Identification of small molecule inhibitors of human COQ7

Article abstract of DOI:10.1016/j.bmc.2019.115182

Given that the associated clinical manifestations of ubiquinone (UQ, or coenzyme Q) deficiency diseases are highly heterogeneous and complicated, effective new research tools for UQ homeostasis studies are awaited. We set out to develop human COQ7 inhibit

Full text of DOI:10.1016/j.bmc.2019.115182

Journal Pre-proofs
Identification of small molecule inhibitors of human COQ7
Keiko Tsuganezawa, Katsuhiko Sekimata, Yukari Nakagawa, Rei Utata, Kana
Nakamura, Naoko Ogawa, Hiroo Koyama, Mikako Shirouzu, Takehiro
Fukami, Kiyoshi Kita, Akiko Tanaka
PII:
S0968-0896(19)31618-9
https://doi.org/10.1016/j.bmc.2019.115182
BMC 115182
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
4 October 2019
23 October 2019
24 October 2019
Please cite this article as: K. Tsuganezawa, K. Sekimata, Y. Nakagawa, R. Utata, K. Nakamura, N. Ogawa, H.
Koyama, M. Shirouzu, T. Fukami, K. Kita, A. Tanaka, Identification of small molecule inhibitors of human
COQ7, Bioorganic & Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.bmc.2019.115182
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Elsevier Ltd. All rights reserved.

Identification of small molecule inhibitors of human COQ7
Keiko Tsuganezawa1,2, Katsuhiko Sekimata3, Yukari Nakagawa1,2, Rei Utata1, Kana Nakamura1,2,
Naoko Ogawa1,2, Hiroo Koyama3, Mikako Shirouzu1,2, Takehiro Fukami4, Kiyoshi Kita5 and
Akiko Tanaka1,2
1RIKEN Systems and Structural Biology Center, 2RIKEN Center for Biosystems Dynamics Research, 1-7-
3
22 Suehiro-cho, Tsurumi, Yokohama 230-0045, RIKEN Center for Sustainable Resource Science,
4RIKEN Program for Drug Discovery and Medical Technology Platforms, 2-1 Hirosawa, Wako, Saitama
5
351-0198, School of Tropical Medicine and Global Health, Nagasaki University, 1-12-4, Sakamoto,
Nagasaki 852-8523, Japan.
Corresponding Author: Akiko Tanaka, RIKEN Center for Biosystems Dynamics Research, 1-7-22
Suehiro-cho, Tsurumi, Yokohama 230-0045, Japan. E-mail: aktanaka@riken.jp; Phone: +81-45-503-9451;
Fax: +81-45-503-9450.
Key words: COQ7; Ubiquinone; UQ; CoQ; inhibitor; UQ synthesis.
1

Abstract
Given that the associated clinical manifestations of ubiquinone (UQ, or coenzyme Q) deficiency
diseases are highly heterogeneous and complicated, effective new research tools for UQ
homeostasis studies are awaited. We set out to develop human COQ7 inhibitors that interfere with
UQ synthesis. Systematic structure-activity relationship development starting from a screening hit
compound led to the identification of highly potent COQ7 inhibitors that did not disturb
physiological cell growth of human normal culture cells. These new COQ7 inhibitors may serve
as useful tools for studying the balance between UQ supplementation pathways: de novo UQ
synthesis and extracellular UQ uptake.
2

1. Introduction
Ubiquinone (UQ; also known as coenzyme Q) is a redox-active lipid molecule comprised of
a benzoquinone and a polyisoprenoid side chain. UQ has important roles in energy metabolism
and reactive oxygen species (ROS) homeostasis in cells. UQ functions as an aerobic electron
transport mediator from complex I and II to complex III in eukaryotic mitochondria.1 In the
electron transport system in mitochondria, oxidation-reduction reaction mediated by UQ yields
intracellular ROS as a major by-product,2 and simultaneously, the reduced form of UQ regenerates
antioxidants, including vitamin E.1, 3
Clinically, UQ deficiencies are associated with highly heterogeneous and complicated
manifestations. Primary UQ deficiencies are very rare and genetically caused by mutations in
genes involved in UQ biosynthesis. These mutations affect multiple organ systems including the
central and peripheral nervous systems, kidney, skeletal muscle, heart, and sensory system.4
Secondary UQ deficiencies caused by defects not directly linked to UQ biosynthesis enzymes are
more common than primary UQ deficiencies and are associated with various disorders, including
mitochondrial myopathies.4, 5 The underlying biological mechanisms of attenuated UQ levels in
the patients remain largely unknown. A majority of the UQ-deficient patients respond well to oral
3

supplementation of high-dose UQ10, but in some cases, the clinical response to this therapy is less
than satisfactory.6
Mammalian cells have two UQ supplementation pathways: endogenous synthesis through the
UQ biosynthetic complex in mitochondria,4, 6, 7 and exogenous UQ uptake.8, 9 In the UQ synthetic
pathway, 4-hydroxybenzoic acid (4-HB), precursor of the benzoquinone, and the isoprenoid tail
are enzymatically linked by COQ2, and the resulting molecule undergoes subsequent
modifications of the ring moiety.4 This pathway is regulated in cells by controlling a key enzyme
COQ7.10-12 COQ7 activity is reported to be controlled at the transcription level by NFB,13 at the
mRNA stabilization level by HuR and hnRNP C1/C2,14 and at the phosphorylation level by
PPTC7.15 On the other hand, there is almost no report about the regulation of UQ uptake activity,
other than that most patients with UQ deficiency respond well to oral UQ10 supplementation. A
recent report indicates that the scavenger receptor CD36 is responsible for this function.8 The
precise mechanism which regulates these two pathways to achieve UQ homeostasis in human cells
is largely unknown, but elucidation of the mechanism is necessary to reveal the pathogenesis of
secondary UQ deficiencies. Given a number of questions to be resolved, new effective research
tools are needed for studying the biochemical aspects of UQ deficiency syndrome.
4

