Ridaifen-F conjugated with cell-penetrating peptides inhibits intracellular proteasome activities and induces drug-resistant cell death

Article abstract of DOI:10.1016/j.ejmech.2018.01.045

Ridaifen-F (RID-F) potently inhibits proteolytic activities of the 20S proteasome but poorly inhibits those of the 26S proteasome. Here, we report preparation of several conjugates in which various peptides are connected to RID-F. Conjugates with peptides

Full text of DOI:10.1016/j.ejmech.2018.01.045

European Journal of Medicinal Chemistry 146 (2018) 636e650
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Ridaifen-F conjugated with cell-penetrating peptides inhibits
intracellular proteasome activities and induces drug-resistant cell
death
Makoto Tanaka a, Yunhao Zhu a, Masafumi Shionyu a, Nozomi Ota b, Natsumi Shibata b,
Chihiro Watanabe b, Akihito Mizusawa b, Ryuzo Sasaki a, Tamio Mizukami a,
,
, *
Isamu Shiina b **, Makoto Hasegawa a
a Faculty of Bioscience, Nagahama Institute of Bio-Science and Technology, 1266 Tamura-cho, Nagahama, Shiga 526-0829, Japan
b Department of Applied Chemistry, Faculty of Science, Tokyo University of Science, 1-3 Kagurazaka, Shinjuku-ku, Tokyo 162-8601, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Ridaifen-F (RID-F) potently inhibits proteolytic activities of the 20S proteasome but poorly inhibits those
of the 26S proteasome. Here, we report preparation of several conjugates in which various peptides are
connected to RID-F. Conjugates with peptides consisting of seven amino acid residues significantly
inhibited the 26S proteasome. Particularly, RID-F conjugated to an octaarginine peptide (R8, a so-called
cell-penetrating peptide) inhibited intracellular proteasome activities and induced cell death in drug-
resistant KMS-11 myeloma cells. RID-F conjugated to hydrophobic peptides also inhibited the 26S pro-
teasome but failed to induce cell death, suggesting poor penetration into cells. We infer that the R8
peptide has dual functions: (1) rapid penetration of conjugates into the cell increases intracellular drug
concentrations sufficient for exhibition of its effect, and (2) recognition of the conjugates by the 26S
Received 27 October 2017
Received in revised form
27 December 2017
Accepted 15 January 2018
Available online 31 January 2018
Keywords:
Proteasome inhibitor
Tamoxifen derivative
Cell-penetrating peptide
Docking study
proteasome stimulates drug entry into the catalytic chamber. In the presence of ATPgS, RID-F conjugates
containing R8 inhibited the 26S proteasome more potently than in the presence of ATP, suggesting
efficient entry of drugs into the catalytic chamber in a similar fashion to the substrate. Taken together
with docking simulations of RID-F conjugate interactions with proteasome active sites, the second
function of R8 peptide is plausible. Thus, the conjugation of nonpeptidic proteasome inhibitors to a cell-
penetrating peptide could represent a viable strategy for overcoming the drug-resistance of tumor cells.
© 2018 Elsevier Masson SAS. All rights reserved.
1. Introduction
(b1e7) that form respective heptameric ring structures (
a-ring and
b
-ring). Two -rings and two -rings are stacked in an abba
a
b
In eukaryotic cells, protein degradation by the ubiquitin-
proteasome system plays a pivotal role in the regulation of various
cellular processes including proliferation, differentiation, apoptosis,
gene transcription, signal transduction, and quality control of
nascent proteins [1]. The 26S proteasome, which is constitutively
present in eukaryotic cells, comprises a barrel-shaped core particle
(20S proteasome) and the 19S regulatory particles (RPs) capping one
or both ends of the 20S proteasome barrel [2e4]. The 20S protea-
arrangement, generating the barrel-like structure of the 20S pro-
teasome. The proteolytic active sites are located inside the barrel,
where the
tivities with distinct substrate specificities:
peptide hydrolase (PGPH), 2 for trypsin-like (T-L), and
chymotrypsin-like (CT-L) activities [5,6].
b
1,
b
2, and
b
5 subunits are responsible for protease ac-
1 for peptidylglutamyl
5 for
b
b
b
The 19S RPs regulate translocation of substrates into the 20S
proteasome barrel. There are two types of 19S RP sub-complexes,
known as the lid and the base, which contain different subunits
with different functions. Deubiquitylation and unfolding of sub-
strates are required prior to translocation of substrates into the
catalytic chamber. The lid removes ubiquitin from poly-
ubiquitinated substrates [7,8], while the base unfolds the globular
domains of substrates [9,10]. One of the components in the base, a
some consists of seven
a-subunits (a1e7) and seven b-subunits
* Corresponding author.
** Corresponding author.
E-mail addresses: shiina@rs.kagu.tus.ac.jp (I. Shiina), m_hasegawa@nagahama-i-
bio.ac.jp (M. Hasegawa).
https://doi.org/10.1016/j.ejmech.2018.01.045
0223-5234/© 2018 Elsevier Masson SAS. All rights reserved.

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
637
heterohexameric ring with AAA-ATPase modules, plays a critical
role in substrate translocation into the 20S proteasome barrel and
substrate unfolding. ATP binding to the base opens the gate for
substrates to be translocated into the barrel, and subsequent hy-
drolysis of the bound ATP is required for substrate unfolding
mediated by reverse chaperone or unfoldase associated with AAA-
ATPases.
Disruption of the ubiquitin-proteasome system causes accu-
mulation of incompatible regulatory proteins that trigger an
apoptotic cascade, eventually leading to growth arrest and cell
death. In comparison with normal cells, tumor cells are generally
more sensitive to the proapoptotic effects of proteasome inhibition
[11]. This was ascertained in multiple myeloma cells whose growth
pharmaceutical interest. Studies of conjugating CPPs to several
drugs, including methotrexate [26], doxorubicin [27e31], and
organometal conjugates [32e34], have been reported. Tat peptide
conjugated to doxorubicin significantly increases the intracellular
concentration of the drug in both drug-sensitive and drug-resistant
cell lines [29].
In this study, we prepared conjugates of RID-F and various
peptides, including oligo-arginines, and compared their inhibitory
potency on the protease activities of the 20S and 26S proteasomes.
Notably, RID-F conjugates containing R8 showed a strong inhibitory
potency toward 26S proteasome activity in comparison with that of
free RID-F. Furthermore, these RID-F conjugates overcame KMS-
11 cell drug resistance, rapidly inducing cell death. We confirmed
that these conjugates inhibited proteolytic activities of the intra-
cellular proteasome. These results suggest that conjugation of
nonpeptidic PIs with CPPs is a viable strategy to overcome tumor-
cell drug resistance.
strongly relies on the transcription factor NF
B to NF B in the cytoplasm prevents translocation of NF
the nucleus. Stress-induced phosphorylation of I B results in
proteasome-dependent degradation of I B and thus NF B activa-
tion. Inhibition of the proteasome likely represses myeloma cell
growth by maintaining NF B in an inactive state. Thus, proteasome
kB [12,13]. Binding of
Ik
k
k
B into
k
k
k
k
2. Results
inhibitors (PIs) have emerged as a new class of molecular-targeted
anticancer drugs [14].
2.1. Synthetic procedures
Several PIs are in clinical use as anticancer drugs [13,14]. One
inhibitor, the peptide boronate bortezomib, is approved for the
treatment of multiple myeloma and mantle cell lymphoma [15,16].
The toxic boronate pharmacophore, however, is associated with
severe adverse effects, which is why an epoxomicin analog, tetra-
peptide epoxyketone carfilzomib, was developed and approved
[17]. More recently, another peptide, boronate ixazomib, which is
available for oral uptake, has been approved for the treatment of
multiple myeloma in combination with two other drugs, lenalido-
mide and dexamethazone, after first-line therapy [18]. In addition
to the treatment of hematological malignancies, PIs have been
proposed as promising drug candidates for solid tumors, inflam-
mation, immune diseases, ischemic stroke, and tuberculosis [19].
This situation encourages the development of new types of PIs with
improved efficacy and fewer adverse effects [20].
To find new noncovalent and nonpeptidic PIs, we previously
examined proteasome inhibition using a series of tamoxifen de-
rivatives and found that ridaifen (RID)-A, -B, -D, and -F inhibited
protease activities of the 20S proteasome [21]. RID-F (1) was the
most potent inhibitor of the three different enzymatic activities.
However, subsequent investigation indicated that RID-F derivatives
were not effective against multidrug-resistant human multiple
myeloma cells (KMS-11). Drug resistance can be caused by poor
drug uptake, active excretion, intracellular degradation or modifi-
cation, concentration of drugs into a specific subcellular compart-
ment, and mutations that render target proteins insensitive to
drugs. Although it is difficult to identify the mechanisms that give
rise to drug resistance in individual cells, it is likely that poor drug
uptake or active excretion is a prevalent mechanism, and therefore
modification of drugs to improve uptake is a promising strategy.
Several proteins, such as Tat protein from human immunodefi-
ciency virus-1 (HIV-1), can spontaneously penetrate cells [22]. The
minimal domain needed for penetration was defined as a short
sequence of 10e16 amino acids, termed a “cell-penetrating pep-
tide” (CPP) [23]. Native Tat peptide is composed of several cationic
amino acids, including six arginine and two lysine residues.
Furthermore, the cationic peptide, octaarginine (R8), can promote
cellular uptake of conjugated compounds. Although several hy-
potheses explaining how these peptides penetrate cells have been
presented [24,25], the underlying mechanism still remains unclear.
Nevertheless, CPPs can be used for the translocation of peptides,
proteins, ribonucleic acids, oligonucleotide mimics, and nano-
particles through cell membranes. Since CPPs can also transport
nonpermeable drugs into living cells, they are of considerable
We previously presented the synthesis and characterization of a
novel tamoxifen derivative, RID-F (1), which inhibits 20S protea-
some activity. RID-F was synthesized chemically as described in
previous reports [21,35,36]. RID-F derivatives (RID-F-COOH (15)
and RID-F-CH2NH2 (19)) were prepared by the Mukaiyama reduc-
tive coupling reaction [37,38], as shown in Schemes 1 and 2. The
1,3-phenylenedimethanol (2) was first mono-protected with a p-
methoxybenzyl group and the following two-carbon elongation of
the intermediate 4 to afford 1-(30-((p-methoxybenzyloxy)methyl)
phenyl)propan-1-one (6). Then, 6 was treated with an excess
amount of 4,40-dihydroxybenzophenone (7) in the presence of the
low-valent titanium species generated from titanium(IV) chloride
with zinc powder to afford the desired cross-coupling product 8 in
52% yield. The phenol moieties in 8 were temporarily protected as
tert-butyldimethylsilyl (TBS) groups and finally transformed into
the corresponding aminoethyl ethers by alkylation of 14. Fortu-
nately, the benzyl protective group in the tetra-substituted olefin 14
was simultaneously cleaved in the alkylation reaction under
nucleophilic conditions to provide the desired RID-F derivative 15
including the free carboxylic acid part.
Next, RID-F derivative 19, including the free primary amine part,
was also synthesized using a similar procedure as described in
Scheme 2. Aminoethyl groups were introduced into the bisphenol
17 in 77% yield, and the desired free primary amine 19 was suc-
cessfully prepared from benzyl oxime 18 by reduction using lithium
aluminum hydride in 77% yield.
Peptides and the RID-F derivatives were linked at the N- or C-
terminus of the peptides by amide bond formation (Scheme 3). In
the peptides used for the conjugation, the opposite terminus to the
site linked with the RID-F derivatives was blocked by modification
with an amide (C-terminus) or acetyl group (N-terminus). The
chemical structures of representative compounds, RID-F (1), RID-F-
GR8 (27), and R8G-RID-F (31), are depicted in Fig. 1. The purities of
all the conjugates were confirmed by analytical reversed-phase
high-performance liquid chromatography (RP-HPLC) analysis and
matrix-assisted laser desorption/ionization-time of flight (MALDI-
TOF) mass measurements, and were above 95% (Supplementary
Fig. 4). The m/z values of these compounds were observed to be
identical with the theoretical values (Supplementary Table 1).
2.2. Proteasome inhibition by RID-F-CPP conjugates
To assess the potency of proteasome inhibition of RID-F (1) and

638
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
Scheme 1. Synthesis of RID-F-COOH (15).

