). Yield: 33%; 1H NMR
(400 MHz, CDCl3) δ: 8.25 (1H, dd, J = 8.0, 2.0 Hz), 8.21 (1H, d,
J = 8.0 Hz), 7.78 (1H, dd, J = 8.0, 2.0 Hz), 7.43 (1H, s), 7.31 (1H, t,
J = 8.0 Hz), 7.24 (1H, dd, J = 8.0, 1.2 Hz), 2.54 (3H, s); 13C NMR
(100 MHz, CDCl3) δ: 176.13, 155.71, 151.55, 146.74, 134.58, 126.33,
125.91, 125.15, 123.58, 123.06, 122.61, 119.02, 117.85, 21.89; IR
(KBr): 3836.2, 3745.5, 1670.2, 1627.8, 1600.8, 1467.7, 1438.8, 1313.4,
1238.2, 1180.4, 1151.4 cmꢀ 1; mp: 164–165 ◦C; MS (EI) m/z: 244 (M+);
HRMS (EI) m/z: 244.0282 (M+) [calcd for C14H9ClO2: 244.0291].
1H NMR (500 MHz, CDCl3) δ: 8.59 (1H, m), 8.35 (1H, dd, J = 8.0,
2.0 Hz), 8.20 (1H, d, J = 8.0 Hz), 7.85–7.82 (2H, m), 7.62–7.59 (1H,
m), 4.26–4.22 (2H, m), 1.45–1.41 (3H, m); 13C NMR (100 MHz, ace-
tone d6) δ: 182.47, 180.54, 165.54, 160.12, 137.85, 135.61, 135.43,
134.56, 132.69, 128.08, 127.54, 127.36, 124.28, 119.98, 119.19, 65.13,
14.23; IR (KBr): 3298.0, 3197.8, 3095.5, 2935.5, 1708.8, 1674.1,
1606.6, 1585.4, 1417.6, 1261.4, 1228.6, 1064.6 cmꢀ 1
;
mp:
+
+
◦
284–286 C; MS (EI) m/z: 296 (M ); HRMS (EI) m/z: 296.0695 (M )
4.1.4.7. 5-Chloro-2-methyl-9H-xanthen-9-one (X5). Yield: 21%; 1H NMR
(400 MHz, CDCl3) δ: 8.26 (1H, d, J = 8.0 Hz), 8.12 (1H, s), 7.79 (1H,
[calcd for C14H9ClO2: 296.0685].
4.1.4. General procedure for construction of the xanthone skeleton
To a stirred solution of the corresponding o-nitrobenzaldehyde
(1.32 mmol), PPh3 (26 mg, 0.10 mmol), anhydrous CuCl2 (9 mg,
0.07 mmol), and dried K3PO4 (618 mg, 2.91 mmol) in toluene (5 mL)
were added to the corresponding phenol (1.72 mmol) at room tem-
perature, and the resulting mixture was heated at 100 ◦C for 45 h under
an air atmosphere. After cooling, the reaction was quenched with 10%
NaOH aq. (5 mL) and the aqueous mixture was extracted with CH2Cl2
(5 mL × 3). The organic extracts were combined, dried over Na2SO4,
and evaporated to give a black solid, which was chromatographed on
SiO2 (15 g, hexane/EtOAc = 50/1–5/1) to afford the corresponding
xanthone as a pale yellow solid.46
dd, J
= 7.4, 1.6 Hz), 7.59 (1H, d, J = 8.8 Hz), 7.52 (1H, d,
J = 8.8 Hz), 7.31 (1H, dd, J = 8.0, 7.4 Hz), 2.49 (3H, s).64
4.1.4.8. 6-Chloro-2-methyl-9H-xanthen-9-one (X6). Yield: 19%; 1H NMR
(400 MHz, CDCl3) δ: 8.28 (1H, d, J = 8.4 Hz), 8.11 (1H, s), 7.56 (1H,
dd, J
= 8.8, 2.6 Hz), 7.51 (1H, d, J = 2.0 Hz), 7.40 (1H, d,
J = 8.8 Hz), 7.34 (1H, dd, J = 8.4, 2.0 Hz), 2.48 (3H, s).65
4.1.4.9. 2-Chloro-7-methyl-9H-xanthen-9-one (X7). Yield: 23%; 1H NMR
(400 MHz, CDCl3) δ/ppm: 8.30 (1H, d, J = 2.6 Hz), 8.11 (1H, s), 7.66
(1H, dd, J = 9.2, 2.6 Hz), 7.56 (1H, dd, J = 8.4, 2.6 Hz), 7.46 (1H, d,
J = 9.2 Hz), 7.41 (1H, d, J = 8.4 Hz), 2.48 (3H, s).66
4.1.4.1. 2-Methyl-9H-xanthen-9-one (28). Yield: 31%; 1H NMR
(400 MHz, CDCl3) δ: 8.35 (1H, dd, J = 8.0, 2.0 Hz), 8.13 (1H, d,
4.1.4.10. 1-Chloro-7-methyl-9H-xanthen-9-one (X8). Yield: 5%; 1H NMR
(400 MHz, CDCl3) δ: 8.09 (1H, s), 7.56 (1H, dd, J = 8.4, 7.6 Hz), 7.52
(1H, dd, J = 8.8, 2.4 Hz), 7.41 (1H, dd, J = 8.4, 1.2 Hz), 7.37 (1H, d,
J = 7.6 Hz), 7.34 (1H, d, J = 8.8 Hz), 2.47 (3H, s).67
J
= 2.0 Hz), 7.72 (1H, ddd, J = 8.4, 6.8, 2.0 Hz), 7.55 (1H, dd,
J = 8.4, 2.0 Hz), 7.49 (1H, d, J = 8.4 Hz) , 7.41 (1H, d, J = 8.4 Hz),
7.38 (1H, ddd, J = 8.0, 6.8, 1.2 Hz), 2.48 (3H, s).47
4.1.4.11. 4-Chloro-2-methyl-9H-xanthen-9-one (X9). Yield: 18%; 1H
NMR (500 MHz, CDCl3) δ: 8.33 (1H, dd, J = 8.0, 1.5 Hz), 8.05–8.04
(1H, m), 7.76 (1H, ddd, J = 8.0, 7.0, 1.5 Hz), 7.64 (1H, m), 7.61 (1H,
4.1.4.2. 3-Methyl-9H-xanthen-9-one (29). Yield: 47%; 1H NMR
(400 MHz, CDCl3) δ/ppm: 8.34 (1H, dd, J = 8.0, 1.2 Hz), 8.23 (1H, d,
J
= 8.0 Hz), 7.72 (1H, ddd, J = 8.0, 7.2, 1.2 Hz), 7.48 (1H, d,
d, J = 8.0 Hz), 7.41 (1H, ddd, J = 8.0, 7.0, 1.0 Hz), 2.46 (3H, s); 13
C
J = 8.0 Hz), 7.37 (1H, ddd, J = 8.0, 7.2, 1.2 Hz), 7.30 (1H, s), 7.20
NMR (100 MHz, CDCl3) δ: 176.64, 155.80, 149.87, 135.96, 135.01,
(1H, d, J = 8.0 Hz).47
134.01, 126.69, 124.84, 124.28, 122.64, 122.25, 121.35, 118.19, 20.68;
IR (KBr): 3865, 1663, 1591, 1462, 1317, 1256, 1225, 1148, 1022 cmꢀ 1
;
1
mp: 167–168 ◦C, MS (EI) m/z: 244 (M+); HRMS (EI) m/z: 244.0292 (M+)
4.1.4.3. 1-Chloro-6-methyl-9H-xanthen-9-one (X1). Yield: 9%; H NMR
(400 MHz, CDCl3) δ: 8.19 (1H, d, J = 8.0 Hz), 7.55 (1H, dd, J = 8.4,
8.0 Hz), 7.40 (1H, dd, J = 8.4, 1.2 Hz), 7.36 (1H, dd, J = 8.0,
1.2 Hz), 7.23 (1H, s), 7.19 (1H, dd, J = 8.0, 1.2 Hz), 2.50 (3H, s).61
[calcd for C14H9ClO2: 244.0291].