In this study, we have identified new inhibitors of COQ7, the key UQ synthetic enzyme.16, 17
COQ7 associates with COQ9 and forms large UQ biosynthetic complex with other enzymes
involved in UQ synthesis on the matrix face of the inner mitochondrial membrane.18 Human COQ7
is a DMQ mono-oxygenase consisting of 179 amino acids, and its active site is predicted to have
a typical di-iron center.19, 20 COQ7 converts demethoxy-UQ (DMQ) to demethyl-UQ (DMeQ),
which is then converted to UQ by COQ3.21 Thus, cells under COQ7 deficiency are unable to
produce UQ, and consequently accumulate DMQ.22, 23 To date, clioquinol (CQ), 5-chloro-7-indo-
8-hydroxyquinoline, is the only reported inhibitor of COQ7 hydroxylase activity.24 However, CQ
is known to form a stable complex with copper (II) ion and thereby exhibits serious cytotoxicity
in culture cells. Kawamura et al. studied the cytotoxicity of CQ and reported that inhibition of
superoxide dismutase-1 and resulted enhancement of ROS production led the CQ-treated cells to
death.25
Searches for other COQ7 inhibitors have met with little success, presumably because of the
lack of desirable enzymatic assays that can evaluate COQ7 hydroxylase activity. So far, stable
COQ7 forming complex with COQ9 has not yet been purified, thus reported recombinant COQ7
preparation has weak hydroxylase activity.26 To overcome this difficulty, we evaluated COQ7
inhibition activity of test compounds by their effects on UQ synthesis in co-cultured cells.
5

Compounds we identified in this study revealed that the balance between two UQ supplementation
pathways varies widely depending on the cell line.
2. Results and Discussion
2.1. Cell-based screening for hCOQ7 inhibitors
Since the UQ content in cells is known to vary widely among animal tissues,27 accumulation
levels of DMQ10 caused by inhibition of UQ synthesis may vary depending on the cell line.
Therefore, to select a sufficiently sensitive cell line for our cell-based UQ assay, we tested 5 human
cell lines: BxPC3, C3A, PANC-1, PC3, and HeLa. When cultured under normal conditions with
no inhibitor, the quinone fractions of these cells contained only UQ10 (section 4.3.2, Fig. 3A). After
the cells were co-cultured with 10 M CQ for 1 day, cells are expected to accumulate DMQ10 as
previously reported (section 4.3.2, Fig. 3B).24 The amounts of DMQ10 and UQ10 in each extract
were analyzed by using high-pressure liquid chromatography (HPLC), and the proportion of
DMQ10 content relative to the total quinone content (DMQ10 peak height / [UQ10 peak height +
DMQ10 peak height] (%)) was calculated for each extract. The DMQ10 content of extracts from
C3A and HeLa cells exceeded 20%, whereas extracts from BxPC3, PANC-1, and PC3 cells
contained less than 5% DMQ10. These observations suggested that only C3A and HeLa cells
6

actively synthesize UQ10 under this culture condition. When we extend the co-culture with 10 M
CQ to 2 days, the DMQ10 contents of the extracts were largely unchanged from those after 1 day
co-culture.
HeLa cells were selected for use in our cell-based screening assay. They were co-cultured for
2 days with each screening compound at 30 M, cell pellets were prepared from the co-cultures,
and their quinone fractions were extracted and analyzed. Among the 141 compounds tested,
compound 1 (3-Phenyl-1-(pyridine-2-yl)-1H-pyrazol-5-ol) caused remarkable levels of DMQ10
accumulation (>30% of total quinone).
Compound 1 was purchased from Labotest, and HeLa cells were co-cultured for 2 days with
the compound at 10 M. The DMQ10 content of the resulted extract was 19.3% + 5.2% (mean +
SD of triplicate results). Thus, we decided to initiate systematic structure-activity development
starting from compound 1 for more potent COQ7 inhibitors.
2.2. Development of improved compounds
2.2.1. Synthesis
Scheme 1 shows the synthesis of compounds 1–19. Pyrazole derivatives 1 and 3–19 were
synthesized by cyclization of the corresponding hydrazine derivatives and -ketoacetates in
EtOH.28 Chlorination of compound 1 to compound 2 was accomplished by heating in POCl3.
7