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
639
Scheme 2. Synthesis of RID-F-CH2NH2 (19).
its CPP conjugates (RID-F-GR8 (27) and R8G-RID-F (31)), three
proteolytic activities (PGPH, T-L, and CT-L) were examined with an
enzymatic kinetics assay using fluorogenic substrates. To promote
the passage of peptidyl substrates to the catalytic chamber, sodium
dodecyl sulfate (SDS) was present at 0.03% for assaying the 20S
proteasome and ATP was present at 1 mM for assaying the 26S
proteasome. The inhibition curve of each compound is depicted in
Supplementary Fig. 1. Their IC50 values are summarized in Table 1.
Free RID-F inhibited PGPH and CT-L activities of the 20S pro-
teasome, but did not inhibit the three activities of the 26S protea-
some. RID-F-GR8 (27) and R8G-RID-F (31) inhibited T-L activity of
the 20S proteasome. Although these two conjugates containing R8
did not inhibit the CT-L activity of the 20S proteasome, they
strongly inhibited the CT-L activity of the 26S proteasome. Thus, the
inhibitory potency and specificity of the RID-F conjugates alter
depending on the proteasome type (20S and 26S). Such alterations
are also found with conventional inhibitors such as bortezomib,
vinblastine, and lactacystin. The gate of the 20S proteasome
chamber is normally closed so that substrates and PIs cannot enter
the chamber freely [39]. The presence of SDS enables ingress of
these substrates and inhibitors into the chamber [40]. On the other
hand, the AAA ring of the 19S RP in the 26S proteasome forms a
channel through which these substances enter the 20S chamber in
an ATP-dependent manner. Alterations in the potency and
specificity of PIs are due to structural differences in the two types of
proteasomes; however, structural information alone is insufficient
for delineating the molecular mechanisms underlying these alter-
ations. Regardless of the mechanism, the marked inhibition of the
CT-L activity of the 26S proteasome by RID-F conjugates containing
R8 encouraged further study.
Basic peptides with 4e10 R residues inhibit CT-L and PGPH ac-
tivities of the 20S proteasome at submicromolar levels [41]. How-
ever, free GR8 (C-terminal amidated) or R8G peptides (N-terminal
acetylated) did not inhibit the protease activities of the 26S pro-
teasome (Table 1), indicating that RID-F-CPP conjugate-mediated
inhibition of the 26S proteasome was not attributable to the basic
peptides in the conjugates.
ATP is typically added for assaying the 26S proteasome. It is of
interest to examine whether ATP hydrolysis is required for degra-
dation of small substrates (in our assay) by the 26S proteasome.
ATP
used to address this question. As shown in Fig. 2A, 26S proteasome
CT-L activity was higher in the presence of ATP S than ATP, while
gS, which is hydrolyzed at a much slower rate than ATP, was
g
there was no effect on 20S proteasome CT-L activity. Thus, ATP
hydrolysis is not required for 26S proteasome CT-L activity for small
substrates. Next, the effect of ATP
RID-F conjugates was examined. Interestingly, in the presence of
ATP S, both RID-F-GR8 (27) and R8G-RID-F (31) showed stronger
gS on the inhibitory potency of
g

640
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
Scheme 3. Conjugation of RID-F derivatives and peptides at the N-terminus (A) or C-terminus (B).

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
641
Fig. 1. Structure of RID-F and CPP conjugates.
Table 1
IC50 values of CPP-conjugated RID-F against the human 20S and 26S proteasome.
20S IC50 M)
(
m
26S IC50
(mM)
b
1 (PGPH)
b
2 (T-L)
b
5 (CT-L)
b1 (PGPH)
b2 (T-L)
b5 (CT-L)
RID-F (1)
0.72 0.10
>50
>40
>8
>8
0.7 0.08
4.28 0.54
>8.0
>100
1.02 0.02
>50
>40
>8
>8
0.0005 0.0004
4.58 0.43
0.16 0.01
>100
>50
>40
>40
>40
0.019 0.005
>40
>40
>100
>100
RID-F-GR8 (27)
R8G-RID-F (31)
H-GR8-NH2
Ac-R8G-OH
Bortezomib
Vinblastine
Lactacystin
1.2 0.3
3.5 0.8
>8
45
4
5.5 0.2
9.4 1.2
>40
>40
2.1 0.5
>40
>40
>40
>40
>40
14.50 2.29
>40
>8
1.93 0.29
4.90 0.34
3.25 0.54
4.96 0.17
Abbreviations: IC50, half-maximum inhibitory concentration; CT-L, chymotrypsin-like activity; T-L, trypsin-like activity; PGPH, peptidylglutamyl peptide hydrolase activity.
Data are given as means SD (n ¼ 3).
inhibition than in the presence of ATP; IC50 values of RID-F-GR8 and
R8G-RID-F against 26S proteasome CT-L activity significantly shif-
activities of the 20S and 26S proteasomes. The potency of nine
types of RID-F-peptide conjugates, including basic and hydrophobic
residues, was examined. Inhibition curves are shown in
Supplementary Fig. 2. The IC50 values are summarized in Table 2.
When the 20S proteasome was examined, RID-F-peptide con-
jugates with different lengths of arginine repeats, RID-F-GR4 (25),
RID-F-GR6 (26), R4G-RID-F (29), and R6G-RID-F (30), showed sig-
nificant inhibition of T-L activity. From their IC50 values, the N-
terminal peptide conjugates, RID-F-GR4 (25) and RID-F-GR6 (26),
were approximately 10 times more potent than C-terminal peptide
conjugates, R4G-RID-F (29) and R6G-RID-F (30). Three hydrophobic
ted to lower values of 2.4
C). Free RID-F did not inhibit CT-L activity in the presence of either
ATP S or ATP (Fig. 2D).
mM and 6.0 mM, respectively (Fig. 2B and
g
2.3. Proteasome inhibition by RID-F conjugates containing various
peptides
We evaluated how the sequence and length of peptides conju-
gated to RID-F affected the inhibitory potency for proteolytic

642
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
(A)
(B)
RID-F-GR8
ATP
ATPγS
100
26S ATPγS
26S ATP
50
20S ATPγS
20S ATP
0
30000
25000
20000
15000
10000
5000
0
10-7
0.1
10-6
10-5
10
10-4
100
1
26S ATPγS
Concentration (M)
(C)
R8G-RID-F
26S ATP
ATP
ATPγS
100
20S ATP
20S ATPγS
50
0
10-7
0.1
10-6
10-5
10
10-4
100
1
Concentration (M)
(D)
80
0
20
40
60
RID-F
Time (min)
ATP
ATPγS
100
50
0
10-7
0.1
10-6
10-5
10
10-4
100
1
Concentration (M)
Fig. 2. (A) Kinetics of fluorogenic substrate degradation in the presence of ATP or ATP
gS. Values are means SEM of three independent experiments. (BeD) Inhibition of the 26S
proteasome in the presence of ATP (blue circles) or ATP S (red squares). Results are means SEM of three independent experiments. (For interpretation of the references to color in
g
this figure legend, the reader is referred to the Web version of this article.)
Table 2
Inhibition of the human 20S and 26S proteasome by peptide-conjugated RID-F.
20S IC50
(
mM)
26S IC50
(mM)
b1 (PGPH)
b2 (T-L)
b5 (CT-L)
b
1 (PGPH)
b2 (T-L)
b5 (CT-L)
RID-F-GR4 (25)
RID-F-GR6 (26)
R4G-RID-F (29)
R6G-RID-F (30)
RID-F-GLLE (20)
RID-F-GLLVY (21)
LLVYG-RID-F (28)
RID-F-GLLVYGG (22)
RID-F-GLLLGGGG (24)
>8
>8
>8
3.3 1.0
4.5 0.3
0.89 0.08
0.28 0.01
0.99 0.04
0.89 0.08
0.35 0.06
0.20 0.10
3.7 0.3
3.2 0.2
>8
8.9 1.2
>8
6.9 0.8
3.8 0.6
>8
>8
>8
>8
6.1 1.1
0.89 0.09
0.37 0.14
0.89 0.16
1.1 0.2
>40
>40
>40
>40
>40
33
>40
22
17
>40
>40
>40
>40
>40
>40
>40
>40
32
>40
>40
>40
>40
24
26
>40
9.1 1.0
5.0 0.9
1
3
9
7
7
3
Abbreviations: IC50, half-maximum inhibitory concentration; CT-L, chymotrypsin-like activity; T-L, trypsin-like activity; PGPH, peptidylglutamyl peptide hydrolase activity.
Data are given as means SD (n ¼ 3). Inhibition of the 20S core particle and 26S proteasome was evaluated in concentration ranges of 0.5e8 M and 0.1e40 M, respectively.
m
m
peptide conjugates, RID-F-GLLE (20), RID-F-GLLVY (21), and LLVYG-
RID-F (28), strongly inhibited CT-L and PGPH activities, but these
activities were insensitive to RID-F conjugates that contained R.
Thus, the chemical properties of the peptides present in the RID-F
conjugate influence the inhibitory specificity for the 20S protea-
some; conjugates that contain basic peptides favorably inhibit T-L
activity, and those that contain hydrophobic peptides do so for CT-L
and PGPH activities. Notably, all of the conjugates that contained
less than six amino acids showed minimal inhibition of the 26S
proteasome. However, conjugates with longer peptides, such as
RID-F-GLLVYGG (22) and RID-F-GLLLGGGG (24), showed significant
inhibition of the 26S proteasome. Taken together with the results of
RID-F-GR8 (27) and R8G-RID-F (31), conjugation with a peptide
consisting of seven or more amino acid residues was required for
significant inhibition of 26S proteasome activities. This require-
ment suggests that the peptide portion of the RID-F conjugate in-
teracts with the 26S proteasome, although the site of interaction
remains to be studied.
2.4. Cellular proteasome inhibition by RID-F conjugates and their
cytotoxic effect on multidrug-resistant myeloma cells
Since RID-F CPP conjugates, but not free RID-F, inhibited the 26S
proteasome (see Table 1), we examined the effects of the conjugate