4.1.4.12. 4,5-Dichloro-2-methyl-9H-xanthen-9-one (X10). Yield: 18%;
1H NMR (500 MHz, CDCl3) δ: 8.24 (1H, dd, J = 8.0, 1.5 Hz), 8.03 (1H,
m), 7.83 (1H, dd, J = 8.0, 1.5 Hz), 7.67 (1H, d, J = 2.5 Hz), 7.35 (1H,
t, J = 8.0 Hz), 2.48 (3H, s); 13C NMR (100 MHz, CDCl3) δ: 176.07,
151.48, 149.63, 136.37, 135.19, 134.75, 125.24, 124.76, 124.23,
123.31, 122.81, 122.63, 122.20, 20.75; IR (KBr): 3894, 3751, 2328,
1665, 1564, 1437, 1377, 1308, 1186, 1126, 1067 cmꢀ 1; mp:
4.1.4.4. 2-Chloro-6-methyl-9H-xanthen-9-one (X2). Yield: 11%; 1H NMR
(400 MHz, CDCl3) δ: 8.29 (1H, d, J
= 2.4 Hz), 8.21 (1H, d,
J = 8.4 Hz), 7.65 (1H, dd, J = 9.2, 2.4 Hz), 7.45 (1H, d, J = 9.2 Hz),
7.30 (1H, s), 7.22 (1H, dd, J = 8.4, 1.2 Hz), 2.52 (3H, s).62
4.1.4.5. 3-Chloro-6-methyl-9H-xanthen-9-one (X3). Yield: 29%; 1H NMR
+
+
◦
254–255 C; MS (EI) m/z: 278 (M ); HRMS (EI) m/z: 277.9893 (M )
10

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
[calcd for C14H8Cl2O2: 277.9901].
68%; 1H NMR (400 MHz, DMSO d6) δ: 8.71 (1H, d, J = 2.4 Hz), 8.36
(1H, dd, J = 8.8, 2.4 Hz), 8.17 (1H, dd, J = 8.0, 2.0 Hz), 8.09 (1H,
4.1.4.13. 2,4-Dimethyl-9H-xanthen-9-one (X11). Yield: 26%; 1H NMR
(400 MHz, CDCl3) δ: 8.34 (1H, dd, J = 8.0, 1.6 Hz), 7.98 (1H, s), 7.72
(1H, ddd, J = 8.4, 6.8, 1.6 Hz), 7.54 (1H, d, J = 8.4 Hz), 7.40 (1H, s),
7.37 (1H, ddd, J = 8.0, 6.8, 1.6 Hz), 2.54 (3H, s), 2.43 (3H, s).68
dd, J
= 8.0, 2.0 Hz), 7.83 (1H, d, J = 8.8 Hz), 7.51 (1H, t,
J
= 8.0 Hz); 13C NMR (100 MHz, DMSO d6) δ: 175.32, 166.00,
157.63, 151.04, 135.82, 135.74, 127.83, 127.24, 125.13, 125.10,
122.69, 121.81, 120.66, 119.02; IR (KBr): 3155, 3038, 1730, 1645,
1616, 1603, 1489, 1470, 1435, 1312 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z:
274 (M+); HRMS (EI) m/z: 274.0026 (M+) [calcd for C14H7ClO4:
274.0033].
4.1.5. General procedure for the synthesis of xanthone derivatives 30–42
To a stirred solution of the xanthone (28 or 29 or X1–X11,
0.11 mmol) in AcOH (1 mL)/Ac2O (1 mL) was added CrO3 (104 mg,
1.04 mmol) at 0 ◦C, and the resulting mixture was heated at 100 ◦C for
3 h. After cooling, the reaction mixture was filtered through a pad of
Celite and washed with EtOAc (1 mL × 3). The filtrate and washings
were combined and evaporated to give a black solid, which was chro-
matographed on SiO2 (5 g, hexane/EtOAc = 5/1–1/1) to afford the
corresponding xanthone derivatives as a white solid.
4.1.5.8. 6-Chloro-9-oxo-9H-xanthene-2-carboxylic acid (37). Yield:
53%; 1H NMR (400 MHz, DMSO d6) δ: 8.71 (1H, d, J = 2.4 Hz), 8.35
(1H, dd, J = 8.8, 2.4 Hz), 8.20 (1H, d, J = 8.8 Hz), 7.93 (1H, d,
J = 1.6 Hz), 7.76 (1H, d, J = 8.8 Hz), 7.57 (1H, dd, J = 8.8, 1.6 Hz);
13C NMR (100 MHz, DMSO d6) δ: 175.43, 166.56, 157.56, 157.46,
155.97, 140.13, 136.09, 128.04, 127.51, 125.37, 120.76, 120.15,
118.46, 118.33; IR (KBr): 3414, 2922, 1676, 1605, 1445, 1425,
1261 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z: 274 (M+); HRMS (EI) m/z:
274.0034 (M+) [calcd for C14H7ClO4: 274.0033].