OEt
O
HO
N
Cl
N
a
b
R1-NHNH2
+
O
R1
R1
N
N
R2
R2
R2
1,3-19
2
Scheme 1. Reagents and conditions.
a) EtOH, 90 °C; b) POCl3, 110 °C.
2.2.2. Structure-activity relationships of the hCOQ7 inhibitors
COQ7 inhibition activity and cytotoxic effect of the synthetic compounds were determined
(Table 1). Inhibition of COQ7 activity was evaluated by the cell-based quinone analysis of extracts
from HeLa cells co-cultured with each test compound at 10 M for 2 days, and DMQ10 and UQ10
contents in each culture well (ng/well), and calculated DMQ10 content (%) were determined. The
effect of each compound on COQ7 activity was evaluated as DMQ10 content (%) as discussed
before (see section 2.1.), because total quinone content (ng) was affected by the number of cells
harvested. The cytotoxic effect of the compounds was evaluated by growth of human normal cell
line WI-38 co-cultured with various concentrations (0-30 M) of tested compounds for 4 days.
The concentration of each compound that caused 50% cell growth inhibition (GI50) was
determined and listed in Table 1.
Table 1. UQ10 synthesis-inhibiting activity of pyrazole derivatives.
8

X
N
R1
N
R2
DMQ10
content*
(%)
WI-38
GI50
(M)
Compound
No.
DMQ10*
(ng/well)
UQ10*
R1
R2
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
X
(ng/well)
1
2
3
4
5
6
7
8
9
OH
Cl
1.3
0
17.1
35.7
28.5
33.7
23.7
20.7
22.7
12.2
14.8
7.1
0
20.8
>30
>30
>30
>30
>30
>30
19.0
28.9
N
N
OH
OH
OH
OH
OH
OH
OH
0
0
N
0
0
0
0
N
Cl
0
0
N
Cl
0
0
N
Cl
13.2
1.8
52.0
10.8
N
F
N
9

MeO2S
H2N
Me
10
11
12
13
14
15
16
17
18
19
Ph
Ph
Ph
Ph
Ph
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
2.2
2.6
7.3
0
17.8
23.4
15.4
6.7
11.0
10.0
32.2
0
>30
>30
26.3
25.4
9.4
N
N
N
MeO
F3C
N
N
9.2
5.5
5.8
0
9.4
49.5
20.7
32.4
0
5.9
9.1
N
N
N
N
N
12.1
24.4
8.8
12.7
25.0
17.5
9.6
6.6
9.3
42.9
51.4
8.8
*: mean value of duplicate results.
No accumulation of DMQ10 was observed in cells co-cultured with compound 2. Therefore,
10

replacement of the hydroxyl group in compound 1 with chlorine (2) abolished the COQ7 inhibition
activity, suggesting that the hydroxyl group is essential for the interaction with COQ7. In addition,
compounds 3-5 failed to cause DMQ10 accumulation, suggesting that the 2-pyridyl structure is
critically necessary for the activity (Table 1).
Next, the contribution of the substituent on the 2-pyridyl group of compound 1 to DMQ10
accumulation was studied. Introduction of the 3-chloro and 4-chloro groups (6 and 7) resulted in
decreased activity. In contrast, the 5-chloro analog (8) was more potent than 1, suggesting that
introduction of an appropriate substituent at the 5-position may lead to increased COQ7 inhibition.
This finding prompted us to study the effects of other functional groups at the 5-position. In our
limited examples, the order of DMQ10 contents (%) in HeLa cells was as follows: chlorine (8) >
trifluoromethyl (14) > methyl (12) > sulfonyl methyl (10) > fluorine (9) > amine (11) > methoxy
(13).
When alkyl groups with various sizes (compounds 15-18) and an aralkyl group (compound
19) were introduced to the R2 group, it appeared that bulky groups (compounds 15, 16, 18, and
19) gave more desirable DMQ10 accumulation than the phenyl group (compound 1). However,
these compounds showed increased cytotoxicity.
Considering the balance between the desirable COQ7 inhibitory activity and the undesirable
11

cytotoxicity against the human normal cell line indicated by the GI50 values, compounds 8 and 12
appeared to have the most favorable characteristics (Table 1).
2.3. Characterization of the COQ7 inhibitors
2.3.1. Effects of the COQ7 inhibitors on human culture cells
Since compounds 8 and 12 had less growth inhibition against human normal culture cells
(Table 1), their effects on human cancer-derived cell lines (C3A, HeLa, PANC-1, and PC-3) were
investigated. The GI50 values of the two compounds against the four cell lines were determined,
and we found that all the values were higher than 12 M, except the value of compound 8 against
C3A cells (discussed later). When co-cultured with CQ, a known inhibitor of COQ7, C3A and
HeLa cells highly accumulated DMQ10, whereas PANC-1 and PC-3 cells did not (see section 2.1.).
These observations demonstrated the weak toxicity of compounds 8 and 12 against these cell lines,
regardless of the levels of accumulated DMQ10.
Since weak growth inhibition of C3A cells was observed with compounds 8 (GI50, 9.0 + 1.1
M) and 12 (GI50, 18.5 + 1.2 M)), biological damage induced by these COQ7 inhibitors were
studied using C3A cells. The first likely target of inhibitor-induced damage was electron mediation
in the respiratory chain, but we thought that the accumulated DMQ10 can function instead of UQ10
in the respiration and thus is nontoxic for human cells, as it is known for C. elegance.22 To confirm
12