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
643
on the cellular proteasome using two cell lines, multidrug-resistant
human multiple myeloma cells (KMS-11) and drug-sensitive mul-
tiple myeloma cells (RPMI8226), and a Proteasome-Glo™ Assay kit
(Promega Co.), which enables the direct measurement of intracel-
lular proteasome activities in cultured cells [42]. Multidrug-
resistance of KMS-11 cells was confirmed by using cisplatin
(Table 3). Inhibition of three proteasome activities by RID-F (1),
RID-F-GR8 (27), and R8G-RID-F (31) was measured in the presence
of various concentrations of the conjugates, and IC50 values were
calculated from inhibition curves (Supplementary Fig. 3). Free RID-
F did not inhibit CT-L, PGPH, and T-L activities of the intracellular
proteasome in KMS-11 cells, whereas R8G-RID-F and RID-F-GR8
significantly inhibited all three activities to a similar extent; IC50
values were 2e8 mM (Table 3). On the other hand, RID-F-GR8, and
R8G-RID-F comparably inhibited CT-L and PGPH activities in
RPMI8226 cells, but RID-F inhibited only CT-L activity with slightly
lower potency.
Fig. 3. Time dependency of KMS-11 cell viability in the presence of RID-F (blue circles),
RID-F-GR8 (red squares), and R8G-RID-F (green triangles) (10 mM) determined with an
Similarly, cell survival rates were measured by quantifying ATP
levels to determine Cyt50 values for each conjugate. As shown in
Table 3, RID-F-GR8 and R8G-RID-F induced cell death of KMS-11 and
ATP assay. The viability of mock-treated cells served as a standard (100%). Results are
means SEM of three independent experiments. (For interpretation of the references
to color in this figure legend, the reader is referred to the Web version of this article.)
RPMI8226 cells with similar potency (Cyt50 ¼ 4e8
mM). Remark-
ably, these RID-F conjugates exerted their inhibitory effects on the
proteasome and cell survival with similar potency. Fig. 3 shows
time-dependent death of KMS-11 cells cultured in the presence of
of Arg entered the S1 pocket instead of the guanidyl group (Fig. 4B).
In the previous study, it was suggested that the vinyl benzene group
of RID-F or the homopiperidine ring of some RID-F derivatives is
10 mM of each compound. Viable cell counts were markedly
important for interacting with the S1 pocket of the b1 subunit [21].
decreased upon culture for 6 h in the presence of RID-F-GR8 or R8G-
RID-F, whereas the effect of free RID-F was undetectable. On the
other hand, RID-F, RID-F-GR8, and R8G-RID-F induced comparable
levels of cell death in RPMI8226 cells.
The estimated binding modes of RID-F-GR and RG-RID-F in the
present study suggest that a different manner of homopiperidine
ring usage is important for interactions between CPP-conjugated
RID-F and the proteasome. Moreover, although the main chain di-
rection of the amino acid moiety was different between RID-F-GR
and RG-RID-F, the estimated binding modes were somewhat
similar between each other; the substrate-binding cleft composed
of the S1 to S5 pockets was almost buried due to the interactions of
the amino acid moiety with the S1 pocket and the RID-F moiety
2.5. Molecular docking simulations
To determine how RID-F-peptide conjugates bind the protea-
some to exhibit their inhibitory effect, docking studies using
AutoDock Vina [43] were conducted. To date, several three-
dimensional (3D) structures of the human proteasome have been
reported. We used structural data of the human 20S proteasome
complexed with dihydroeponemycin (Protein Data Bank ID: 5lf1
[44]). Estimated binding modes of RID-F-GR (Fig. 4A) and RG-RID-F
(Fig. 4B) were obtained from clusters composed of 23 and 26
docking poses, respectively (see Experimental Protocols for details).
The estimated binding affinities of RID-F-GR and RG-RID-F
were ꢀ9.1 (the second highest value in 100 docking poses)
and ꢀ9.5 kcal/mol (the highest value in 100 docking poses),
respectively. In the estimated binding mode of RID-F-GR, the gua-
nidyl group of the Arg residue in RID-F-GR entered the S1 pocket
with the b6 subunit. This binding mode prevented substrates from
interacting with the binding cleft widely and mediated the inhib-
itory effect of CPP-conjugated RID-F.
3. Discussion
RID-F, a novel tamoxifen derivative, inhibited the human 20S
proteasome [21]. Our subsequent studies of the 26S proteasome,
however, showed that RID-F poorly inhibited its proteolytic activ-
ities. Furthermore, RID-F failed to kill KMS-11 cells, a multidrug-
resistant human myeloma cell line. To overcome these drawbacks
of RID-F, we prepared various RID-F-peptide conjugates. This
concept was based on the findings of CPPs (see Introduction). In this
paper, we hypothesized the dual functions of peptide moieties in
RID-F-peptide conjugates with respect to inhibition of the cellular
[19] of the
with the
rings also interacted with similar sites in the
b
b
5 subunit, and the two homopiperidine rings interacted
6 subunit (Fig. 4A). For RG-RID-F, two homopiperidine
6 subunit to the ones
b
used for RID-F-GR binding, but the amino group in the main chain
Table 3
Proteasome inhibition and cytotoxicity in KMS-11 cells and RPMI8226 cells.
KMS-11
RPMI8226
IC50
CT-L
N.D.
6.3 1.6
2.1 0.7
N.D.
(
m
M)
Cyt50
N.D.
7.2 1.0
4.5 0.1
N.D.
(
m
M)
IC50
(
m
M)
Cyt50 (mM)
PGPH
T-L
CT-L
PGPH
T-L
RID-F (1)
N.D.
N.D.
11.0 0.8
5.7 0.2
2.1 0.6
N.D.
N.D.
e
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
e
3.6 0.6
4.3 0.3
2.7 0.4
N.D.
N.D.
1.18 0.88
RID-F-GR8 (27)
R8G-RID-F (31)
H-GR8-NH2
Ac-R8G-OH
cisplatin
4.8 0.8
2.8 1.2
N.D.
N.D.
e
8.0 1.8
4.1 0.3
N.D.
N.D.
e
6.6 0.2
3.1 0.1
N.D.
N.D.
e
N.D.
e
N.D.
N.D.
Abbreviations: N.D., not detected; IC50, half-maximum inhibitory concentration; CT-L, chymotrypsin-like activity; T-L, trypsin-like activity; PGPH, peptidylglutamyl peptide
hydrolase activity; -, not determined. Data are given as means SD (n ¼ 3).

644
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
Fig. 4. Estimated binding modes of RID-F-GR (A) and RG-RID-F (B) with the CT-L site of the human proteasome. The
b5 and b6 subunits of the human proteasome are shown as a
surface model colored in cyan and green, respectively. Thr1 of the 5 subunit is colored in red. The docked RID-F-GR and RG-RID-F are shown as a ball-and-stick model with carbon,
b
nitrogen, and oxygen atoms in gray, blue, and red, respectively. Orange lines indicate the boundaries of the docking grid box for the AutoDock Vina run. All images were generated
using UCSF Chimera 1.11.2 [56]. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
proteasome. Firstly, peptide moieties are recognized by the pro-
teasome as substrates, thereby facilitating the entry of conjugates
into the catalytic chamber. Secondly, peptide moieties promote
penetration of conjugates into cells, thereby increasing intracellular
concentrations of RID-F to levels sufficient for proteasome inhibi-
tion. We found that RID-F conjugates containing R8 inhibited not
only the 26S proteasome but also the intracellular proteasome in
KMS-11 cells. Furthermore, these compounds induced cell death.
Interestingly, although these conjugates inhibited 26S proteasome
CT-L activity only, they inhibited all three proteolytic activities in
KMS-11 cells. The reason is unclear, but may involve unknown
intracellular factor(s) that are missing in the isolated 26S
proteasome.
MRP2 [49,50]. MRP2 is an efflux transporter, and PDZK1 mediates
subcellular localization of target proteins (localization of MRP2 into
the plasma membrane in this case). In our experiments, KMS-
11 cells were not sensitive to RID-F and cisplatin, in contrast to
RPMI8226 cells, which were. This suggests that KMS-11 cells
possess the multidrug resistance previously proposed [48]. RID-F-
GR8 (27) and R8G-RID-F (31) not only inhibited proteasomal activity
in KMS-11 cells but also rapidly induced cell death. Taken together
with the 26S proteasome data, R8 peptide in the RID-F conjugates
appears to have dual functions: stimulation of drug entry into the
proteolytic chamber and stimulation of cellular uptake. Further-
more, R8 may contribute to increased drug access to the CT-L active
site. The presence of 1 mM of the RID-F conjugates, however, was
The inhibitory potencies of RID-F-peptide conjugates against the
26S proteasome were dependent on the lengths and chemical
properties of the peptides. Peptides consisting of at least seven
amino acids were necessary for inhibition. Notably, hydrophobic
peptides inhibited 26S CT-L activity with potencies comparable to
those of R8. However, RID-F conjugates containing hydrophobic
peptides did not influence cell survival (data not shown), suggest-
ing that these conjugates do not enter the cell efficiently. Never-
theless, this is a useful finding to be taken into consideration for
further improvement of RID-F conjugates.
poorly cytotoxic against KMS-11 cells, indicating that further
structural improvement of the RID-F conjugates is required.
Our results collectively suggest the proposed mechanism in
Fig. 5. Structural properties of RID-F (1) suggest that it can diffuse
efficiently across the cell membrane. RID-F may act on the 20S form
free from the 19S RPs, but does not significantly inhibit the major
intracellular form, the 26S proteasome. Furthermore, RID-F (1) may
be readily exported from cells, thus leading to poor efficacy against
KMS-11 cells. RID-F-GR8 (27) and R8G-RID-F (31) are rapidly taken
up by cells and efficiently translocated into the 26S proteasome.
Within the proteasome proteolytic chamber, the peptide moiety
may be cleaved, and RID-F conjugated to one or two amino acids
mainly binds to the CT-L site (see Fig. 4). With the rapid increase in
the number of CPP studies, drug delivery systems using CPPs that
target tumor cells with high potency will likely become available
[23,51]. However, for CPPs to be effective in vivo, it will be necessary
to develop strategies that prevent cleavage of the conjugated
peptides by blood proteases. This could be achieved by chemically
modifying the peptide structures by, for example, macrocyclization,
Our experimental results with the 26S proteasome and ATPgS
clearly indicate that 19S RPs are important for substrate entry into
the CT-L active site, although ATP hydrolysis is not required.
Instead, ATP hydrolysis may impede entry, because higher CT-L
activity was seen in the presence of ATPgS than ATP. ATP hydroly-
sis would be required for unfolding of globular domains of large
proteins before degradation. This is consistent with cryo-electron
microscopic analyses of the 26S proteasome in the presence of
ATP
unfolded proteins requires ATP binding but not ATP hydrolysis. The
presence of ATP S instead of ATP may retain the 19S RP in an active
gS [39,45e47]; 26S proteasomal degradation of peptides and
or by incorporating D-amino acids and N-methyls [52]. Studies are
g
underway in our laboratory to develop highly potent and tumor-
specific RID derivatives using random peptide library techniques
and optimization of chemical structures.
conformation, opening the substrate entry gate for a prolonged
time.
Unexpectedly, RID-F-GR8 and R8G-RID-F showed higher inhibi-
tion of 26S CT-L activity in the presence of ATPgS than ATP. At
4. Experimental protocols
present, the precise reason is unclear, but it is plausible that entry of
these compounds into the 20S chamber is stimulated in a similar
fashion to that of the substrate, and somehow, they are accessible to
the CT-L active site with high efficiency. Gibbs free energy changes
in our molecular docking simulations indicated that RID-F-GR and
RG-RID-F preferably bind the CT-L active site (data not shown).
Previous reports indicate that KMS-11 cells are resistant to
several anticancer drugs [48] via overexpression of PDZK1 protein,
which interacts with a multidrug resistance-associated protein
4.1. Chemical synthesis
RID-F was synthesized according to the methods described in
our previous reports [21,35,36]. The synthetic pathways for pro-
ducing the new compounds are depicted in Schemes 1 and 2.
Substituted RID-F derivatives RID-F-COOH (15) and RID-F-CH2NH2
(19), which could be used as precursors of peptide conjugates, were
prepared using the Mukaiyama reductive coupling reaction [37,38]