4.1.5.1. 9-Oxo-9H-xanthene-2-carboxylic acid (30). Yield: 79%; 1H
NMR (400 MHz, DMSO d6) δ: 8.73 (1H, d, J = 2.0 Hz), 8.34 (1H, dd,
J
J
= 8.8, 2.0 Hz), 8.22 (1H, dd, J = 8.0, 2.0 Hz), 7.92 (1H, ddd,
= 8.4, 7.4, 2.0 Hz), 7.77 (1H, d, J = 8.8 Hz), 7.72 (1H, d,
4.1.5.9. 7-Chloro-9-oxo-9H-xanthene-2-carboxylic acid (38). Yield:
26%; 1H NMR (400 MHz, DMSO d6) δ: 8.69 (1H, d, J = 2.4 Hz), 8.34
(1H, dd, J = 8.8, 2.4 Hz), 8.11 (1H, d, J = 2.8 Hz), 7.94 (1H, dd,
J = 8.4 Hz), 7.54 (1H, dd, J = 8.0, 7.4 Hz).69
4.1.5.2. 9-Oxo-9H-xanthene-3-carboxylic acid (31). Yield: 31%; 1H
NMR (400 MHz, DMSO d6) δ: 8.29 (1H, d, J = 8.8 Hz), 8.21 (1H, dd,
J = 8.0, 1.6 Hz), 8.10 (1H, s), 7.95 (1H, d, J = 8.8 Hz), 7.91 (1H, ddd,
J = 8.8, 7.2, 1.6 Hz), 7.71 (1H, d, J = 8.8 Hz), 7.51 (1H, dd, J = 8.0,
7.2 Hz).70
J
= 8.4, 2.8 Hz), 7.76 (1H, d, J = 8.4 Hz), 7.76 (1H, d,
J = 8.8 Hz).61
4.1.5.10. 8-Chloro-9-oxo-9H-xanthene-2-carboxylic acid (39). Yield:
42%; 1H NMR (400 MHz, DMSO d6) δ: 8.67 (1H, d, J = 2.4 Hz), 8.31
(1H, dd, J = 8.8, 2.4 Hz), 7.82 (1H, dd, J = 8.4, 8.0 Hz), 7.71 (1H, d,
J = 8.8 Hz), 7.67 (1H, d, J = 8.4 Hz), 7.54 (1H, dd, J = 8.0, 1.2 Hz);
13C NMR (100 MHz, DMSO d6) δ: 174.43, 166.09, 157.17, 156.92,
135.46, 135.24, 132.73, 128.04, 127.60, 126.90, 121.48, 118.41,
118.07, 117.93; IR (KBr): 3555, 3483, 3414, 3364, 2332, 1674, 1601,
1493, 1429, 1279, 1173, 1128 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z: 274
(M+); HRMS (EI) m/z: 274.0034 (M+) [calcd for C14H7ClO4: 274.0033].
4.1.5.3. 8-Chloro-9-oxo-9H-xanthene-3-carboxylic acid (32). Yield:
58%; 1H NMR (400 MHz, DMSO d6) δ: 8.23 (1H, d, J = 8.0 Hz), 8.04
(1H, s), 7.94 (1H, d, J = 8.0 Hz), 7.81 (1H, dd, J = 8.4, 8.0 Hz), 7.66
(1H, d, J = 8.4 Hz), 7.52 (1H, d, J = 8.0 Hz); 13C NMR (100 MHz,
DMSO d6) δ: 174.62, 166.03, 157.51, 154.17, 137.28, 135.25, 132.70,
127.38, 126.81, 124.45, 124.31, 118.72, 118.20, 117.98; IR (KBr):
1699, 1668, 1597, 1495, 1456, 1425, 1323, 1292, 1167 cmꢀ 1; mp:
+
+
◦
281–282 C; MS (EI) m/z: 274 (M ); HRMS (EI) m/z: 274.0034 (M )
4.1.5.11. 4-Chloro-9-oxo-9H-xanthene-2-carboxylic acid (40). Yield:
54%; 1H NMR (400 MHz, DMSO d6) δ: 8.58 (1H, d, J = 1.6 Hz), 8.34
[calcd for C14H7ClO4: 274.0033].
(1H, s), 8.17 (1H, d, J
= 6.8 Hz), 7.92 (1H, br), 7.73 (1H, d,
J = 7.6 Hz), 7.54 (1H, br); 13C NMR (100 MHz, DMSO d6) δ: 175.24,
165.23, 155.14, 153.58, 136.32, 134.75, 126.97, 126.34, 126.14,
125.52, 122.32, 122.26, 120.78, 118.41; IR (KBr): 3566, 3192, 3045,
2448, 2357, 2332, 1674, 1599, 1560, 1489, 1464, 1250, 1211, 1180,
1153, 1113, 1088 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z: 274 (M+); HRMS
(EI) m/z: 274.0034 (M+) [calcd for C14H7ClO4: 274.0033].
