Flavonoids from the peels of Citrus unshiu Markov. and their inhibitory effects on RANKL-induced osteoclastogenesis through the downregulation of c-Fos signaling in vitro

Article abstract of DOI:10.1016/j.bioorg.2020.104613

Phytochemical investigation of Citrus unshiu peels led to the isolation of eight new flavonols (7–9, 11–15) and sixteen known compounds (1–6, 10, 16–24). Their structures were elucidated using spectroscopic analysis (1D, 2D NMR, and HR-MS). Besides, all isolated compounds (1–24) were evaluated for their inhibitory effects on receptor activator of RANKL-induced osteoclastogenesis in BMMs. Among them, dimethylmikanin (1), quercetogetin (2), 3,3′,4′,5,7,8-hexamethoxyflavone (3), 3-methoxynobiletin (4) showed a significant inhibitory effect on RANKL-induced osteoclast differentiation at a concentration of 10 μM. Moreover, 3-methoxynobiletin (4) suppressed RANKL-induced osteoclastogenesis by decreasing the number of osteoclasts and osteoclast actin-ring formation in a dose-dependent manner without causing any cytotoxic effects on BMMs. At the molecular level, 3-methoxynobiletin (4) inhibited RANKL-induced c-Fos expression and subsequently NFATc1 activation, as well as the expression of osteoclastogenesis-related marker genes c-Src and CtsK. These findings suggested that 3-methoxynobiletin (4) attenuated osteoclast differentiation by inhibiting RANKL-mediated c-Fos signaling and that it may have therapeutic potential for treating or preventing bone resorption-related diseases, such as osteoporosis.

Full text of DOI:10.1016/j.bioorg.2020.104613

Bioorganic Chemistry 107 (2021) 104613
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Flavonoids from the peels of Citrus unshiu Markov. and their inhibitory
effects on RANKL-induced osteoclastogenesis through the downregulation
of c-Fos signaling in vitro
a
,
b
c
c
c
d
Thi Oanh Vu , Phuong Thao Tran , Wonyoung Seo , Jeong Hyung Lee , Byung Sun Min ,
a,b,*
Jeong Ah Kim
a
College of Pharmacy, Research Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 41566, Republic of Korea
Vessel-Organ Interaction Research Center, VOICE (MRC), College of Pharmacy, Kyungpook National University, Daegu 41566, Republic of Korea
Department of Biochemistry, College of Natural Sciences, Kangwon National University, Chuncheon, Gangwon-Do 24341, Republic of Korea
College of Pharmacy, Drug Research and Development Center, Daegu Catholic University, Gyeongbuk 38430, Republic of Korea
b
c
d
A R T I C L E I N F O
A B S T R A C T
Keywords:
Phytochemical investigation of Citrus unshiu peels led to the isolation of eight new flavonols (7–9, 11–15) and
sixteen known compounds (1–6, 10, 16–24). Their structures were elucidated using spectroscopic analysis (1D,
Citrus unshiu peels
Flavonols
2
D NMR, and HR-MS). Besides, all isolated compounds (1–24) were evaluated for their inhibitory effects on
Flavones
receptor activator of RANKL-induced osteoclastogenesis in BMMs. Among them, dimethylmikanin (1), querce-
RANKL
′
′
togetin (2), 3,3 ,4 ,5,7,8-hexamethoxyflavone (3), 3-methoxynobiletin (4) showed a significant inhibitory effect
on RANKL-induced osteoclast differentiation at a concentration of 10 M. Moreover, 3-methoxynobiletin (4)
Osteoclastogenesis
μ
suppressed RANKL-induced osteoclastogenesis by decreasing the number of osteoclasts and osteoclast actin-ring
formation in a dose-dependent manner without causing any cytotoxic effects on BMMs. At the molecular level, 3-
methoxynobiletin (4) inhibited RANKL-induced c-Fos expression and subsequently NFATc1 activation, as well as
the expression of osteoclastogenesis-related marker genes c-Src and CtsK. These findings suggested that 3-
methoxynobiletin (4) attenuated osteoclast differentiation by inhibiting RANKL-mediated c-Fos signaling and
that it may have therapeutic potential for treating or preventing bone resorption-related diseases, such as
osteoporosis.
1
. Introduction
lead to subsequent activation of the nuclear factor of activated T cells 1
(
NFATc1), the master regulator of osteoclastogenesis. The activation of
Osteoporosis is the most common metabolic bone disorder and re-
NFATc1 directly regulates the expression of several osteoclastic genes,
such as tartrate-resistant acid phosphatase (TRAP), CtsK, and c-Src
[5,6].
mains an increasingly significant problem that affects 200 million in-
dividuals worldwide [1]. Osteoporosis is characterized by low bone
mass and microarchitectural deterioration of bone tissue, leading to
enhanced bone fragility and increased fracture risk. An imbalance be-
tween osteoclast-mediated bone resorption and osteoblast-induced bone
formation is the main cause of the disease [2]. Osteoclasts are large
multinucleate cells that differentiate from hematopoietic precursor cells
of the monocyte-macrophage lineage following the stimulation of two
crucial cytokines, macrophage colony-stimulating factor (Mꢀ CSF) and
the receptor activator of nuclear factor (NF)-κB ligand (RANKL) [3,4].
The binding of RANKL and its receptor (RANK) triggers numerous
downstream signaling events, such as NF-κB and c-Fos signaling, which
Citrus unshiu Markov. (Rutaceae), known as Satsuma, Satsuma
mandarins, or Satsuma oranges in Western society, are widely cultivated
in subtropical countries, such as Korea, Japan, China, and Russia [7].
Their dried fruit peels have been used to improve a variety of digestive
dysfunctions, including tympanites, nausea, vomiting, and dyspepsia
[8]. The peels have also been traditionally used to improve bronchial
and asthmatic conditions as well as cardiac and blood circulation
problems in East Asia [9]. Previous chemical and pharmacological
studies reported that many actives components, including essential oils,
terpenoids, flavonoids, phenolic compounds were isolated from the
*
Corresponding author at: College of Pharmacy, Research Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 41566, Republic of Korea.
E-mail address: jkim6923@knu.ac.kr (J.A. Kim).
https://doi.org/10.1016/j.bioorg.2020.104613
Received 6 November 2020; Received in revised form 18 December 2020; Accepted 28 December 2020
Available online 5 January 2021
0
045-2068/© 2021 Elsevier Inc. All rights reserved.