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
645
Fig. 5. Schematic illustration of the proposed mechanism via which peptide conjugates overcome multidrug resistance of KMS-11 cells. Abbreviations: CP, core particle; CPP, cell-
penetrating peptide; CT-L, chymotrypsin-like activity.
and the introduction of aminoethyl ether moieties possessing
azepanyl groups. The synthesis procedure of each compound is
described below.
the mixture and evaporation of the solvent, the crude product was
obtained.
To a solution of the crude product in THF (17.6 mL) was added
ethylmagnesium bromide (1.0 M, 13.2 mL, 13.2 mmol) at 0 ꢁC. After
the reaction mixture had been stirred for 2 h at room temperature,
saturated aqueous ammonium chloride was added at 0 ꢁC. The
mixture was extracted with ethyl acetate, and the organic layer was
washed with water and brine, and dried over sodium sulfate. After
filtration of the mixture and evaporation of the solvent, the crude
product was purified by column chromatography on silica (eluant:
hexane/ethyl acetate ¼ 10/1) to afford 5 (2.38 g, 86%).
4.1.1. (3-((40-Methoxybenzyloxy)methyl)phenyl)methanol (3)
To
a
solution of 1,3-phenylenedimethanol (2) (3.51 g,
25.4 mmol) in tetrahydrofuran (THF; 84.4 mL) at 0 ꢁC was added
55% sodium hydride (dispersion in paraffin liquid, 1.11 g,
25.4 mmol). After dropping 4-methoxybenzyl chloride (3.43 mL,
25.4 mmol), tetrabutylammonium iodide (1.00 g, 2.71 mmol) was
added. The reaction mixture was stirred for 1 h at room tempera-
ture and for 3 h at 60 ꢁC, and then saturated aqueous ammonium
chloride was added at 0 ꢁC. The mixture was extracted with ethyl
acetate, and the organic layer was dried over sodium sulfate. After
filtration of the mixture and evaporation of the solvent, the crude
product was purified by column chromatography on silica (eluant:
hexane/ethyl acetate ¼ 6/1) to afford 3 (2.15 g, 33%).
1H NMR (CDCl3, 500 MHz):
d 7.35e7.28 (6H, m, Ar), 6.91e6.88
(2H, m, Ar), 4.61 (1H, td, J ¼ 7.0, 3.5 Hz, 1-H), 4.53 (2H, s, Bn), 4.50
(2H, s, Bn), 3.81 (3H, s, OCH3), 1.87e1.71 (3H, m, 2-H and OH), 0.92
(3H, t, J ¼ 7.0 Hz, 3-H); 13C NMR (CDCl3, 500 MHz):
d
159.20 (400),
144.79, 138.50, 130.26, 129.42, 128.43, 126.92, 125.34, 125.21, 113.78
(300), 75.89 (1), 71.86 (PMPCH2 or PMBOCH2), 71.77 (PMPCH2 or
PMBOCH2), 55.24 (CH3O), 31.84 (2), 10.13 (3); HR MS: calcd for
1H NMR (CDCl3, 300 MHz):
d
7.37e7.28 (6H, m, Ar), 6.92e6.88
(2H, m, Ar), 4.71 (2H, s, 1-H), 4.53 (2H, s, Bn), 4.51 (2H, s, Bn), 3.81
(3H, s, OCH3); 13C NMR (CDCl3, 300 MHz): 159.17 (40), 141.09,
C
18H22O3Na [MþNa]þ 309.1461, found 309.1457.
d
138.58, 130.17, 129.39, 128.53, 126.95, 126.27, 126.14, 113.77 (30),
71.85 (PMPCH2 or PMBOCH2), 71.66 (PMPCH2 or PMBOCH2), 65.04
(CH2OH), 55.21 (CH3O); HR MS: calcd for C16H18O3Na [MþNa]þ
281.1148, found 281.1160.
4.1.3. 1-(30-((400-Methoxybenzyloxy)methyl)phenyl)propan-1-one
(6)
To a solution of DMSO (0.45 mL, 6.88 mmol) in dichloromethane
(29.4 mL) was added oxalyl chloride (0.45 mL, 5.16 mmol)
at ꢀ78 ꢁC. The reaction mixture was stirred for 15 min at ꢀ78 ꢁC,
and then a solution of 5 (937 mg, 3.27 mmol) in dichloromethane
(5.0 mL) was added. After the reaction mixture had been stirred for
15 min, triethlyamine (1.44 mL, 10.3 mmol) was added. After the
reaction mixture was stirred for 2 h at room temperature, water
was added at 0 ꢁC. The mixture was extracted with dichloro-
methane, and the organic layer was washed with water and brine,
and dried over sodium sulfate. After filtration of the mixture and
evaporation of the solvent, the crude product was purified by col-
umn chromatography on silica (eluant: hexane/ethyl acetate ¼ 6/1)
to afford 6 (930 mg, 96%).
4.1.2. 1-(30-((400-Methoxybenzyloxy)methyl)phenyl)propan-1-ol (5)
To a solution of dimethyl sulfoxide (DMSO; 1.16 mL, 17.7 mmol)
in dichloromethane (78.0 mL) was added oxalyl chloride (1.15 mL,
13.2 mmol) at ꢀ78 ꢁC. The reaction mixture was stirred for
20 min at ꢀ78 ꢁC, and then a solution of 3 (2.15 g, 8.32 mmol) in
dichloromethane (10.0 mL) was added. After the reaction mixture
had been stirred for 15 min, triethlyamine (3.68 mL, 26.4 mmol)
was added. After the reaction mixture was stirred for 3 h at room
temperature, water was added at 0 ꢁC. The mixture was extracted
with dichloromethane, and the organic layer was washed with
water and brine, and dried over sodium sulfate. After filtration of
1H NMR (CDCl3, 500 MHz):
d 7.94 (1H, s, Ar), 7.90e7.88 (1H, m,

646
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
Ar), 7.57e7.55 (1H, m, Ar), 7.46e7.42 (1H, m, Ar), 7.32e7.28 (2H, m,
Ar), 6.92e6.88 (2H, m, Ar), 4.57 (2H, s, Bn), 4.52 (2H, s, Bn), 3.82 (3H,
HR MS: calcd for C43H58O4Si2Na [MþNa]þ 717.3766, found 717.3735.
s, OCH3), 3.01 (2H, q, J ¼ 7.2 Hz, 2-H), 1.23 (3H, t, J ¼ 7.2 Hz, 3-H); 13
C
4.1.6. (3-(10,10-Bis(400-(tert-butyldimethylsilyloxy)phenyl)but-10-en-
20-yl)phenyl)methanol (10)
NMR (CDCl3, 500 MHz):
d
200.76 (1), 159.31 (400), 138.98, 137.08,
132.15, 130.03, 129.45, 128.65, 127.22, 127.19, 113.85 (300), 72.08
(PMPCH2 or PMBOCH2), 71.34 (PMPCH2 or PMBOCH2), 55.28
(CH3O), 31.85 (2), 8.23 (3); HR MS: calcd for C18H20O3Na [MþNa]þ
307.1305, found 307.1293.
To a solution of 9 (182 mg, 0.262 mmol) in dichloromethane/
phosphate buffer (pH 7; 9/1 ratio; 6.55 mL) was added 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone
(DDQ;
119 mg,
0.524 mmol) at 0 ꢁC. After the reaction mixture was stirred for
1.5 h at room temperature, phosphate buffer (pH 7) was added at
0 ꢁC. The mixture was extracted with dichloromethane, and the
organic layer was washed with brine and dried over sodium sulfate.
After filtration of the mixture and evaporation of the solvent, the
crude product was purified by thin-layer chromatography on silica
(eluant: hexane/ethyl acetate ¼ 3/1) to afford 10 (109 mg, 72%).
4.1.4. 4,40-(200-(3000-((40000-Methoxybenzyloxy)methyl)phenyl)but-100-
ene-100,100-diyl)diphenol (8)
To a suspension of zinc powder (2.76 g, 42.2 mmol) in THF
(21.1 mL) at ꢀ10 ꢁC was added titanium(Ⅳ) chloride (2.1 mL,
19.1 mmol). The reaction mixture was diluted with THF (5.0 mL)
and refluxed at 90 ꢁC (bath temperature) for 2 h, and then a mixture
of 4,40-dihydroxybenzophenone (7) (2.15 g, 10.0 mmol) and 6
(890 mg, 3.13 mmol) in THF (54.0 mL) was added to the mixture at
0 ꢁC. After the reaction mixture had been refluxed at 90 ꢁC (bath
temperature) for 2 h in the dark, 10% aqueous potassium carbonate
was added to the mixture at 0 ꢁC in the light. The mixture was
filtered through a short pad of Celite® with ethyl acetate, and the
filtrate was extracted with ethyl acetate. The organic layer was
washed with brine, and dried over sodium sulfate. After filtration of
the mixture and evaporation of the solvent, the crude product was
purified by thin-layer chromatography on silica (eluant: hexane/
ethyl acetate ¼ 6/1) to afford 8 (756 mg, 52%).
1H NMR (CDCl3, 400 MHz):
d 7.16e7.00 (6H, m, Ar), 6.82e6.80
(2H, m, Ar), 6.71e6.67 (2H, m, Ar), 6.48e6.45 (2H, m, Ar), 4.53 (2H,
d, J ¼ 5.6 Hz, CH2O), 2.49 (2H, q, J ¼ 7.3 Hz, 3-H), 0.99 (9H, s, TBS),
0.94e0.91 (12H, 4-H and TBS), 0.22 (6H, s, TBS), 0.09 (6H, s, TBS);
13C NMR (CDCl3, 400 MHz): 154.27 and 153.51 (400),142.99,140.64,
d
140.26, 138.41, 136.61, 136.42, 131.87, 130.49, 129.02, 128.49, 127.97,
124.53, 119.47 and 118.89 (300), 65.40 (CH2OH), 28.73 (30), 25.66 and
25.64 (Si(CH3)2C(CH3)3), 18.17 and 18.15 (Si(CH3)2C(CH3)3), 13.67
(40), ꢀ4.39 and ꢀ4.51 (Si(CH3)2C(CH3)3); HR MS: calcd for
C
35H50O3Si2Na [MþNa]þ 597.3191, found 597.3184.
4.1.7. 3-(10,10-Bis(400-(tert-butyldimethylsilyloxy)phenyl)but-10-en-
20-yl)benzaldehyde (11)
1H NMR (CD3OD, 500 MHz):
d 7.19e7.16 (3H, m, Ar), 7.09e7.06
(3H, m, Ar), 7.03e7.00 (2H, m, Ar), 6.88e6.86 (2H, m, Ar), 6.77e6.74
(2H, m, Ar), 6.68e6.65 (2H, m, Ar), 6.41e6.38 (2H, m, Ar), 4.36 (2H,
s, Bn), 4.22 (2H, s, Bn), 3.76 (3H, s, OCH3), 2.48 (2H, q, J ¼ 7.50 Hz, 3-
H), 0.90 (3H, t, J ¼ 7.50 Hz, 4-H); 13C NMR (CD3OD, 300 MHz):
To a solution of 10 (99.4 mg, 0.173 mmol) in dichloromethane
(3.46 mL) was added DesseMartin periodinane (DMP; 117 mg,
0.277 mmol) at room temperature. After the reaction mixture was
stirred for 2 h at room temperature, diethyl ether and 10% sodium
thiosulfate/saturated aqueous sodium hydrogen carbonate at a 1/1
ratio (volume) were added at 0 ꢁC. The mixture was extracted with
diethyl ether, and the organic layer was washed twice with brine,
and dried over sodium sulfate. After filtration of the mixture and
evaporation of the solvent, the crude product was purified by thin-
layer chromatography on silica (eluant: hexane/ethyl acetate ¼ 10/
1) to afford 11 (93.1 mg, 94%).
d
160.81 (40000), 157.24 (4 or 40), 156.39 (4 or 40), 144.51, 141.33,
140.22, 138.92, 136.47, 136.13, 133.12, 131.60, 131.38, 130.95, 130.70,
130.14, 129.03, 126.74, 115.82 (3 or 30 or 30000), 115.12 (3 or 30 or 30000),
114.75 (3 or 30 or 30000), 72.58 (PMPCH2 or PMBOCH2), 72.10 (PMPCH2
or PMBOCH2), 55.67 (CH3O), 29.86 (300), 13.99 (400); HR MS: calcd for
C
31H30O4Na [MþNa]þ 489.2036, found 489.2031.
4.1.5. (((20-(300-((4000-Methoxybenzyloxy)methyl)phenyl)but-10-ene-
10,10-diyl)bis(1,4-phenylene))bis(oxy))bis(tert-butyldimethylsilane)
(9)
1H NMR (CDCl3, 500 MHz):
d 9.87 (1H, s, CHO), 7.62e7.60 (2H, m,
Ar), 7.35e7.29 (2H, m, Ar), 7.11e7.08 (2H, m, Ar), 6.83e6.80 (2H, m,
Ar), 6.71e6.68 (2H, m, Ar), 6.49e6.46 (2H, m, Ar), 2.54 (2H, q,
J ¼ 7.5 Hz, 3-H), 1.00 (9H, s, TBS), 0.93 (3H, t, J ¼ 7.5 Hz, 4-H), 0.90
(9H, s, TBS), 0.23 (6H, s, TBS), 0.08 (6H, s, TBS); 13C NMR (CDCl3,
To a solution of 8 (38.1 mg, 0.0817 mmol) and imidazole
(44.5 mg, 0.654 mmol) in dimethylformamide (DMF; 2.04 mL) at
0 ꢁC were added tert-butylchlorodimethylsilane (49.2 mg,
0.326 mmol) and N,N-dimethyl-4-aminopyridine (DMAP; 0.20 mg,
0.00164 mmol). After the reaction mixture was stirred for 14 h at
room temperature, water was added at 0 ꢁC. The mixture was
extracted with diethyl ether, and the organic layer was washed
with 1 N sodium hydroxide solution, water, and brine, and dried
over sodium sulfate. After filtration of the mixture and evaporation
of the solvent, the crude product was purified by thin-layer chro-
matography on silica (eluant: hexane/ethyl acetate ¼ 9/1) to afford
9 (46.6 mg, 82%).
500 MHz):
d
192.44 (CHO), 154.53 and 153.86 (400), 143.90, 139.69,
139.44, 136.16, 136.13, 136.10, 135.80, 131.95, 131.15, 130.43, 128.56,
127.14, 119.55 and 119.16 (300), 28.59 (30), 25.66 and 25.64
(Si(CH3)2C(CH3)3), 18.18 (Si(CH3)2C(CH3)3), 13.57 (40), ꢀ4.38
and ꢀ4.50 (Si(CH3)2C(CH3)3); HR MS: calcd for C35H48O3Si2Na
[MþNa]þ 595.3034, found 595.3039.
4.1.8. Benzyl 3-(10,10-bis(400-(tert-butyldimethylsilyloxy)phenyl)but-
10-en-20-yl)benzoate (13)
To a solution of 11 (16.0 mg, 0.0279 mmol) in THF (0.47 mL) were
added 2-methyl-2-butene (0.030 mL, 0.279 mmol), tert-butyl
alcohol (0.47 mL), sodium chlorite (6.3 mg, 0.0559 mmol), and a
solution of sodium dihydrogen phosphate (16.6 mg, 0.140 mmol) in
water (0.47 mL). After the reaction mixture had been stirred for
3.5 h, brine was added at 0 ꢁC. The mixture was extracted with
chloroform, and the organic layer was dried over sodium sulfate.
After filtration of the mixture and evaporation of the solvent, the
crude product was obtained.
1H NMR (CDCl3, 500 MHz):
d 7.24e7.21 (2H, m, Ar), 7.15e7.12
(1H, m, Ar), 7.10e7.07 (4H, m, Ar), 7.03e7.01 (1H, m, Ar), 6.88e6.87
(2H, m, Ar), 6.82e6.79 (2H, m, Ar), 6.72e6.69 (2H, m, Ar), 6.47e6.44
(2H, m, Ar), 4.40 (2H, s, Bn), 4.31 (2H, s, Bn), 3.81 (3H, s, OCH3), 2.48
(2H, q, J ¼ 7.5 Hz, 3-H), 1.00 (9H, s, TBS), 0.92 (3H, t, J ¼ 7.5 Hz, 4-H),
0.89 (9H, s, TBS), 0.22 (6H, s, TBS), 0.06 (6H, s, TBS); 13C NMR (CDCl3,
500 MHz):
d
159.15 (4000), 154.26 (1), 153.48 (1), 142.80, 140.78,
138.23, 137.69, 136.71, 136.37, 131.89, 130.53, 130.42, 129.44, 129.23,
129.02, 127.84, 125.44, 119.46 (2), 118.85 (2), 113.75 (3000), 71.66
(PMPCH2 or PMBOCH2), 71.13 (PMPCH2 or PMBOCH2), 55.27 (CH3O),
28.91 (30), 25.68 and 25.63 (Si(CH3)2C(CH3)3), 18.18 and 18.12
(Si(CH3)2C(CH3)3), 13.65 (40), ꢀ4.38 and ꢀ4.52 (Si(CH3)2C(CH3)3);
To a solution of the crude product in DMF (1.00 mL) were added
55% sodium hydride (dispersion in paraffin liquid, 2.9 mg,
0.0670 mmol) and benzyl bromide (3.98
mL, 0.0335 mmol). After
the reaction mixture had been stirred for 11 h, saturated aqueous