4.1.5.4. 7-Chloro-9-oxo-9H-xanthene-3-carboxylic acid (33). Yield:
56%; 1H NMR (400 MHz, DMSO d6) δ: 8.29 (1H, d, J = 8.0 Hz), 8.13
(1H, d, J = 2.4 Hz), 8.10 (1H, d, J = 1.2 Hz), 7.97 (1H, dd, J = 8.0,
1.2 Hz), 7.95 (1H, dd, J = 9.2, 2.4 Hz), 7.78 (1H, d, J = 9.2 Hz).70
4.1.5.5. 6-Chloro-9-oxo-9H-xanthene-3-carboxylic acid (34). Yield:
71%; 1H NMR (400 MHz, DMSO d6) δ: 8.27 (1H, d, J = 7.6 Hz), 8.20
(1H, d, J = 8.4 Hz), 8.07 (1H, s), 7.97 (1H, d, J = 7.6 Hz), 7.91 (1H,
s), 7.56 (1H, dd, J = 8.4, 2.4 Hz); 13C NMR (100 MHz, CDCl3) δ:
175.17, 165.92, 156.10, 155.22, 140.19, 137.12, 127.94, 126.70,
125.25, 124.60, 123.73, 120.15, 119.11, 118.25; IR (KBr): 2357, 1659,
4.1.5.12. 4,5-Dichloro-9-oxo-9H-xanthene-2-carboxylic acid (41). Yield:
61%; 1H NMR (400 MHz, DMSO d6) δ: 8.58 (1H, d, J = 2.0 Hz), 8.40
(1H, d, J = 2.0 Hz), 8.14 (1H, dd, J = 8.0, 2.0 Hz), 8.10 (1H, dd,
J = 8.0, 2.0 Hz), 7.53 (1H, td, J = 8.0, 2.0 Hz); 13C NMR (100 MHz,
DMSO d6) δ: 174.98, 165.18, 153.36, 150.78, 136.05, 135.04, 127.42,
126.29, 125.74, 125.15, 122.66, 122.47, 122.18, 122.10; IR (KBr):
3906, 3751, 3076, 2357, 2343, 2330, 1682, 1601, 1582, 1468, 1410,
1279, 1225, 1190, 1146 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z: 308 (M+);
HRMS (EI) m/z: 307.9644 (M+) [calcd for C14H6Cl2O4: 307.9643].
1607, 1583, 1560, 1491, 1433, 1298, 1207, 1173, 1094, 1072 cmꢀ 1
;
mp: >300 ◦C; MS (EI) m/z: 274 (M+); HRMS (EI) m/z: 274.0031 (M+)
[calcd for C14H7ClO4: 274.0033].
4.1.5.6. 5-Chloro-9-oxo-9H-xanthene-3-carboxylic acid (35). Yield:
83%; 1H NMR (400 MHz, DMSO d6) δ: 8.29 (1H, d, J = 8.4 Hz), 8.16
(1H, dd, J = 7.6, 1.6 Hz), 8.12 (1H, s), 8.08 (1H, dd, J = 8.0, 1.6 Hz),
7.99 (1H, dd, J = 8.4, 1.2 Hz), 7.50 (1H, dd, J = 8.0, 7.6 Hz); 13C
NMR (100 MHz, DMSO d6) δ: 175.26, 165.82, 154.82, 151.26, 137.17,
135.68, 126.74, 125.04, 124.90, 124.76, 123.38, 122.74, 121.79,
119.10; IR (KBr): 3065, 1703, 1674, 1603, 1495, 1474, 1429, 1304,
1209, 1177 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z: 274 (M+), HRMS (EI) m/z:
274.0034 (M+) [calcd for C14H7ClO4: 274.0033].
4.1.5.13. 4-Methyl-9-oxo-9H-xanthene-2-carboxylic acid (42). Yield:
99%; 1H NMR (400 MHz, DMSO d6) δ: 13.24 (1H, br), 8.58 (1H, s), 8.24
(1H, s), 8.20 (1H, d, J = 7.6 Hz), 7.92 (1H, dd, J = 8.4, 8.0 Hz), 7.75
(1H, d, J = 8.4 Hz), 7.52 (1H, dd, J = 8.0, 7.6 Hz), 2.58 (3H, s); 13C
NMR (100 MHz, DMSO d6) δ: 176.03, 166.32, 156.49, 155.40, 135.90,
135.81, 128.06, 126.00, 125.98, 125.50, 124.93, 120.84, 120.64,
118.50, 15.32; IR (KBr): 2370, 2353, 2324, 1697, 1663, 1606, 1597,
1470, 1431, 1406, 1321, 1292, 1225 cmꢀ 1; mp: >300 ◦C; MS (EI) m/z:
254 (M+); HRMS (EI) m/z: 254.0588 (M+) [calcd for C15H10O4:
4.1.5.7. 5-Chloro-9-oxo-9H-xanthene-2-carboxylic acid (36). Yield:
11

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
254.0579].
4.6. ADME assay
Solubility assay: The Japanese Pharmacopeia (JP) 1st fluid (pH 1.2)
or JP 2nd fluid (pH 6.8) for dissolution testing was used for solubility
measurements. A test solution of the test compound was prepared by
4.2. Materials
Wild-type TTR and V30M were expressed in Escherichia coli as
described previously.71 Chrysazin (1), Emodin (3), Aloe Emodin (5),
Alizarin Safirol SE (11), Nuclear Fast Red (12), Leucoquinizarin (14),
Anthrarufin (15), Diacerein (16) and Daunorubicin Hydrochloride (17)
were purchased from Tokyo Chemical Industry Co. Chrysophanol (2),
Physcion (4) and Purpurin (7) were purchased from LKT Laboratories.
Isopurpurin (9) and Quinizarin (13) were purchased from Sigma-Aldrich
Japan. Rhein (6), Alizarin (8) and Quinalizarin (10) were purchased
from AdooQ-BioScience, FUJIFILM Wako Pure Chemical, and Vitas-M
Laboratory, respectively.
diluting 10 mM DMSO stock solution 2
μL:165 μL in JP 1st or 2nd fluid
and mixed at 37 ◦C for 4 h by rotation at 1000 rpm. After loading the
mixed solution into 96-well MultiScreen Filter Plates (product number
MSHVN4510, 0.45
μm hydrophilic PVDF membrane; Millipore, Bed-
ford, MA), filtration was performed by centrifugation. The filtrates were
mixed with acetonitrile and analyzed by HPLC-UV (254 nm). Solubility
was calculated by comparing the peak area of the filtrate mixture with
that of a 100
μM standard solution. When the peak area of the filtrate
mixture was larger than the peak area of the standard solution, it was
described as >100
μM. At least two technical replicates were
performed.