this point, we determined the GI50 values of the compounds against C3A cells cultured in medium
containing 10 mM 2-deoxyglucose (2-DG), which inhibits glycolysis.29-32 Under these conditions,
cells utilize aerobic respiration only for their primary means of energy production. Thus
consequently C3A cells exhibited growth retardation (generation time extended approximately 2-
fold). However, the GI50 values of the COQ7 inhibitors stayed largely unchanged (GI50 values:
compound 8, 6.1 + 0.32 M; compound 12, 20.2 + 1.99 M), suggesting that the accumulated
DMQ10 does not inhibit cellular aerobic respiration in our assay system. It was in good agreement
with previous studies by Wang et al., in which mitochondrial preparations from coq7 KO mice
required only mild UQ addition for the electron transport.23, 33
The next likely target of the inhibitor-induced damage is the anti-oxidant defense systems.
Therefore, we added 0.3, 1, and 3 mM CoCl2 to C3A cultures to increase the generation of cellular
ROS.34, 35 Under these conditions, cell viabilities with 0.3, 1, and 3 mM CoCl2 at 24 hr were
decreased to 78%, 51%, and 33%, respectively. Simultaneous addition of compound 8 or 12 at 10
M was tested, and the addition did not increase the amount of cell death to meaningful extents.
Recent reports on transgenic mice study7 and pathogenesis of UQ deficiency4, 6 indicated that
severe UQ deficiency, per se, does not increase the levels of oxidative stress. Together, these
13

findings suggested that our COQ7 inhibitors do not limit C3A cell survival through inhibiting
aerobic respiration or increasing oxidative stress.
HeLa cells were co-cultured for 2 days with CQ, compounds 8 and 12 at various
concentrations, and their quinone fractions were analyzed (Fig.1). In the case of CQ treatment,
DMQ accumulation was observed with 5 M addition, but it decreased with 20 M addition due
to the severe cytotoxicity of CQ. On the other hand, extracts of cells treated with higher
concentrations of the new compounds showed increased DMQ10 contents (%). Saturation of
DMQ10 content occurred at approximately 60% of the total quinone content, and the addition of
compound 8 at 10 M was more effective to increase the DMQ10 content as compared with when
compound 12 was employed. These results showed usefulness of our new inhibitors to block UQ
synthesis without disturbing the condition of culture cells.
100%
80%
60%
40%
20%
0%
M)
5 10 20 5 10 20 2 10 20
CQ Compound8 Compound12
14

Fig. 1 Analysis of UQ10 and DMQ10 in HeLa cells treated with COQ7 inhibitors.
After 2 days of co-culture with CQ (left), compound 8 (middle), or 12 (right) at the indicated
concentrations, HeLa cells were harvested and their UQ10 and DMQ10 contents (%) were analyzed.
The gray part of each bar indicates the amount of UQ10, and the black part indicates that of DMQ10.
The mean values of duplicate results are shown.
2.3.2. UQ supplementation pathways in human culture cells
The quinone content of various cell lines cultured for 1 day with the inhibitors at 10 M was
analyzed (Fig. 2). To our surprise, substantial DMQ10 accumulation occurred in both PANC-1 cells
(11.4% by compound 8, and 8.7% by compound 12) and PC3 cells (27.6% by compound 8, and
14.0% by compound 12). In contrast, DMQ10 accumulations did not apparently occur in either
PANC-1 or PC3 cells treated with CQ (section 2.1). These observations indicated that treatment
with our non-toxic COQ7 inhibitors revealed that culture cells of all the human strains tested
actively synthesize UQ10 through COQ7.
25
20
15
10
5
0
PANC-1
PC3
C3A
HeLa
15

Fig. 2 COQ7 inhibitors reveal that the balance between UQ supplementation pathways varies
depending on the cell line.
PANC-1, PC3, C3A, and HeLa cells were co-cultured for 1 day with compound 8 or 12 at 10 M,
and the resulting quinone contents (ng) were determined. Each bar indicates the total amount of
UQ10 and DMQ10 in the untreated control culture (Day 0), control culture treated with DMSO (Day
1), and the cultures treated with compound 8 (Day 1 +8) or 12 (Day 1 +12). The gray part of each
bar represents the amount of UQ10, and the black part represents the amount of DMQ10. The mean
values of duplicate results are shown.
The increase of the total amount of UQ10 and DMQ10 during 1 day (ng per culture well) was
studied (Fig. 2). That in the untreated C3A culture was 7.4 ng, and in the HeLa culture was 6.8 ng.
On the contrary, the amount of accumulated DMQ10 in C3A culture treated with compound 8 for
1 day was 7.8 ng, and that in HeLa culture was 7.9 ng. These results suggested that UQ
supplementation in C3A and HeLa cells was mainly due to the de novo UQ synthesis through
COQ7. In contrast, the total amount of quinones increased by 5.5 ng in PANC-1 culture for 1 day,
but the DMQ10 accumulation was less than 2 ng. These findings suggested that uptake of
extracellular UQ10 primarily accounts for UQ supplementation in PANC-1 cells. Therefore, our
COQ7 inhibitors revealed that the balance between UQ supplementation pathways, de novo UQ10
synthesis, and uptake of extracellular UQ10, varies widely depending on the cell line.
Thelin et al. reported that the half-life of UQ9 in rat varied between 49 and 125 hours,
depending on the tissue evaluated.27 According to the UQ half-life, 10% to 30% of UQ10 present
16