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
647
ammonium chloride was added at 0 ꢁC. The mixture was extracted
with ethyl acetate, and the organic layer was dried over sodium
sulfate. After filtration of the mixture and evaporation of the sol-
vent, the crude product was purified by thin-layer chromatography
on silica (eluant: hexane/ethyl acetate ¼ 4/1) to afford 13 (15.2 mg,
80%).
azepanyl), 0.89 (3H, t, J ¼ 7.2 Hz, 4-H); 13C NMR (CDCl3, 500 MHz):
d
156.51 and 155.57 (400), 141.99, 141.35, 137.48, 136.83, 136.30,
133.18, 132.10, 130.64, 130.04, 128.99, 128.18, 127.54, 125.26, 114.05
and 113.38 (300), 63.91 and 63.61 (1000), 55.35 and 55.13 (2000), 54.70
and 54.46 (azepane), 29.66 (30), 27.06 and 27.02 (azepane), 24.14
and 23.62 (azepane), 13.58 (40); HR MS: calcd for C39H51N2O4
[MþH]þ 611.3843, found 611.3856.
1H NMR (CDCl3, 500 MHz):
d 7.87e7.87 (1H, m, Ar), 7.81e7.79
(1H, m, Ar), 7.44e7.33 (6H, m, Ar), 7.24e7.22 (1H, m, Ar), 7.18e7.15
(1H, m, Ar), 7.09e7.08 (2H, m, Ar), 6.82e6.80 (2H, m, Ar), 6.70e6.67
(2H, m, Ar), 6.48e6.46 (2H, m, Ar), 5.32 (2H, s, Bn), 2.52 (2H, q,
J ¼ 7.5 Hz, 3-H), 1.00 (9H, s, TBS), 0.93e0.90 (12H, m, 4-H and TBS),
0.22 (6H, s, TBS), 0.07 (6H, s, TBS); 13C NMR (CDCl3, 500 MHz):
4.1.11. (E)-3-(10,10-Bis(400-(tert-butyldimethylsilyloxy)phenyl)but-
10-en-20-yl)benzaldehyde O-benzyl oxime (16a) and (E)-3-((E)-10-
(400-(tert-Butyldimethylsilyloxy)phenyl)-10-(4000-hydroxyphenyl)but-
10-en-20-yl)benzaldehyde O-benzyl oxime or (E)-3-((Z)-10-(400-(tert-
Butyldimethylsilyloxy)phenyl)-10-(4000-hydroxyphenyl)but-10-en-20-
yl)benzaldehyde O-benzyl oxime (16b)
d
166.53 (COOBn), 154.40 and 153.68 (400), 143.14, 139.85, 139.22,
136.28, 136.10, 135.99, 134.85, 131.94, 130.69, 130.45, 129.65, 128.56,
128.14, 128.08, 127.81, 127.24, 119.48 and 119.08 (300), 66.54
(PhCH2O), 28.64 (30), 25.66 and 25.63 (Si(CH3)2C(CH3)3), 18.17 and
18.15 (Si(CH3)2C(CH3)3), 13.59 (40), ꢀ4.38 and ꢀ4.52
(Si(CH3)2C(CH3)3); HR MS: calcd for C42H54O4Si2Na [MþNa]þ
703.3482, found 703.3588.
To
a solution of 11 (43.3 mg, 0.0756 mmol) in methanol
(0.76 mL) at room temperature were added O-benzylhydroxyl-
amine (18.6 mg, 0.151 mmol) and sodium carbonate (32.0 mg,
0.302 mmol). The reaction mixture was stirred for 19 h at room
temperature, and then brine was added at 0 ꢁC. The mixture was
extracted with ethyl acetate, and the organic layer was dried over
sodium sulfate. After filtration of the mixture and evaporation of
the solvent, the crude product was purified by thin-layer chroma-
tography on silica (eluant: hexane/ethyl acetate ¼ 10/1) to afford
16a (17.6 mg, 34%) and 16b (22.2 mg, 52%).
4.1.9. Benzyl 3-(10,10-bis(400-hydroxyphenyl)but-10-en-20-yl)
benzoate (14)
To a solution of 13 (52.1 mg, 0.0767 mmol) in THF (2.50 mL) at
room temperature was added tetrabutylammonium fluoride (TBAF)
in THF (1.0 M, 0.30 mL, 0.300 mmol) at 0 ꢁC. The reaction mixture
was stirred for 1 h at room temperature, and saturated aqueous
ammonium chloride was added at 0 ꢁC. The mixture was extracted
with ethyl acetate, and the organic layer was dried over sodium
sulfate. After filtration of the mixture and evaporation of the sol-
vent, the crude product was purified by thin-layer chromatography
on silica (eluant: hexane/ethyl acetate ¼ 1/2) to afford 14 (21.0 mg,
61%).
4.1.12. 16a
1H NMR (CDCl3, 400 MHz):
d 8.01 (1H, s, 5-H), 7.44e7.31 (7H, m,
Ar), 7.14e7.04 (4H, m, Ar), 6.83e6.80 (2H, m, Ar), 6.72e6.69 (2H, m,
Ar), 6.50e6.47 (2H, m, Ar), 5.19 (2H, s, Bn), 2.51 (2H, q, J ¼ 7.6 Hz, 3-
H), 1.01 (9H, s, TBS), 0.95e0.92 (12H, m, TBS and 4-H), 0.23 (6H, s,
TBS), 0.10 (6H, s, TBS); 13C NMR (CD3OD, 400 MHz):
d 154.35 and
153.66 (400), 149.25 (CHNOBn), 143.25, 140.14, 138.78, 137.52, 136.45,
136.10, 131.86, 131.69, 131.64, 130.48, 128.54, 128.40, 128.10, 127.92,
124.39, 119.48 and 119.02 (300), 76.31 (PhCH2O), 28.72 (30), 25.67
(Si(CH3)2C(CH3)3), 18.17 (Si(CH3)2C(CH3)3), 13.61 (40), ꢀ4.38
and ꢀ4.48 (Si(CH3)2C(CH3)3); HR MS: calcd for C42H56NO3Si2
[MþH]þ 678.3793, found 678.3815.
1H NMR (CDCl3, 500 MHz):
d 7.88e7.87 (1H, m, Ar), 7.82e7.80
(1H, m, Ar), 7.43e7.33 (5H, m, Ar), 7.26e7.24 (1H, m, Ar), 7.21e7.18
(1H, m, Ar), 7.10e7.08 (2H, m, Ar), 6.81e6.79 (2H, m, Ar), 6.70e6.68
(2H, m, Ar), 6.46e6.43 (2H, m, Ar), 5.32 (2H, s, Bn), 2.51 (2H, q,
J ¼ 7.5 Hz, 3-H), 0.90 (3H, t, J ¼ 7.5 Hz, 4-H); 13C NMR (CD3OD,
500 MHz): d
167.85 (COOBn), 157.41 and 156.61 (400), 144.86, 141.26,
140.26, 137.60, 136.08, 135.83, 135.65, 133.17, 132.24, 131.56, 130.89,
129.62, 129.18, 128.98, 128.09, 115.83 and 115.27 (300), 67.61
(PhCH2O), 29.56 (30), 13.98 (40); HR MS: calcd for C30H26O4Na
[MþNa]þ 473.1723, found 473.1744.
4.1.13. 16b
1H NMR (CDCl3, 400 MHz):
d 8.02 (1H, s, 5-H E or Z), 8.00 (1H, s,
5-H E or Z), 7.43e7.28 (8H, m, Ar), 7.13e7.05 (4H, m, Ar), 6.82e6.79
(2H, m, Ar), 6.73e6.68 (2H, m, Ar), 6.48e6.45 (2H, m, Ar), 5.19 (2H,
s, Bn E or Z), 5.18 (2H, s, Bn E or Z), 4.70 (1H, s, OH E or Z), 4.47 (1H, s,
OH E or Z), 2.48 (2H, q, J ¼ 7.6 Hz, 3-H), 0.99e0.90 (12H, m, TBS and
4-H E or Z), 0.22 and 0.08 (6H, s, TBS E or Z); 13C NMR (CDCl3,
4.1.10. 3-(10,10-Bis(400-(2000-(azepan-10000-yl)ethoxy)phenyl)but-10-
en-20-yl)benzoic acid (RID-F-COOH, 15)
To a solution of 14 (21.0 mg, 0.0466 mmol) in DMF (1.20 mL) was
added 55% sodium hydride (dispersion in paraffin liquid, 12.2 mg,
0.280 mmol). The reaction mixture was stirred for 15 min at 50 ꢁC,
and then 1-(2-chloroethyl)azepane hydrochloride (30.5 mg,
0.154 mmol) was added in portions at room temperature. After the
reaction mixture had been stirred for 3 h at 50 ꢁC, saturated
aqueous ammonium chloride was added at 0 ꢁC. The mixture was
extracted with dichloromethane, and the organic layer was dried
over sodium sulfate. After filtration of the mixture and evaporation
of the solvent, the crude product was purified by thin-layer chro-
matography on silica (eluant: ammoniacal chloroform/meth-
anol ¼ 9/1) to afford the crude product (96.4 mg). After removing
the solid (57.6 mg) precipitated from the mixture, the residue was
azeotroped with toluene to afford RID-F-COOH (15) (22.1 mg, 77%).
500 MHz): d
154.34 and 154.25 and 153.63 and 153.46 (400 or 4000),
149.31 (CHNOBn), 143.18, 143.11, 140.23, 140.16, 138.48, 137.41,
136.43, 136.04, 135.59, 132.12, 131.84, 131.66, 130.71, 130.45, 128.42,
128.30, 128.19, 128.13, 127.94, 127.88, 124.47, 119.52 and 119.06 (300),
114.90 and 114.29 (3000), 76.29 (PhCH2O), 28.82 and 28.71 (30), 25.65
(Si(CH3)2C(CH3)3), 18.17 (Si(CH3)2C(CH3)3), 13.61 (40), ꢀ4.39
and ꢀ4.49 (Si(CH3)2C(CH3)3); HR MS: calcd for C36H41NO3SiNa
[MþNa]þ 586.2748, found 586.2762.
4.1.14. (E)-3-(10,10-Bis(400-hydroxyphenyl)but-10-en-20-yl)
benzaldehyde O-benzyl oxime (17)
To a solution of 16a (16.3 mg, 0.0240 mmol) in THF (0.60 mL) at
room temperature was added TBAF in THF (1.0 M, 72.1 mL,
0.0721 mmol) at 0 ꢁC. The reaction mixture was stirred for 30 min at
room temperature, and saturated aqueous ammonium chloride
was added at 0 ꢁC. The mixture was extracted with dichloro-
methane, and the organic layer was dried over sodium sulfate. After
filtration of the mixture and evaporation of the solvent, the crude
product was purified by thin-layer chromatography on silica
(eluant: hexane/ethyl acetate ¼ 2/1) to afford 17 (9.5 mg, 88%).
1H NMR (CDCl3, 500 MHz):
d 11.59 (1H, br s, COOH), 7.90 (1H, s,
Ar), 7.79 (1H, d, J ¼ 7.5 Hz, Ar), 7.17e7.10 (3H, m, Ar), 7.06e7.04 (1H,
m, Ar), 6.86e6.83 (2H, m, Ar), 6.76e6.74 (2H, m, Ar), 6.48e6.46 (2H,
m, Ar), 4.31 (2H, t, J ¼ 4.5 Hz, OCH2), 4.16 (2H, t, J ¼ 4.5 Hz, OCH2),
3.37e3.33 (4H, m, NCH2), 3.23e3.21 (8H, m, azepanyl), 2.48 (2H, q,
J ¼ 7.2 Hz, 3-H), 1.88e1.83 (8H, m, azepanyl), 1.69e1.65 (8H, m,