A PAMPA assay was performed to determine the passive membrane
diffusion rates: A Corning Gentest Pre-coated PAMPA Plate System was
used in the PAMPA permeability test. The acceptor plate was prepared
4.3. Protein crystallography
For the preparation of TTR and inhibitor complexes, 30 mg/mL WT-
TTR or V30M-TTR and 10 mM of compound 7, 25, 32 or 40 (in DMSO)
were mixed in a 9:1 ratio, and the protein samples were incubated for
60 min at room temperature. Single crystals suitable for a diffraction
experiment were successfully obtained by the sitting-drop vapor-diffu-
sion method under the condition [15–26% polyethylene glycol 400,
0.4 M CaCl2, 0.1 M sodium acetate pH 4.9]. Crystals were cryo-
protected by the buffer [32% polyethylene glycol 400, 0.4 M CaCl2,
0.1 M sodium acetate pH 5.0] and flash-frozen using liquid nitrogen. X-
ray diffraction data were collected using beamlines 5A and 17A at the
Photon Factory (Tsukuba, Japan). The diffraction data sets were pro-
cessed with XDS.72 All crystals were isomorphous with typical TTR
crystal structures with space group P21212. Therefore, structure re-
finements were directly performed without molecular replacement
phasing using the apo V30M-TTR (PDB ID: 4PWE) or apo WT-TTR (PDB
ID: 4 N85) structures as initial models.39,40 The structures were refined
using PHENIX.REFINE with manual model building using COOT.73,74
The 3D structures and library data of the inhibitors were created using
PRODRG {Schuttelkopf, 2004 #1122}. Because the inhibitor-binding
sites were located on the crystallographic symmetry axis, the in-
hibitors were refined at 50% occupancy (Fig. 2). The coordinates and
structure factors have been deposited in the Protein Data Bank (Sup-
plementary Tables 1 and 2).
by adding 200
μL of 5% DMSO/0.1 M phosphate buffer (pH7.4) to each
well, and then 300
μL of 10 μM test compounds in 5% DMSO/0.1 M
phosphate buffer (pH 6.4) was added to the donor wells. The acceptor
plate was then placed on top of the donor plate and incubated at 37 ◦C
without agitation for 4 h. At the end of the incubation, the plates were
separated and the solutions from each well of both the acceptor plate
and the donor plate were transferred to 96-well plates and mixed with
acetonitrile. The final concentrations of compounds in both the donor
wells and acceptor wells, as well as the concentrations of the initial
donor solutions, were analyzed by liquid chromatography tandem mass
spectrometry (LC-MS/MS). The permeability of the compounds was
calculated according to a previous report.53 The recovery of tested
compounds was more than 90%. The permeabilities of antipyrine
(100
μM), metoprolol (500 μM) and sulfasalazine (500 μM) as refer-
ence compounds were 13, 1.0 and 0.044 × 10ꢀ 6 cm/s, respectively. At
least two technical replicates were performed.
Hepatic microsomal stability assay: Disappearance of the parent
compound over time was measured by using the amount of drug at time
zero as a reference. After 5 min of preincubation, 1 mM NADPH (final
concentration; the same applies to the following) was added to a mixture
containing 1
μM of the test compound, 0.2 mg/mL of rat liver micro-
somes (Sekisui XenoTech LLC, Kansas City, KS), 1 mM EDTA and 0.1 M
phosphate buffer (pH 7.4) and incubated at 37 ◦C for 30 min with
rotation at 60 rpm. An aliquot of 50
μL of the incubation mixture was
4.4. Acid-Mediated aggregation experiments
sampled and added to 250
μL of chilled acetonitrile/internal standard
(IS). After centrifuging for 15 min at 3150g (4 ◦C), the supernatants
were analyzed by LC-MS/MS. Hepatic microsomal stability (mL/min/
kg, CLint) was calculated according to a previous report,76 using
44.8 mg MS protein/g liver and 40.0 g liver/kg body weight as scaling
factors. At least two technical replicates were performed.
V30M-TTR was incubated at pH 7.0 in the presence of the com-
pounds for 30 min at room temperature prior to protein aggregation.
Acid-mediated protein aggregation was initiated by diluting the protein
solution with 100 mM acetate buffer at pH 4.6. The protein concen-
tration was set to 10
0–100
μM and the compound concentration was set to
Determination of the unbound fraction in rat plasma: An equilibrium
dialysis apparatus was used to determine the unbound fraction for each
compound in rat plasma. A Rapid Equilibrium Dialysis (RED) Device
Single-Use Plate with Inserts, 8 K MWCO (ThermoFisher Scientific,
Waltham, MA) was used. Plasma was spiked with the test compound
μ
M. After incubation for 96 h at pH 4.6 and 37 ◦C, the solution
was 6-fold diluted with 200 mM Tris-HCl at pH 8.0 containing 20
μM
thioflavin T. The fluorescence measurement was performed as described
previously.39 The IC50 values were obtained by fitting the data with the
4-parameter logistic model.75 At least two technical replicates were
performed.
(1
μM), and 200 μL aliquots were loaded into the apparatus and dia-
lyzed versus 350
μ
L of 0.1 M phosphate buffer (pH 7.4) at 37 ◦C for 4 h
by rotation at 1000 rpm. The unbound fraction was calculated as the
ratio of receiver side (buffer) to donor side (plasma) concentrations. At
least two technical replicates were performed.
4.5. Animals
Male Sprague Dawley rats (7-week-old) were purchased from SLC
Inc. (Hamamatsu, Japan). All animals were maintained in air-
conditioned quarters with a room temperature of 20 ± 2 ◦C, relative
humidity of 50 ± 10%, and an alternating 12 h light:dark cycle. The
rats were fed a certified diet (MF; Oriental Yeast Co.) and water ad
libitum. The studies were conducted in accordance with the guidelines
provided by the Animal Care and Use Committee of Osaka University.
4.7. In vivo pharmacokinetics assay
A test compound solution in 10% DMSO/0.1 M phosphate buffer
(pH 8.5) was injected intravenously (i.v.) or orally (p.o.) to male SD rats.
The dosage of the test compounds was 0.1 (tafamidis) or 1 (40 and 42)
mg/5 mL/kg. Blood samples were collected from the caudal vein at 5 (i.
12

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
2 Vatassery GT, Quach HT, Smith WE, Benson BA, Eckfeldt JH. A sensitive assay of
transthyretin (prealbumin) in human cerebrospinal fluid in nanogram amounts by
ELISA. Clin Chim Acta. 1991;197:19–25.
v.), 15, and 30 min and 1, 2, 4, and 8 h. Plasma was prepared by
centrifugation of the blood samples at ꢀ 80 ◦C. Plasma samples were
precipitated with 4 volumes of acetonitrile/IS and centrifuged at
15,000g at 4 ◦C for 10 min. The supernatants were analyzed by LC-MS/
MS. Standard non-compartmental analysis was performed to determine
the pharmacokinetic parameters: the estimated initial concentration
(C0), maximum plasma concentration (Cmax), time to maximum plasma
concentration (Tmax), elimination half-life (t1/2), area under the con-
centration time curve from time zero to infinity (AUC∞), AUC from time
zero to 8 h (AUC0-8h), total clearance (CLtot), and volume of distribution
at steady-state (Vdss). The absolute bioavailability (BA) of the oral dose
was calculated as AUC∞(p.o.)/AUC∞(i.v.). At least two biological rep-
licates were generated for each sample.