in cells might be degraded during 1 day culture. It means that the actual amount of newly
synthesized UQ10 in the “Day 1” sample is somewhat larger than that calculated from the difference
between that in “Day 1” sample and in “Day 0” sample (Fig. 2). Therefore, our inhibitors might
not completely inhibit UQ10 synthesis in these cells, yet as shown in Fig. 2, our inhibitors are useful
tools for analyzing the balance between the UQ supplementation pathways. The information will
contribute to the biochemical understanding of the pathogenesis of secondary UQ deficiency.
3. Conclusions
We have identified a series of new COQ7 inhibitors that interfere with cellular UQ synthesis,
and we used these compounds to show that the balance between the UQ supplementation pathways
varies widely depending on the cell line studied. In particular, compounds 8 and 12 may serve as
effective tools for studying the UQ supplementation pathways, and contribute to the elucidation of
the mechanism that causes secondary UQ deficiency.
4. Materials and Methods
4.1. Screening compounds
A total of 141 compounds were obtained from the Drug Discovery Initiative, at the University
17

of Tokyo, Japan (http://www.ddi.u-tokyo.ac.jp/). CQ was purchased from Fluka.
4.2. Synthetic compounds
4.2.1. General
All reagents and solvents were obtained from commercial sources and used as received. 1H
NMR spectra were recorded with tetramethylsilane (TMS) as the internal standard by using a 270
MHz spectrometer (JNM-Ex, JEOL). Automated column chromatographic separations were
performed on a flash chromatograph (Biotage ZIP Si, Biotage Inc., and Hi Flash amino, Maruzen
Inc.). LC-MS analysis was performed by Aquity Ultra Performance LC (Waters) using an Aquity
UPLC BEH C18 column (2.1 mm x 50 mm, Waters). The column temperature was 40 C, run time
was 2 min, flow rate was 0.6 ml/min, and the eluting solvent was a mixture of acetonitrile and
water containing 0.1% trifluoroacetic acid with a gradient of 10%-90%. The mass spectrometry
data were acquired on a SQD2 quadrupole mass spectrometer (Waters). The purity of all
synthesized compounds exceeded 95%.
4.2.2. 3-Phenyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol (1)
Compound 1 was synthesized as reported before.36 1H NMR (270 MHz, CDCl3)  12.81(1H,
s), 8.27-8.30(1H, m), 8.04-8.08 (1H, m), 7.84-9.94 (3H, m), 7.33-7.46 (3H, m), 7.15-7.20(1H, m),
5.95 (1H, s). LC-MS (ESI): m/z 316 [M+H]+
18

4.2.3. 2-(5-Chloro-3-phenyl-1H-pyrazol-1-yl)pyridine (2)
Compound 1 (50 mg, 0.211 mmol) was heated in POCl3 (350 μl) for 3 h. After cooling to
room temperature, the reaction mixture was carefully poured into ice water, extracted with
dichloromethane, dried over Na2SO4, filtered, and concentrated in vacuo. The residue was purified
by column chromatography over silica gel (hexane/EtOAc) to afford 2 (33 mg, 61%) as a white
1
solid. H NMR (270 MHz, CDCl3)  8.58-8.60 (1H, m), 7.80-7.93 (4H, m), 7.30-7.46 (4H, m),
6.75 (1H, s) LC-MS (ESI): m/z 256 [M+H]+
4.2.4. 1,3-Diphenyl-1H-pyrazol-5-ol (3)
To a solution of phenylhydrazine (182 μM, 200 mg, 1.849 mmole), ethyl benzoylacetate
(382.9 l, 426.5 mg, 2.219 mmole) in EtOH (0.5 ml) was added and stirred at 90 °C overnight in
a sealed tube. After cooling to room temperature, the solid obtained was filtered, washed with
EtOH, and dried to afford 3 (112 mg, 26% yield) as a yellow solid. 1H NMR (270 MHz, CDCl3) 
7.95-8.00 (2H, m), 7.75-7.80 (2H, m), 7.40-7.49 (5H, m), 7.19-7.26 (1H, m), 3.87 (2H, s). LC-MS
(ESI): m/z 237 [M+H]+
4.2.5. 3-Phenyl-1-(pyridin-4-yl)-1H-pyrazol-5-ol (4)
Compounds 4-19 were prepared by using a similar procedure as that for compound 3. 1H
NMR (270 MHz, DMSO-d6)  8.59 (2H, d, J = 6.5 Hz), 7.98 (2H, d, J = 6.5 Hz), 7.84-7.87 (2H,
19