648
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
To a solution of 16b (22.2 mg, 0.0394 mmol) in THF (0.98 mL) at
room temperature was added TBAF in THF (1.0 M, 59.1 L,
and 4-H), 0.93 (3H, t, J ¼ 7.6 Hz, 4-H); 13C NMR (CDCl3, 500 MHz):
m
d
157.57 and 156.76 (400), 142.79, 140.84, 137.98, 136.17, 135.88,
0.0591 mmol) at 0 ꢁC. The reaction mixture was stirred for
30 min at room temperature, and saturated aqueous ammonium
chloride was added at 0 ꢁC. The mixture was extracted with
dichloromethane, and the organic layer was dried over sodium
sulfate. After filtration of the mixture and evaporation of the sol-
vent, the crude product was purified by thin-layer chromatography
on silica (eluant: hexane/ethyl acetate ¼ 2/1) to afford 17 (15.2 mg,
86%).
131.84,130.52,130.46,128.73,127.93,124.59,114.03 and 113.32 (300),
66.36 and 66.18 (1000), 56.45 and 56.31 (2000), 55.87 and 55.81 (aze-
pane), 46.48 (CH2NH2), 28.78 (30), 27.94 and 27.88 (azepane), 27.05
(azepane), 13.71 (40); HR MS: calcd for C39H54N3O2 [MþH]þ
596.4211, found 596.4207.
4.2. Synthesis of RID-F-peptide conjugates
1H NMR (CDCl3, 400 MHz):
d
8.03 (1H, s, 5-H), 7.43e7.30 (7H, m,
To obtain peptide conjugates, peptides were reacted with an
equimolar RID-F derivative carrying a carboxy group or amino
group, 1-(30-dimethylaminopropyl)-3-ethylcarbodiimide and 1-
hydroxybenzotriazole in DMF containing 100 equivalents of trie-
Ar), 7.15e7.05 (4H, m, Ar), 6.82e6.78 (2H, m, Ar), 6.73e6.70 (2H, m,
Ar), 6.47e6.44 (2H, m, Ar), 5.19 (2H, s, Bn), 2.48 (2H, q, J ¼ 7.6 Hz, 3-
H), 0.92 (3H, t, J ¼ 7.6 Hz, 4-H); 13C NMR (CDCl3, 400 MHz):
d 154.35
and 153.55 (400), 149.36 (CHNOBn), 143.15, 140.32, 138.33, 137.47,
136.09, 135.58, 132.10, 131.72, 131.68, 130.71, 128.43, 128.38, 128.22,
127.95, 124.61, 114.99 and 114.37 (300), 76.33 (PhCH2O), 28.83 (30),
13.58 (40); HR MS: calcd for C30H27NO3Na [MþNa]þ 472.1883, found
472.1890.
thylamine.
All
peptides
were
synthesized
using
9-
fluorenylmethoxycarbonyl (Fmoc) chemistry according to stan-
dard methods. After incubating for 6 h at 40 ꢁC, the target conju-
gates were purified from reaction mixtures with a HPLC system
(Gilson, Inc., WI, USA) equipped with a reverse phase column
(COSMOSIL, 5C18-AR-II, 4.6 ꢂ 150 mm, Nacalai Tesque Inc., Kyoto,
Japan). HPLC was performed at a flow rate of 1 mL/min with a linear
gradient of 5e50% acetonitrile in 0.05% trifluoroacetic acid in water.
Purities of the target conjugates were confirmed using a Spiral-TOF
JMS-S3000 mass spectrometer (JEOL Ltd., Tokyo, Japan).
4.1.15. (E)-3-(10,10-Bis(400-(2000-(azepan-10000-yl)ethoxy)phenyl)but-
10-en-20-yl)benzaldehyde O-benzyl oxime (18)
To a solution of 17 (8.5 mg, 0.0189 mmol) in DMF (0.4 mL) was
added 55% sodium hydride (dispersion in paraffin liquid, 8.3 mg,
0.189 mmol). The reaction mixture was stirred for 15 min at 50 ꢁC,
and then 1-(2-chloroethyl)azepane hydrochloride (18.7 mg,
0.0945 mmol) was added in portions at room temperature. After
the reaction mixture had been stirred for 12 h at 50 ꢁC, saturated
aqueous ammonium chloride was added at 0 ꢁC. The mixture was
extracted with dichloromethane, and the organic layer was dried
over sodium sulfate. After filtration of the mixture and evaporation
of the solvent, the crude product was purified by thin-layer chro-
matography on silica (eluant: ammoniacal chloroform/meth-
anol ¼ 30/1) to afford 18 (10.2 mg, 77%).
4.3. Cell culture
RPMI8226 and KMS-11 cell lines were distributed by RIKEN BRC
Cell Bank (Chiba, Japan) and JCRB Cell Bank, National Institutes of
Biomedical Innovation, Health and Nutrition (Osaka, Japan),
respectively. Cells were grown and maintained in RPMI1640 me-
dium (Nacalai Tesque, Kyoto, Japan) supplemented with 10% fetal
bovine serum, 100 units/mL penicillin, and 100
mg/mL streptomycin
at 37 ꢁC in 5% CO2.
1H NMR (CDCl3, 400 MHz):
d 7.42e7.24 (9H, m, Ar), 7.12e6.87
(2H, m, Ar), 6.73e6.70 (2H, m, Ar), 6.57e6.55 (2H, m, Ar), 4.54 (2H,
s, Bn), 4.08 (2H, t, J ¼ 6.4 Hz, OCH2), 3.94 (2H, t, J ¼ 6.4 Hz, OCH2),
2.97 (2H, t, J ¼ 6.4 Hz, NCH2), 2.88 (2H, t, J ¼ 6.4 Hz, NCH2),
2.80e2.69 (8H, m, azepanyl 2-H), 2.49 (2H, q, J ¼ 7.6 Hz, 3-H),
1.67e1.59 (16H, m, azepanyl 3-H and 4-H), 0.92 (3H, t, J ¼ 7.6 Hz, 4-
4.4. Proteasome fluorometric substrate assays
The CT-L, PGPH, and T-L proteasome activities were determined
by measuring the degradation rate of the fluorometric substrates,
Suc-LLVY-MCA, Z-LLE-MCA, and Boc-LRR-MCA, respectively. Puri-
H); 13C NMR (CDCl3, 400 MHz):
d
157.94 and 157.27 (400), 144.26
fied human 20S proteasomes (0.1
USA) were incubated with a fluorometric peptide substrate (50
in the presence of varying concentrations of inhibitory compounds
(0.1e10 M) in 100 L of assay buffer (25 mM HEPES [pH 8.0],
mg; Enzo Life Sciences, Inc., NY,
(CHNOBn), 140.12, 138.51, 138.48, 135.35, 134.69, 134.48, 133.11,
131.91, 131.74, 130.47, 130.38, 129.56, 128.67, 128.42, 128.33, 127.94,
127.65, 127.51, 119.06, 114.16 and 113.69 (300), 73.13 (PhCH2O), 66.44
and 66.21 (1000), 56.42 and 56.27 (2000), 55.88 and 55.83 (azepane),
29.68 (30), 27.95 and 27.88 (azepane), 27.03 (azepane),13.51 (40); HR
MS: calcd for C46H58N3O3 [MþH]þ 700.4473, found 700.4497.
mM)
m
m
0.5 mM ethylenediaminetetraacetic acid, 0.03% SDS) [40]. The re-
action mixture was monitored for the fluorescence of released 7-
amino-4-methylcoumarin (lex ¼ 380 nm, lem ¼ 480 nm) for 1 h at
37 ꢁC. Inhibitory potencies of the compounds (0.1e40
mM) for pu-
4.1.16. (3-(10,10-Bis(400-(2000-(azepan-10000-yl)ethoxy)phenyl)but-10-
en-20-yl)phenyl)methanamine (RID-F-CH2NH2, 19)
rified 26S proteasomes (0.1
ditions to those described above but with another assay buffer
(50 mM HEPES [pH 7.5], 40 mM KCl, 5 mM MgCl2, 50 g of bovine
serum albumin, 0.5 mM ATP, 1 mM dithiothreitol) with the corre-
sponding fluorometric peptide substrate (50 M). The IC50 value,
mg) were measured using similar con-
To a solution of 18 (9.7 mg, 0.0139 mmol) in THF (0.34 mL) at
0 ꢁC was added lithium aluminum hydride (1.6 mg, 0.0422 mmol).
The reaction mixture was stirred for 3 h at room temperature, and
saturated aqueous potassium sodium tartrate was added at 0 ꢁC.
The mixture was extracted with dichloromethane, and the organic
layer was dried over sodium sulfate. After filtration of the mixture
and evaporation of the solvent, the crude product was purified by
thin-layer chromatography on silica (eluant: ammoniacal chloro-
form/methanol ¼ 30/1) to afford RID-F-CH2NH2 (19) (6.4 mg, 77%).
m
m
defined as the compound concentration required for 50% inhibition
of proteasome activity, was determined for each compound from
the respective inhibition curve.
4.5. Cell growth inhibition assays
1H NMR (CDCl3, 400 MHz):
d
7.16e7.12 (3H, m, Ar), 7.02e7.00
RPMI8226 and KMS-11 cells were seeded in triplicate wells in a
(3H, m, Ar), 6.90e6.86 (2H, m, Ar), 6.77e6.73 (2H, m, Ar), 6.55e6.52
(2H, m, Ar), 4.09 (2H, t, J ¼ 6.0 Hz, OCH2), 3.93 (2H, t, J ¼ 6.0 Hz,
OCH2), 3.70 (2H, s, 5-H), 2.98 (2H, t, J ¼ 6.0 Hz, NCH2), 2.87 (2H, t,
J ¼ 6.0 Hz, NCH2), 2.80 (4H, q, azepanyl 2-H), 2.73 (4H, q, azepanyl
2-H), 2.49 (2H, q, J ¼ 7.6 Hz, 3-H), 1.68e1.56 (16H, m, azepanyl 3-H
96-well plate at a density of 5 ꢂ 102 cells/well and incubated with
various concentrations (0.1e10 mM) of inhibitory compounds dis-
solved in DMSO for 24 or 48 h. The amounts of intracellular ATP
were determined by CellTiter-Glo® Luminescent Cell Viability Assay
(Promega Co., WI, USA) according to the manufacturer's protocol.