3 Johnson SM, Wiseman RL, Sekijima Y, Green NS, Adamski-Werner SL, Kelly JW.
Native state kinetic stabilization as a strategy to ameliorate protein misfolding
diseases: a focus on the transthyretin amyloidoses. Acc Chem Res. 2005;38:911–921.
4 Johnson SM, Connelly S, Fearns C, Powers ET, Kelly JW. The transthyretin
amyloidoses: from delineating the molecular mechanism of aggregation linked to
pathology to a regulatory-agency-approved drug. J Mol Biol. 2012;421:185–203.
5 Vieira M, Saraiva MJ. Transthyretin: a multifaceted protein. Biomol Concepts. 2014;5:
45–54.
6 Galant NJ, Westermark P, Higaki JN, Chakrabartty A. Transthyretin amyloidosis: an
under-recognized neuropathy and cardiomyopathy. Clin Sci (Lond). 2017;131:
395–409.
7 Ando Y, Coelho T, Berk JL, et al. Guideline of transthyretin-related hereditary
amyloidosis for clinicians. Orphanet J Rare Dis. 2013;8:31.
8 Ueda M, Ando Y. Recent advances in transthyretin amyloidosis therapy. Transl
Neurodegener. 2014;3:19.
9 Yokoyama T, Mizuguchi M. Inhibition of the amyloidogenesis of transthyretin by
natural products and synthetic compounds. Biol Pharm Bull. 2018;41:979–984.
10 Yokoyama T, Mizuguchi M. Transthyretin amyloidogenesis inhibitors: from
discovery to current developments. J Med Chem. 2020;63:14228–14242.
11 Coelho T, Maia LF, Martins da Silva A, et al. Tafamidis for transthyretin familial
amyloid polyneuropathy: a randomized, controlled trial. Neurology. 2012;79:
785–792.
4.8. LC-MS/MS quantification Method
An LC-MS8060 instrument equipped with a Shimadzu Nexera series
LC system (Shimadzu, Kyoto, Japan) was used. All compounds were
analyzed in multi-reaction monitoring mode under electron spray ioni-
zation conditions. The analytical column used was a CAPCELLPAK C18
12 Coelho T, Maia LF, da Silva AM, et al. Long-term effects of tafamidis for the
treatment of transthyretin familial amyloid polyneuropathy. J Neurol. 2013;260:
2802–2814.
MGIII (3
μm × 2.0 mm ID × 35 mm; OSAKA SODA, Osaka, Japan) at
50 ◦C. The gradient mobile phase consisted of 0.1% formic acid in water
(mobile phase A) and 0.1% formic acid in acetonitrile (mobile phase B)
at a total flow rate of 1 mL/min. The initial mobile phase composition
was 10% B, which was held constant for 0.5 min, increased in a linear
fashion to 90% B over 1 min, then held constant for 0.8 min, and finally
brought back to the initial condition of 10% B over 0.01 min and re-
equilibrated for 1 min. The transitions (precursor ion > product ion)
of tafamidis, 40, 42, and IS (methyl testosterone) are 306.0 > 262.1,
273.0 > 229.1 (negative), 255.1 > 211.2, and 303.1 > 109.1 (pos-
itive), respectively.
13 Berk JL, Suhr OB, Obici L, et al. Diflunisal Trial Consortium. Repurposing diflunisal
for familial amyloid polyneuropathy: a randomized clinical trial. JAMA. 2013;310:
2658–2667.
´
14 Gamez J, Salvado M, Reig N, et al. Transthyretin stabilization activity of the
catechol-O-methyltransferase inhibitor tolcapone (SOM0226) in hereditary ATTR
amyloidosis patients and asymptomatic carriers: proof-of-concept study. Amyloid.
2019;26:74–84.
˜
15 Pinheiro F, Varejao N, Esperante S, et al. Tolcapone, a potent aggregation inhibitor
for the treatment of familial leptomeningeal amyloidosis. FEBS J. 2021;288:310–324.
16 Fox JC, Hellawell JL, Rao S, et al. First-in-human study of AG10, a novel, oral,
specific, selective, and potent transthyretin stabilizer for the treatment of
transthyretin amyloidosis: a phase 1 safety, tolerability, pharmacokinetic, and
pharmacodynamic study in healthy adult volunteers. Clin Pharmacol Drug Dev. 2020;
9:115–129.
17 Vidal R, Garzuly F, Budka H, et al. Meningocerebrovascular amyloidosis associated
with a novel transthyretin mis-sense mutation at codon 18 (TTRD 18G). Am J Pathol.
1996;148:361–366.
Declaration of Competing Interest
18 Benson MD. Leptomeningeal amyloid and variant transthyretins. Am J Pathol. 1996;
148:351–354.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
19 Brett M, Persey MR, Reilly MM, et al. Transthyretin Leu12Pro is associated with
systemic, neuropathic and leptomeningeal amyloidosis. Brain. 1999;122:183–190.
20 Nakagawa K, Sheikh SI, Snuderl M, Frosch MP, Greenberg SM. A new Thr49Pro
transthyretin gene mutation associated with leptomeningeal amyloidosis. J Neurol
Sci. 2008;272:186–190.
Acknowledgments
21 Llull L, Berenguer J, Yagüe J, Graus F. Leptomeningeal amyloidosis due to A25T TTR
mutation: a case report. Neurologia. 2014;29:382–384.
This research was supported by AMED under grant number
JP20lm0203011. This research was partially supported by the Platform
Project for Supporting Drug Discovery and Life Science Research (Basis
for Supporting Innovative Drug Discovery and Life Science Research
(BINDS)) from AMED under grant numbers JP20am0101087 and
JP20am0101123 (support number 2110).
22 Sekijima Y. Transthyretin-type cerebral amyloid angiopathy: a serious complication
in post-transplant patients with familial amyloid polyneuropathy. J Neurol Neurosurg
Psychiatry. 2015;86:124.