m), 7.34-7.47 (3H, m), 5.96 (1H, s). LC-MS (ESI): m/z 238 [M+H]+
4.2.6. 3-Phenyl-1-(pyridin-3-yl)-1H-pyrazol-5-ol (5)
1H NMR (270 MHz, DMSO-d6)  9.08 (1H, s), 8.48 (1H, d, J = 4.8 Hz), 8.22 (1H, d, J = 8.1
Hz), 7.84 (2H, d, J = 7.6 Hz), 7.35-7.55 (4H, m), 6.04 (1H, s). LC-MS (ESI): m/z 238 [M+H]+
4.2.7. 1-(3-Chloropyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (6)
1H NMR (270 MHz, CDCl3)  11.37 (1H, s), 7.95-7.87 (3H, m), 7.81-7.87 (3H, m), 7.34-7.46
(3H, m), 7.16-7.20 (1H, m) 5.98 (1H, s). LC-MS (ESI): m/z 272 [M+H]+
4.2.8. 1-(4-Chloropyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (7)
1H NMR (270 MHz, CDCl3)  8.38 (1H, d, J = 5.7Hz), 8.03 (1H, d, J = 1.6 Hz), 7.97-8.00
(1H, m), 7.78-7.82 (2H, m), 7.46-7.55 (3H, m), 3.91 (2H, s). LC-MS (ESI): m/z 272 [M+H]+.
4.2.9. 1-(5-Chloropyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (8)
1H NMR (270 MHz, CDCl3)  11.94 (1H, s), 8.26-8.27 (1H, m), 8.02-8.05 (1H, m), 7.79-7.89
(3H, m), 7.33-7.48 (3H, m), 5.96 (1H, s). LC-MS (ESI): m/z 272 [M+H]+.
4.2.10. 1-(5-Fluoropyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (9)
1H NMR (270 MHz, CDCl3)  11.93 (1H, s), 8.17 (1H, d, J = 2.7 Hz), 8.07-8.12 (1H, m),
7.81-7.87 (2H, m), 7.63-7.71 (1H, m), 7.33-7.48 (3H, m), 5.96 (1H, s). LC-MS (ESI): m/z 256
[M+H]+.
20

4.2.11. 1-(5-(Methylsulfonyl)pyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (10)
1H NMR (270 MHz, CDCl3)  11.90 (1H, s), 8.86 (1H, d, J = 2.2 Hz), 8.35-8.39 (1H, m),
8.23 (1H, d, J = 8.9 Hz), 7.85-7.89 (2H, m), 7.37-7.49 (3H, m), 6.01 (1H, s), 3.169 (3H, s). LC-
MS (ESI): m/z 316 [M+H]+.
4.2.12. 1-(5-Aminopyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (11)
1H NMR (270 MHz, CDCl3) δ 12.51 (1H, s), 7.83-7.89 (3H, m), 7.73 (1H, d, J = 2.4 Hz),
7.30-7.44 (1H, m), 7.22-7.25 (1H, t), 5.92 (1H, s), 3.72 (2H, s). LC-MS (ESI): m/z 253 [M+H]+.
4.2.13. 1-(5-Methylpyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (12)
1H NMR (270 MHz, CDCl3)  12.84 (1H, s), 8.10 (1H, d, J = 2.2 Hz), 7.95 (1H, d, J = 8.4
Hz), 7.84-7.88 (1H, m), 7.70-7.74 (1H, m), 7.32-7.47 (4H, m), 5.94 (1H, s), 2.38 (3H, s). LC-MS
(ESI): m/z 252 [M+H]+.
4.2.14. 1-(5-Methoxypyridin-2-yl)-3-phenyl-1H-pyrazol-5-ol (13)
1H NMR (270 MHz, CDCl3)  12.39 (1H, s), 7.99-8.02 (1H, m), 7.94-7.96 (1H, m), 7.83-7.87
(2H, m), 7.31-7.50 (4H, m), 5.94 (1H, s), 3.90 (3H, s). LC-MS (ESI): m/z 268 [M+H]+.
4.2.15. 3-Phenyl-1-(5-(trifluoromethyl)pyridin-2-yl)-1H-pyrazol-5-ol (14)
1H NMR (270 MHz CDCl3) δ 12.04 (1H, s), 8.59 (1H, s), 8.09-8.59 (2H, m), 7.84-7.88 (2H,
m), 7.36-7.50 (3H, m), 5.99 (1H, s). LC-MS (ESI): m/z 306 [M+H]+.
21

4.2.16. 3-Cyclohexyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol (15)
1H NMR (270 MHz, CDCl3) δ 12.70 (1H, s), 8.23 (1H, d, J = 5.1 Hz), 7.80-7.90 (2H, m),
7.08-7.13 (1H, m), 5.44 (1H, s), 2.54-2.61 (1H, m), 1.71-1.99 (1H, m), 1.27-1.51 (5H, m). LC-MS
(ESI): m/z 244 [M+H]+.
4.2.17. 3-Cyclopentyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol (16)
1H NMR (270 MHz, CDCl3) δ 12.74 (1H, s), 8.23-8.25 (1H, m), 7.80-7.91 (2H, m), 7.08-7.13
(1H, m), 5.44 (1H, s), 3.00-3.06 (1H, m), 2.00-2.06 (2H, m), 1.62-1.81 (6H, m). LC-MS (ESI): m/z
230 [M+H]+.
4.2.18. 3-Cyclobutyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol (17)
1H NMR (270 MHz, CDCl3) δ12.74 (1H, s), 8.22-8.25 (1H, m), 7.80-7.91 (2H, m), 7.09-7.17
(1H, m), 5.53 (1H, s), 3.46-3.53 (1H, m), 2.17-2.37 (4H, m), 1.89-2.08 (2H, m). LC-MS (ESI): m/z
216 [M+H]+.
4.2.19. 3-(Cyclohexylmethyl)-1-(pyridin-2-yl)-1H-pyrazol-5-ol (18)
1H NMR (270 MHz, CDCl3) δ12.74 (1H, s), 8.22-8.25 (1H, m), 7.80-7.91 (2H, m), 7.09-7.17
(1H, m), 5.53 (1H, s), 3.46-3.53 (1H, m), 2.17-2.37 (overlapped by water signal, 13H, m), 1.89-
2.08 (2H, m). LC-MS (ESI): m/z 258 [M+H]+.
4.2.20. 3-Phenethyl-1-(pyridin-2-yl)-1H-pyrazol-5-ol (19)
22