M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
649
4.6. Measurements of intracellular proteasome activity
Hosoi for technical assistance. This study was supported by a Grant-
in-Aid for Scientific Research (C), JSPS KAKENHI (JP16K08329) to
M.H., and a Health and Labour Sciences Research Grant from the
Ministry of Health, Labour, and Welfare, Japan (11103425) to I.S.
M.S. received support from the Platform Project for Supporting
Drug Discovery and Life Science Research (Basis for Supporting
Innovative Drug Discovery and Life Science Research [BINDS]) from
the Japan Agency for Medical Research and Development (AMED).
RPMI8226 and KMS-11 [53] cells were seeded in triplicate wells
in a 96-well plate at a density of 5 ꢂ 103 cells/well and incubated for
24 or 48 h in the corresponding medium with various concentra-
tions (0.1e10 mM) of inhibitors dissolved in DMSO. The final con-
centration of DMSO was 0.5% in each reaction mixture. The
Proteasome-Glo™ Assay (Promega Co.) reagent [42] was mixed
according to the manufacturer's protocol. In this assay, cell lysis is
achieved directly at the step of addition of substrates, which con-
sists of peptides bound to aminoluciferin, with a detergent-
containing buffer. Aminoluciferin dye production, which is the
substrate of the buffer-contained luciferase, was measured after
incubation for 15 min.
Abbreviations
ATP, adenosine triphosphate; ATPg
S, adenosine 50-O-(3-thio)
triphosphate, CPP, cell-penetrating peptide; CT-L, chymotrypsin-
like; Cyt50, half-maximum cytotoxic concentration; DMF, dime-
thylformamide; DMSO, dimethyl sulfoxide, HPLC, high-
performance liquid chromatography; IC50, half-maximum inhibi-
tory concentration; PGPH, peptidylglutamyl peptide hydrolase; PI,
proteasome inhibitor; R8, octaarginine; RP, regulatory particle; RID,
ridaifen; SDS, sodium dodecyl sulfate; T-L, trypsin-like; THF,
tetrahydrofuran.
4.7. Molecular docking simulations
To dock RID-F-GR and RG-RID-F with the CT-L site of the human
20S proteasome, using AutoDock Vina version 1.1.2 (ADV) [43],
structural data were prepared as follows.
From the known human 20S proteasome structure data, the
complexed structure with dihydroeponemycin (PDB ID: 5lf1 [44])
determined at 2.0 Å resolution was selected. As the cleft formed by
Appendix A. Supplementary data
the b5 and b6 subunits (chain K and L, respectively) was used for
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.ejmech.2018.01.045.
binding to dihydroeponemycin, the cleft was considered as the CT-L
site, and the dimer structure of chains K and L was used for docking.
The dimer structure was rotated to fit the binding site of dihy-
droeponemycin into the docking grid box (24 ꢂ 24 ꢂ 20 Å) for the
ADV run (Fig. 4) and converted to PDBQT format using MGL Tools
version 1.5.6 [54]. The center coordinates of the docking grid box
were determined on the basis of the center of the amino acid res-
idues bound to dihydroeponemycin. To prevent a part of the ligand
from binding to the outside of the proteasome chamber, the center
coordinates of the docking grid box were slightly moved manually.
To handle the dipeptide moiety (Gly-Arg) of RID-F-GR in the
ADV calculation properly, 3D atomic coordinates of RID-F-GR were
generated using two steps. First, 3D atomic coordinates of RID-F-G
were generated from the two-dimensional representation using
Balloon [55] with default parameters. Second, 3D atomic co-
ordinates of Arg were added to the C-terminal of Gly in the RID-F-G
structure using the Protein Builder function in Molecular Operating
Environment software, version 2015.1001 (MOE 2015.1001).
Moreover, the Conformational Search function with the Low-
ModeMD of MOE 2015.1001 was performed to obtain an initial
structure of the docking. The first ranked structure obtained from
the Conformational Search was converted to PDBQT format using
MGL Tools. In the same way, 3D atomic coordinates of RG-RID-F for
the initial structure of the docking were generated using Balloon
and MOE 2015.1001.
References
[1] A. Hershko, A. Ciechanover, The ubiquitin system, Annu. Rev. Biochem. 67
(1998) 425e479. https://doi.org/10.1146/annurev.biochem.67.1.425.
[2] J.M. Peters, Z. Cejka, J.R. Harris, J.A. Kleinschmidt, W. Baumeister, Structural
features of the 26 S proteasome complex, J. Mol. Biol. 234 (1993) 932e937.
https://doi.org/10.1006/jmbi.1993.1646.
[3] D. Voges, P. Zwickl, W. Baumeister, The 26S proteasome: a molecular machine
designed for controlled proteolysis, Annu. Rev. Biochem. 68 (1999)
1015e1068. https://doi.org/10.1146/annurev.biochem.68.1.1015.
[4] D. Finley, Recognition and processing of ubiquitin-protein conjugates by the
proteasome, Annu. Rev. Biochem. 78 (2009) 477e513. https://doi.org/10.
1146/annurev.biochem.78.081507.101607.
[5] M. Orlowski, The multicatalytic proteinase complex, a major extralysosomal
proteolytic system, Biochemistry 29 (1990) 10289e10297.
[6] M. Groll, L. Ditzel, J. Lowe, D. Stock, M. Bochtler, H.D. Bartunik, R. Huber,
Structure of 20S proteasome from yeast at 2.4 A resolution, Nature 386 (1997)
463e471. https://doi.org/10.1038/386463a0.
[7] R. Verma, L. Aravind, R. Oania, W.H. McDonald, J.R. Yates 3rd, E.V. Koonin,
R.J. Deshaies, Role of Rpn11 metalloprotease in deubiquitination and degra-
dation by the 26S proteasome, Science 298 (2002) 611e615. https://doi.org/
10.1126/science.1075898.
[8] T. Yao, R.E. Cohen, A cryptic protease couples deubiquitination and degrada-
tion by the proteasome, Nature 419 (2002) 403e407. https://doi.org/10.1038/
nature01071.
[9] G.C. Lander, E. Estrin, M.E. Matyskiela, C. Bashore, E. Nogales, A. Martin,
Complete subunit architecture of the proteasome regulatory particle, Nature
482 (2012) 186e191. https://doi.org/10.1038/nature10774.
[10] K. Lasker, F. Forster, S. Bohn, T. Walzthoeni, E. Villa, P. Unverdorben, F. Beck,
R. Aebersold, A. Sali, W. Baumeister, Molecular architecture of the 26S pro-
teasome holocomplex determined by an integrative approach, Proc. Natl.
Acad. Sci. U. S. A. 109 (2012) 1380e1387. https://doi.org/10.1073/pnas.
1120559109.
Ten ADV runs were calculated with default parameters except
for exhaustiveness (1024) and num_modes (10). One hundred
docking poses obtained from 10 ADV runs were clustered by
complete linkage clustering using Root Mean Square Deviation
lower bound (l.b. RMSD) values between each docking pose and
grouped with cut-off values of 2.0 l.b. RMSD. The highest-score
docking pose from the largest cluster was selected as an esti-
mated binding mode.
[11] A.F. Kisselev, W.A. van der Linden, H.S. Overkleeft, Proteasome inhibitors: an
expanding army attacking a unique target, Chem. Biol 19 (2012) 99e115.
https://doi.org/10.1016/j.chembiol.2012.01.003.
[12] T. Hideshima, D. Chauhan, P. Richardson, C. Mitsiades, N. Mitsiades, T. Hayashi,
N. Munshi, L. Dang, A. Castro, V. Palombella, J. Adams, K.C. Anderson, NF-
kappa B as a therapeutic target in multiple myeloma, J. Biol. Chem. 277 (2002)
16639e16647. https://doi.org/10.1074/jbc.M200360200.
[13] A.L. Goldberg, Development of proteasome inhibitors as research tools and
cancer drugs, J. Cell Biol. 199 (2012) 583e588. https://doi.org/10.1083/jcb.
201210077.
Conflicts of interest
[14] A. Navon, A. Ciechanover, The 26 S proteasome: from basic mechanisms to
drug targeting, J. Biol. Chem. 284 (2009) 33713e33718. https://doi.org/10.
1074/jbc.R109.018481.
[15] J. Adams, P.J. Elliott, New agents in cancer clinical trials, Oncogene 19 (2000)
6687e6692. https://doi.org/10.1038/sj.onc.1204088.
The authors declare no competing financial interests.
Acknowledgments
[16] M.S. Raab, K. Podar, I. Breitkreutz, P.G. Richardson, K.C. Anderson, Multiple
myeloma, Lancet 374 (2009) 324e339. https://doi.org/10.1016/s0140-
The authors thank Shigehito Sugiyama, Mie Tsuchida, and Miho