˜
˜
23 Beirao JM, Malheiro J, Lemos C, Beirao I, Costa P, Torres P. Ophthalmological
manifestations in hereditary transthyretin (ATTR V30M) carriers: a review of 513
cases. Amyloid. 2015;22:117–122.
˜
24 Maia LF, Magalhaes R, Freitas J, et al. CNS involvement in V30M transthyretin
amyloidosis: clinical, neuropathological and biochemical findings. J Neurol Neurosurg
Psychiatry. 2015;86:159–167.
Author contributions
´
25 Lozeron P, Theaudin M, Mincheva Z, Ducot B, Lacroix C, Adams D. French Network
for FAP (CORNAMYL). Effect on disability and safety of Tafamidis in late onset of
Met30 transthyretin familial amyloid polyneuropathy. Eur J Neurol. 2013;20:
1539–1545.
T.Y., N.T. and M.M. designed the project. T.Y. and R.K. performed
the X-ray crystallography. R.K. analyzed the inhibitory activity of the
compounds. K.I., Y.N., Y.S. and T.O. synthesized the compounds. W.K.
performed the ITC measurements. K.K. and S.N. performed in vitro
ADME and in vivo pharmacokinetic study. T.Y., K.K., N.T. and M.M.
prepared the manuscript based on discussions among all the authors.
26 Loconte V, Menozzi I, Ferrari A, et al. Structure-activity relationships of flurbiprofen
analogues as stabilizers of the amyloidogenic protein transthyretin. J Struct Biol.
2019;208:165–173.
27 Yokoyama T, Mizuguchi M. Crown ethers as transthyretin amyloidogenesis
inhibitors. J Med Chem. 2019;62:2076–2082.
ˆ
28 Cotrina EY, Oliveira A, Leite JP, et al. Repurposing benzbromarone for familial
amyloid polyneuropathy: a new transthyretin tetramer stabilizer. Int J Mol Sci. 2020;
21:7166.
Appendix A. Supplementary data
29 Ciccone L, Nencetti S, Tonali N, et al. Monoaryl derivatives as transthyretin fibril
formation inhibitors: design, synthesis, biological evaluation and structural analysis.
Bioorg Med Chem. 2020;28, 115673.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmc.2021.116292.
30 Kim B, Ko YH, Runfola M, et al. Diphenyl-methane based thyromimetic inhibitors for
transthyretin amyloidosis. Int J Mol Sci. 2021;22:3488.
31 Gong H, He Z, Peng A, et al. Effects of several quinones on insulin aggregation. Sci
Rep. 2014;4:5648.
References
32 Malik EM, Müller CE. Anthraquinones As pharmacological tools and drugs. Med Res
Rev. 2016;36:705–748.
1 Benson MD, Buxbaum JN, Eisenberg DS, et al. Amyloid nomenclature 2018:
recommendations by the International Society of Amyloidosis (ISA) nomenclature
committee. Amyloid. 2018;25:215–219.
33 Baures PW, Peterson SA, Kelly JW. Discovering transthyretin amyloid fibril inhibitors
by limited screening. Bioorg Med Chem. 1998;6:1389–1401.
13

R. Kitakami et al.
Bioorganic & Medicinal Chemistry 44 (2021) 116292
34 Pickhardt M, Gazova Z, von Bergen M, et al. Anthraquinones inhibit tau aggregation
and dissolve Alzheimer’s paired helical filaments in vitro and in cells. J Biol Chem.
2005;280:3628–3635.
54 Basak S, Mal D. Applications of [4+2] anionic annulation and carbonyl-ene reaction
in the synthesis of anthraquinones, tetrahydroanthraquinones, and
pyranonaphthoquinones. J Org Chem. 2017;82:11035–11051.
35 Maia F, Almeida Mdo R, Gales L, et al. The binding of xanthone derivatives to
transthyretin. Biochem Pharmacol. 2005;70:1861–1869.
55 Elkazaz S, Jones PB. Photochemical hydroxylation of 1-methyl-9,10-anthraquinones:
synthesis of 9’-hydroxyaloesaponarin II. J OrgChem. 2010;75:412–416.
56 Ruggli P, Brunner E. Anthracene derivatives. I. o-Nitroaldehydes of the
anthraquinone series. Helv Chim Acta. 1925;8:155–168.
36 Convertino M, Pellarin R, Catto M, Carotti A, Caflisch A. 9,10-Anthraquinone hinders
beta-aggregation: how does a small molecule interfere with Abeta-peptide amyloid
fibrillation? Protein Sci. 2009;18:792–800.
57 Quayle OR, Reid EE. Some substituted o-benzoylbenzoic acids and the corresponding
anthraquinones. J Am Chem Soc. 1925;47:2357–2361.
37 Yokoyama T, Ueda M, Ando Y, Mizuguchi M. Discovery of γ-mangostin as an
amyloidogenesis inhibitor. Sci Rep. 2015;5:13570.
58 Ashnagar A, Bruce JM. Synthesis of 2-bis(1-methylimidazol-2-yl)methoxycarbonyl-
9,10-anthraquinone. Int J Chem Tech Res. 2011;3:1937–1940.
38 Huang K, Jiang L, Li H, Ye D, Zhou L. Development of anthraquinone analogues as
phosphoglycerate mutase 1 inhibitors. Molecules. 2019;24:845.
59 Leon JW, Whitten DG. Photodegradation of oligomeric polyesters containing
anthraquinone and 1,2-diamine units. Single electron transfer induced cation radical
bond cleavage in the solid state. J Am Chem Soc. 1995;117:2226–2235.
60 Witte PD, Lemli J. Synthesis of c-9-14C-1,8-dihydroxy-3-carboxyanthraquinone.
J. Labelled Compd Radiopharm. 1998;25:23–33.
39 Yokoyama T, Kosaka Y, Mizuguchi M. Crystal structures of human transthyretin
complexed with glabridin. J Med Chem. 2014;57:1090–1096.
40 Yokoyama T, Kosaka Y, Mizuguchi M. Inhibitory activities of propolis and its
promising component, caffeic acid phenethyl ester, against amyloidogenesis of
human transthyretin. J Med Chem. 2014;57:8928–8935.
61 Williams AC, Camp N. Product class 4: benzopyranones and benzopyranthiones. Sci
Synth. 2003;14:347–638.