1H NMR (270 MHz, CDCl3) δ 12.70 (1H, s), 8.21-8.26 (1H, m), 7.81-7.91 (2H, m), 7.27-7.33
(4H, m), 7.07-7.23 (2H, m), 5.43 (1H, s), 2.86-3.04 (4H, m). LC-MS (ESI): m/z 266 [M+H]+.
4.3. Cell culture and cell-based assays
4.3.1. Cell culture
Human pancreas adenocarcinoma BxPC-3 cells, human hepatocellular carcinoma derived
C3A cells, human pancreas epithelioid carcinoma PANC-1 cells, and human prostate
adenocarcinoma PC-3 cells, were purchased from American Type Culture Collection. Human
cervix epidermoid carcinoma HeLa cells and human normal fibroblast WI-38 cells were obtained
from the RIKEN BioResource Center. HeLa, WI-38 and PANC-1 cells were cultured in Dulbecco's
Modified Eagle's Medium (GIBCO, Carlsbad, CA, USA), containing 10% fetal bovine serum
(FBS) and penicillin (5,000 U/ml)–streptomycin (5 mg/ml), at 37 °C in a 5% CO2 atmosphere.
BxPC-3 cells were cultured in RPMI-1640 Medium (GIBCO), C3A cells were in Eagle’s
Minimum Essential Medium (GIBCO), and PC-3 cells were in F-12K Medium (GIBCO),
containing 10% FBS and the antibiotics. The cells were pre-cultured overnight and then
subsequently cultured with each compound or 0.5% DMSO (control).
The effects of compounds on the growth of the human cells were evaluated by using the
CellTiter-Glo 2.0 assay (Promega). Various concentrations of the test compounds were added to
23

cells pre-cultured for 24 hr. The cells were cultured for further 3 days (PANC-1, PC-3, C3A, and
HeLa), or 4 days (WI-38), and the numbers of viable cells were determined. GI50 values, reported
as the mean values of triplicate results, were determined through linear interpolation from
concentration-response curves. To inhibit glycolysis 2-DG was added to the culture medium at
the concentration of 10 mM.29
4.3.2. Analysis of quinones
Cells (2 or 4 x 105 cells per well) were inoculated and cultured overnight, after which test
compounds were added and co-cultured for further 24 or 48 hr. Control samples were co-cultured
with 0.5%-1.0% DMSO. In most cases, we prepared 2 culture wells for each experimental
condition.
Harvested cells were suspended in phosphate-buffered saline (PBS) containing 5% K3Fe(CN)6.
After addition of 0.12 g UQ9 as the internal standard, quinone fractions were extracted from the
cell suspension by vigorous mixing in 7 volumes of ethanol/n-hexane (2:5, v/v) solution for 10
min, and then centrifuged at 9,100 x g for 5 min at room temperature. The supernatants were pooled
and dried in an evaporation device, and the residues were dissolved in 30 l methanol and analyzed
by HPLC with a UV detector (275 nm) (UV-2075, JASCO, Tokyo, Japan). A 10 l aliquot of the
quinone fraction was injected to a reverse-phase HPLC column (CAPCELL PAK C18IF S2, 2 m,
24

2.0 x 50 mm, OSAKA SODA, Osaka, Japan) and eluted under isocratic conditions (0.15 ml/min,
30 °C) by using diisopropyl ether/methanol (1:9, v/v), as described previously.22, 37 The spectral
characteristics of each quinone were monitored by using a photodiode array UV-visible detector
MD-2018 (JASCO).
4.3.3. Cell-based evaluation for hCOQ7 inhibitors
To select COQ7 inhibitors from test compounds, we measured the accumulated amounts of
the precursor molecule, DMQ10, in co-cultured cells. A specific peak detected by our HPLC device
was assigned as the DMQ10 peak in the following manner. An overnight culture of HeLa cells was
further co-cultured with 1% DMSO (negative control) or 10 M CQ (positive control) for 1 day,
and then cells were harvested, and their quinone fractions were extracted to measure the cellular
contents of DMQ10 and UQ10. The extract of negative control cells yielded two peaks, at 8.1 and
11.1 min (Fig. 3A), which we assigned to UQ9 and UQ10, respectively. UQ10 was the endogenous
UQ in human cells, and UQ9 was the internal standard that we added to the cell pellets. Extracts
of CQ-treated cells yielded a third peak, at 10.2 min (Fig. 3B), corresponding to DMQ10 as reported
previously.24 The treatment with CQ inhibited COQ7 that catalyzes DMQ10 to UQ10, and thus the
extract of treated cells contained DMQ10 fraction.
25