650
M. Tanaka et al. / European Journal of Medicinal Chemistry 146 (2018) 636e650
6736(09)60221-x.
[36] I. Shiina, Y. Sano, K. Nakata, T. Kikuchi, A. Sasaki, M. Ikekita, Y. Nagahara,
Y. Hasome, T. Yamori, K. Yamazaki, Synthesis and pharmacological evaluation
of the novel pseudo-symmetrical tamoxifen derivatives as anti-tumor agents,
Biochem. Pharmacol. 75 (2008) 1014e1026. https://doi.org/10.1016/j.bcp.
2007.11.005.
[37] T. Mukaiyama, Titanium tetrachloride in organic synthesis [new synthetic
methods (21)], Angew. Chem. Int. Ed. Engl 16 (1977) 817e826. https://doi.
org/10.1002/anie.197708171.
[38] D.D. Yu, B.M. Forman, Simple and efficient production of (Z)-4-
hydroxytamoxifen, a potent estrogen receptor modulator, J. Org. Chem. 68
(2003) 9489e9491. https://doi.org/10.1021/jo035164n.
[39] P. Sledz, P. Unverdorben, F. Beck, G. Pfeifer, A. Schweitzer, F. Forster,
W. Baumeister, Structure of the 26S proteasome with ATP-gammaS bound
provides insights into the mechanism of nucleotide-dependent substrate
translocation, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 7264e7269. https://doi.
org/10.1073/pnas.1305782110.
[40] K. Tanaka, T. Yoshimura, A. Ichihara, Role of substrate in reversible activation
of proteasomes (multi-protease complexes) by sodium dodecyl sulfate,
J. Biochem 106 (1989) 495e500.
[41] A. Kloss, P. Henklein, D. Siele, M. Schmolke, S. Apcher, L. Kuehn,
P.W. Sheppard, B. Dahlmann, The cell-penetrating peptide octa-arginine is a
potent inhibitor of proteasome activities, Eur. J. Pharm. Biopharm. 72 (2009)
219e225. https://doi.org/10.1016/j.ejpb.2008.10.016.
[42] R.A. Moravec, M.A. O'Brien, W.J. Daily, M.A. Scurria, L. Bernad, T.L. Riss, Cell-
based bioluminescent assays for all three proteasome activities in a homo-
geneous format, Anal. Biochem. 387 (2009) 294e302. https://doi.org/10.1016/
j.ab.2009.01.016.
[17] D.J. Kuhn, Q. Chen, P.M. Voorhees, J.S. Strader, K.D. Shenk, C.M. Sun, S.D. Demo,
M.K. Bennett, F.W. van Leeuwen, A.A. Chanan-Khan, R.Z. Orlowski, Potent
activity of carfilzomib,
a novel, irreversible inhibitor of the ubiquitin-
proteasome pathway, against preclinical models of multiple myeloma,
Blood 110 (2007) 3281e3290. https://doi.org/10.1182/blood-2007-01-
065888.
[18] P. Moreau, T. Masszi, N. Grzasko, N.J. Bahlis, M. Hansson, L. Pour, I. Sandhu,
P. Ganly, B.W. Baker, S.R. Jackson, A.M. Stoppa, D.R. Simpson, P. Gimsing,
A. Palumbo, L. Garderet, M. Cavo, S. Kumar, C. Touzeau, F.K. Buadi, J.P. Laubach,
D.T. Berg, J. Lin, A. Di Bacco, A.M. Hui, H. van de Velde, P.G. Richardson PG1,
TOURMALINE-MM1 study group. Oral ixazomib, lenalidomide, and dexa-
methasone for multiple myeloma, N. Engl. J. Med. 374 (2016) 1621e1634.
https://doi.org/10.1056/NEJMoa1516282.
[19] L. Borissenko, M. Groll, 20S proteasome and its inhibitors: crystallographic
knowledge for drug development, Chem. Rev. 107 (2007) 687e717. https://
doi.org/10.1021/cr0502504.
[20] P. Beck, C. Dubiella, M. Groll, Covalent and non-covalent reversible protea-
some inhibition, Biol. Chem. 393 (2012) 1101e1120. https://doi.org/10.1515/
hsz-2012-0212.
[21] M. Hasegawa, Y. Yasuda, M. Tanaka, K. Nakata, E. Umeda, Y. Wang,
C. Watanabe, S. Uetake, T. Kunoh, M. Shionyu, R. Sasaki, I. Shiina, T. Mizukami,
A novel tamoxifen derivative, ridaifen-F, is a nonpeptidic small-molecule
proteasome inhibitor, Eur. J. Med. Chem. 71 (2014) 290e305. https://doi.
org/10.1016/j.ejmech.2013.11.009.
[22] A.D. Frankel, C.O. Pabo, Cellular uptake of the tat protein from human im-
munodeficiency virus, Cell 55 (1988) 1189e1193. https://dx.doi.org/10.1016/
0092-8674(88)90263-2.
[43] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of
[23] S. Reissmann, Cell penetration: scope and limitations by the application of
cell-penetrating peptides, J. Pept. Sci. 20 (2014) 760e784. https://doi.org/10.
1002/psc.2672.
docking with a new scoring function, efficient optimization, and multi-
threading, J. Comp. Chem 31 (2010) 455e461. https://doi.org/10.1002/jcc.
21334.
[24] S. Futaki, T. Suzuki, W. Ohashi, T. Yagami, S. Tanaka, K. Ueda, Y. Sugiura,
Arginine-rich peptides. An abundant source of membrane-permeable pep-
tides having potential as carriers for intracellular protein delivery, J. Biol.
Chem. 276 (2001) 5836e5840. https://doi.org/10.1074/jbc.M007540200.
[25] L.R. Wright, J.B. Rothbard, P.A. Wender, Guanidinium rich peptide transporters
and drug delivery, Curr. Protein Pept. Sci. 4 (2003) 105e124. https://doi.org/
10.2174/1389203033487252.
[26] G. Koczan, A.C. Ghose, A. Mookerjee, F. Hudecz, Methotrexate conjugate with
branched polypeptide influences Leishmania donovani infection in vitro and
in experimental animals, Bioconjugate Chem. 13 (2002) 518e524. https://doi.
org/10.1021/bc015530e.
[27] C. Rousselle, P. Clair, J.M. Lefauconnier, M. Kaczorek, J.M. Scherrmann,
J. Temsamani, New advances in the transport of doxorubicin through the
blood-brain barrier by a peptide vector-mediated strategy, Mol. Pharmacol. 57
(2000) 679e686. https://dx.doi.org/10.1124/mol.57.4.679.
[28] Y. Li, X. Zheng, Z. Cao, W. Xu, J. Zhang, M. Gong, Self-assembled peptide
(CADY-1) improved the clinical application of doxorubicin, Int. J. Pharm 434
(2012) 209e214. https://doi.org/10.1016/j.ijpharm.2012.06.003.
[29] P. Zhang, A.G. Cheetham, L.L. Lock, H. Cui, Cellular uptake and cytotoxicity of
drug-peptide conjugates regulated by conjugation site, Bioconjugate Chem.
24 (2013) 604e613. https://doi.org/10.1021/bc300585h.
[30] S. Biswas, P.P. Deshpande, F. Perche, N.S. Dodwadkar, S.D. Sane, V.P. Torchilin,
Octa-arginine-modified pegylated liposomal doxorubicin: an effective treat-
ment strategy for non-small cell lung cancer, Canc. Lett. 335 (2013) 191e200.
https://doi.org/10.1016/j.canlet.2013.02.020.
[31] C. Liu, X.N. Liu, G.L. Wang, Y. Hei, S. Meng, L.F. Yang, L. Yuan, Y. Xie, A dual-
mediated liposomal drug delivery system targeting the brain: rational con-
struction, integrity evaluation across the blood-brain barrier, and the trans-
porting mechanism to glioma cells, Int. J. Nanomedicine 12 (2017)
2407e2425. https://doi.org/10.2147/ijn.s131367.
[32] K. Splith, W. Hu, U. Schatzschneider, R. Gust, I. Ott, L.A. Onambele, A. Prokop,
I. Neundorf, Protease-activatable organometal-Peptide bioconjugates with
enhanced cytotoxicity on cancer cells, Bioconjugate Chem. 21 (2010)
1288e1296. https://doi.org/10.1021/bc100089z.
[44] J. Schrader, F. Henneberg, R.A. Mata, K. Tittmann, T.R. Schneider, H. Stark,
G. Bourenkov, A. Chari, The inhibition mechanism of human 20S proteasomes
enables next-generation inhibitor design, Science 353 (2016) 594e598.
https://doi.org/10.1126/science.aaf8993.
[45] P. Unverdorben, F. Beck, P. Sledz, A. Schweitzer, G. Pfeifer, J.M. Plitzko,
W. Baumeister, F. Forster, Deep classification of a large cryo-EM dataset de-
fines the conformational landscape of the 26S proteasome, Proc. Natl. Acad.
Sci. U. S. A. 111 (2014) 5544e5549. https://doi.org/10.1073/pnas.1403409111.
[46] A. Schweitzer, A. Aufderheide, T. Rudack, F. Beck, G. Pfeifer, J.M. Plitzko,
E. Sakata, K. Schulten, F. Forster, W. Baumeister, Structure of the human 26S
proteasome at a resolution of 3.9 A, Proc. Natl. Acad. Sci. U. S. A. 113 (2016)
7816e7821. https://doi.org/10.1073/pnas.1608050113.
[47] S. Chen, J. Wu, Y. Lu, Y.B. Ma, B.H. Lee, Z. Yu, Q. Ouyang, D.J. Finley,
M.W. Kirschner, Y. Mao, Structural basis for dynamic regulation of the human
26S proteasome, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) 12991e12996.
https://doi.org/10.1073/pnas.1614614113.
[48] J. Inoue, T. Otsuki, A. Hirasawa, I. Imoto, Y. Matsuo, S. Shimizu, M. Taniwaki,
J. Inazawa, Overexpression of PDZK1 within the 1q12-q22 amplicon is likely
to be associated with drug-resistance phenotype in multiple myeloma, Am. J.
Pathol. 165 (2004) 71e81. https://doi.org/10.1016/S0002-9440(10)63276-2.
[49] M. Kool, M. de Haas, G.L. Scheffer, R.J. Scheper, M.J. van Eijk, J.A. Juijn, F. Baas,
P. Borst, Analysis of expression of cMOAT (MRP2), MRP3, MRP4, and MRP5,
homologues of the multidrug resistance-associated protein gene (MRP1), in
human cancer cell lines, Canc. Res. 57 (1997) 3537e3547.
[50] O. Kocher, N. Comella, A. Gilchrist, R. Pal, K. Tognazzi, L.F. Brown, J.H. Knoll,
PDZK1, a novel PDZ domain-containing protein up-regulated in carcinomas
and mapped to chromosome 1q21, interacts with cMOAT (MRP2), the
multidrug resistance-associated protein, Lab. Invest. 79 (1999) 1161e1170.
[51] E. Ruoslahti, Tumor penetrating peptides for improved drug delivery, Adv.
Drug Deliv. Rev. 110e111 (2017) 3e12. https://doi.org/10.1016/j.addr.2016.
03.008.
[52] K. Josephson, A. Ricardo, J.W. Szostak, mRNA display: from basic principles to
macrocycle drug discovery, Drug Discov. Today 19 (2014) 388e399. https://
doi.org/10.1016/j.drudis.2013.10.011.
[33] J. Hoyer, U. Schatzschneider, M. Schulz-Siegmund, I. Neundorf, Dimerization
of a cell-penetrating peptide leads to enhanced cellular uptake and drug
delivery, Beilstein J. Org. Chem. 8 (2012) 1788e1797. https://doi.org/10.3762/
bjoc.8.204.
[34] W. Hu, K. Splith, I. Neundorf, K. Merz, U. Schatzschneider, Influence of the
metal center and linker on the intracellular distribution and biological activity
of organometal-peptide conjugates, J. Biol. Inorg. Chem. 17 (2012) 175e185.
https://doi.org/10.1007/s00775-011-0840-5.
[35] I. Shiina, Y. Sano, K. Nakata, M. Suzuki, T. Yokoyama, A. Sasaki, T. Orikasa,
T. Miyamoto, M. Ikekita, Y. Nagahara, Y. Hasome, An expeditious synthesis of
tamoxifen, a representative SERM (selective estrogen receptor modulator), via
the three-component coupling reaction among aromatic aldehyde, cinna-
myltrimethylsilane, and beta-chlorophenetole, Bioorg. Med. Chem. 15 (2007)
7599e7617. https://doi.org/10.1016/j.bmc.2007.09.008.
[53] M. Namba, T. Ohtsuki, M. Mori, A. Togawa, H. Wada, T. Sugihara, Y. Yawata,
T. Kimoto, Establishment of five human myeloma cell lines, In Vitro Cell. Dev.
Biol. 25 (1989) 723e729.
[54] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A.J. Olson, AutoDock4 and AutoDockTools4: automated docking with selective
receptor flexibility, J. Comp. Chem 30 (2009) 2785e2791. https://doi.org/10.
1002/jcc.21256.
[55] M.J. Vainio, M.S. Johnson, Generating conformer ensembles using a multi-
objective genetic algorithm, J. Chem. Inf. Model. 47 (2007) 2462e2474.
https://doi.org/10.1021/ci6005646.
[56] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt,
E.C. Meng, T.E. Ferrin, UCSF Chimeraea visualization system for exploratory
research and analysis, J. Comp. Chem 25 (2004) 1605e1612. https://doi.org/
10.1002/jcc.20084.