`
41 Pavelka K, Bruyere O, Cooper C, et al. Diacerein: Benefits, risks and place in the
management of osteoarthritis. An opinion-based report from the ESCEO. Drugs Aging.
2016;33:75–85.
62 Pan D, Wang Y, Li M, et al. Visible-light-induced aerobic oxidation of benzylic C(sp3)-
H of alkylarenes promoted by DDQ, tert-butyl nitrite, and acetic acid. Synlett. 2019;
30:218–224.
42 Zhang TJ, Li SY, Yuan WY, et al. Discovery and biological evaluation of some
(1H–1,2,3-triazol-4-yl)methoxybenzaldehyde derivatives containing an
anthraquinone moiety as potent xanthine oxidase inhibitors. Bioorg Med Chem Lett.
2017;27:729–732.
63 Zhang X, Yang L, Wu Y, et al. Microwave-assisted transition-metal-free
intramolecular Ullmann-type O-arylation in water for the synthesis of xanthones and
azaxanthones. Tetrahedron Lett. 2014;55:4883–4887.
43 Zhang TJ, Li SY, Yuan WY, Zhang Y, Meng FH. Design, synthesis, and molecular
docking studies of N-(9,10-anthraquinone-2-carbonyl)amino acid derivatives as
xanthine oxidase inhibitors. Chem Biol Drug Des. 2018;91:893–901.
44 Bu X, Deady LW, Finlay GJ, Baguley BC, Denny WA. Synthesis and cytotoxic activity
of 7-Oxo-7H-dibenz[f, i, j]isoquinoline and 7-Oxo-7H-benzo[e]perimidine
derivatives. J Med Chem. 2001;44:2004–2014.
64 Eckstein M, Marona H. Preparation and properties of some 2-
mercaptomethylxanthones. Pol J Chem. 1980;54:1281–1285.
65 Waszkielewicz AM, Gunia A, Szkaradek N, et al. Synthesis and evaluation of
pharmacological properties of some new xanthone derivatives with piperazine
moiety. Bioorg Med Chem Lett. 2013;23:4419–4423.
66 Zhang H, Shi R, Gan P, et al. Palladium-catalyzed oxidative double C-H
functionalization/carbonylation for the synthesis of xanthones. Angew Chem Int Ed.
2012;51:5204–5207.
45 Mitter PC, Das-Gupta S, Bachhwat S. Anthraquinone series. J Indian Chem Soc. 1934;
11:893–897.
46 Menendez CA, Nador F, Radivoy G, Gerbino DC. One-step synthesis of xanthones
catalyzed by a highly efficient copper-based magnetically recoverable nanocatalyst.
Org Lett. 2014;16:2846–2849.
67 Rao H, Ma X, Liu Q, Li Z, Cao S, Li CJ. Metal-free oxidative coupling: Xanthone
formation via direct annulation of 2-aryloxybenzaldehyde using
tetrabutylammonium bromide as a promoter in aqueous medium. Adv Synth Catal.
2013;355:2191–2196.
47 Venkanna A, Goud PVK, Prasad PV, Shanker M, Rao PV. One-step synthesis of
xanthones and thio-xanthones by using magnetically separable copper ferrite
nanoparticles as a catalyst under ligand free conditions. Chemistry Select. 2016;1:
2271–2276.
68 Wang P, Rao H, Hua R, Li CJ. Rhodium-catalyzed xanthone formation from 2-ary-
loxybenzaldehydes via cross-dehydrogenative coupling (CDC). Org Lett. 2012;14:
902–905.
48 Baures PW, Oza VB, Peterson SA, Kelly JW. Synthesis and evaluation of inhibitors of
transthyretin amyloid formation based on the non-steroidal anti-inflammatory drug,
flufenamic acid. Bioorg Med Chem. 1999;7:1339–1347.
69 Kottani R, Majjigapu JRR, Kurchan A, Majjigapu K, Gustafson TP, Kutateladze AG.
Photoinduced signal amplification through controlled externally sensitized
fragmentation in masked sensitizers. J Am Chem Soc. 2006;128:14794–14795.
70 Giri R, Goodell JR, Xing C, et al. Synthesis and cancer cell cytotoxicity of substituted
xanthenes. Bioorg Med Chem. 2010;18:1456–1463.
49 Adamski-Werner SL, Palaninathan SK, Sacchettini JC, Kelly JW. Diflunisal analogues
stabilize the native state of transthyretin. Potent inhibition of amyloidogenesis. J Med
Chem. 2004;47:355–374.
71 Matsubara K, Mizuguchi M, Kawano K. Expression of a synthetic gene encoding
human transthyretin in Escherichia coli. Protein Expr Purif. 2003;30:55–61.
72 Kabsch W. XDS. Acta Crystallogr D Biol Crystallogr. 2010;66:125–132.
50 Bulawa CE, Connelly S, Devit M, et al. Tafamidis, a potent and selective transthyretin
kinetic stabilizer that inhibits the amyloid cascade. Proc Natl Acad Sci USA. 2012;
109:9629–9634.
´
73 Adams PD, Afonine PV, Bunkoczi G, et al. PHENIX: a comprehensive Python-based
51 Gales L, Macedo-Ribeiro S, Arsequell G, Valencia G, Saraiva MJ, Damas AM. Human
transthyretin in complex with iododiflunisal: structural features associated with a
potent amyloid inhibitor. Biochem J. 2005;388:615–621.
system for macromolecular structure solution. Acta Crystallogr DBiol Crystallogr.
2010;66:213–221.
74 Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta
Crystallogr D Biol Crystallogr. 2004;60:2126–2132.
52 Razavi H, Palaninathan SK, Powers ET, et al. Benzoxazoles as transthyretin amyloid
fibril inhibitors: synthesis, evaluation, and mechanism of action. Angew Chem Int Ed
Engl. 2003;42:2758–2761.
75 Sebaugh JL. Guidelines for accurate EC50/IC50 estimation. Pharm Stat. 2011;10:
128–134.
53 Chen X, Murawski A, Patel K, Crespi CL, Balimane PV. A novel design of artificial
membrane for improving the PAMPA model. Pharm Res. 2008;25:1511–1520.
76 Shah P, Kerns E, Nguyen DT, et al. An Automated high-throughput metabolic
stability assay using an integrated high-resolution accurate mass method and
automated data analysis software. Drug Metab Dispos. 2016;44:1653–1661.
14