Fig. 3 Representative chromatograms of quinone fractions extracted from human cells.
Quinone fractions were extracted from human HeLa cells co-cultured for 1 day with 1% DMSO
(A) or 10 µM CQ (B). HeLa cells transfected with COQ7-siRNA were harvested and tested at 4
days (C) and 7 days (D) after the first transfection.
To confirm the assignment of the detected HPLC peak at 10.2 min to DMQ10 that accumulated
due to COQ7 inhibition, we analyzed the ubiquinone extract from cells with decreased COQ7
levels. HeLa cells were transfected with small interfering RNA (siRNA) to knock down COQ7
protein
as
reported
before.38
The
siRNA
targeting
human
COQ7
(5’-
AGUCGGGCAGCUGUGGAUCGAAUAA-3’) and control siRNA (Stealth RNAi siRNA
Negative Control Med GC Duolex #2) were purchased from Invitrogen. In short, 50 nM siRNA
and Lipofectamine RNAiMAX transfection reagent (Invitrogen) diluted in Opti-MEM media
(Invitrogen) were incubated for 20 min to prepare the transfection complexes, and HeLa cells were
transfected with the transfection complexes according to the manufacturer’s instructions.
After 3 days of culture, the amount of COQ7 in the transfected cells was examined by western
26

blotting using primary antibody (anti-hCOQ7 (F-9), Santa Cruz), and its elimination was
confirmed. The ubiquinone fractions extracted from the treated cells after 4 and 7 days showed
increase of the HPLC peaks at 10.2 min, which indicated that the peak at 10.2 min corresponds to
DMQ10 that accumulated due to the insufficiency of COQ7 (Figs. 3C and D). Therefore, we
confirmed that the cell-based measurement of the DMQ10 amount was available to evaluate COQ7
activity.
Acknowledgements
We thank Drs. Daniel Ken Inaoka (Nagasaki University) and Teruki Honma (RIKEN) for
their helpful discussion. This work was financially supported by the RIKEN Program for Drug
Discovery and Medical Technology Platforms. Generous support from the Drug Discovery
Initiative (The University of Tokyo) for providing chemical samples is gratefully appreciated.
Declarations of interest: None.
References
1. Zhao RZ, Jiang S, Zhang L, Yu ZB. Mitochondrial electron transport chain, ROS generation
27

and uncoupling (Review). Int. J. Mol. Med. 2019; 44: 3-15.
2. Zorov DB, Juhaszova M, Sollott SJ. Mitochondrial reactive oxygen species (ROS) and ROS-
induced ROS release. Physiol. Rev. 2014; 94: 909-950.
3. Bentinger M, Tekle M, Dallner G. Coenzyme Q--biosynthesis and functions. Biochem.
Biophys. Res. Commun. 2010; 396: 74-79.
4. Alcazar-Fabra M, Trevisson E, Brea-Calvo G. Clinical syndromes associated with
Coenzyme Q10 deficiency. Essays Biochem. 2018; 62: 377-398.
5. Quinzii CM, Hirano M. Primary and secondary CoQ(10) deficiencies in humans. Biofactors.
2011; 37: 361-365.
6. Acosta MJ, Vazquez Fonseca L, Desbats MA, Cerqua C, Zordan R, Trevisson E, Salviati L.
Coenzyme Q biosynthesis in health and disease. Biochim. Biophys. Acta. 2016; 1857: 1079-
1085.
7. Wang Y, Hekimi S. Understanding Ubiquinone. Trends Cell Biol. 2016; 26: 367-378.
8. Anderson CM, Kazantzis M, Wang J, Venkatraman S, Goncalves RL, Quinlan CL, Ng R,
Jastroch M, Benjamin DI, Nie B, Herber C, Van AA, Park MJ, Yun D, Chan K, Yu A, Vuong
P, Febbraio M, Nomura DK, Napoli JL, Brand MD, Stahl A. Dependence of brown adipose
tissue function on CD36-mediated coenzyme Q uptake. Cell reports. 2015; 10: 505-515.
28

9. Santos-Ocana C, Do TQ, Padilla S, Navas P, Clarke CF. Uptake of exogenous coenzyme Q
and transport to mitochondria is required for bc1 complex stability in yeast coq mutants. J.
Biol. Chem. 2002; 277: 10973-10981.
10. Marbois BN, Clarke CF. The COQ7 gene encodes a protein in saccharomyces cerevisiae
necessary for ubiquinone biosynthesis. J. Biol. Chem. 1996; 271: 2995-3004.
11. Jonassen T, Marbois BN, Kim L, Chin A, Xia YR, Lusis AJ, Clarke CF. Isolation and
sequencing of the rat Coq7 gene and the mapping of mouse Coq7 to chromosome 7. Arch.
Biochem. Biophys. 1996; 330: 285-289.
12. Jonassen T, Proft M, Randez-Gil F, Schultz JR, Marbois BN, Entian KD, Clarke CF. Yeast
Clk-1 homologue (Coq7/Cat5) is a mitochondrial protein in coenzyme Q synthesis. J. Biol.
Chem. 1998; 273: 3351-3357.
13. Brea-Calvo G, Siendones E, Sanchez-Alcazar JA, de Cabo R, Navas P. Cell survival from
chemotherapy depends on NF-kappaB transcriptional up-regulation of coenzyme Q
biosynthesis. PLoS One. 2009; 4: e5301.
14. Cascajo MV, Abdelmohsen K, Noh JH, Fernandez-Ayala DJ, Willers IM, Brea G, Lopez-
Lluch G, Valenzuela-Villatoro M, Cuezva JM, Gorospe M, Siendones E, Navas P. RNA-
binding proteins regulate cell respiration and coenzyme Q biosynthesis by post-transcriptional
29