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
peels of C. unshiu, and they were indicated to have the antioxidant, anti-
carcinogenicity, anti-allergic, anti-diabetic, anti-aging, anti-cancer, and
lipid-lowering activities [8]. Therefore, Citrus peels have a significant
potential to be used as a raw material for cosmetics, pharmaceuticals,
and functional foods.
Korea.
2.4. Extraction and isolation
Dried C. unshiu peels (10.0 kg) were thoroughly extracted three times
◦
In order to discover more naturally occurring compounds with novel
structures and notable bioactive properties for treating or preventing
bone resorption-related diseases, such as osteoporosis, phytochemical
investigations of the C. unshiu peels were undertaken. As a result, 24
compounds were isolated, including eight new flavonols (7–9, and
with MeOH under reflux at 65 C for 4 h each. MeOH extract was
concentrated in a vacuum to obtain a brown residue (3.3 kg), which was
suspended in H O and partitioned with solvents to produce five fractions
2
of n-hexane (70.5 g), methylene chloride (MC, 105.6 g), ethyl acetate
(EtOAc, 66.7 g), n-butanol (30.1 g), and a water-soluble layer (2.93 kg).
The MC extract (105.6 g) was subjected to a vacuum liquid chro-
matography (VLC) silica gel column and eluted with gradient mixtures
of n-hexane–EtOAc (1:0–1:1, v/v) and MC–MeOH (1:0–0:1, v/v) to yield
1
1–15). All isolates (1–24) were further evaluated for their inhibitory
effects on RANKL-induced osteoclast formation in bone marrow-derived
macrophages (BMMs).
1
2 fractions (1A–1L). Fraction 1I (12.8 g) was chromatographed on a C-
2
. Material and methods
18 gel CC using a gradient mixture of MeOH–H O (2:1–1:0, v/v) to
2
obtain eight fractions (2A–2I). Fraction 2C (1.85 g) was re-subjected to
2
.1. Chemicals, reagents, and antibodies
the MCI gel CC and eluted with an isocratic mixture of MeOH–H O (3:1,
2
v/v) to yield five fractions (3A–3E). Fractions 3B (412.4 mg), 3C (603.7
RANKL was purchased from RandD Systems (Minneapolis, Minne-
mg), and 3D (135.0 mg) were further purified by a Sephadex LH-20 CC
and eluted with MeOH to obtain compounds 19 (24.4 mg), 3 (7.0 mg),
21 (4.0 mg), 1 (2.0 mg) and 17 (3.0 mg), 16 (30.2 mg), 20 (670.8 mg),
and 4 (335.9 mg). Compound 18 (187.7 mg) was yielded after recrys-
tallization in MC–MeOH (1:1, v/v) from fraction 3E (314.0 mg). Fraction
2B (6.0 g) was applied to silica gel CC and eluted with a gradient mixture
of MC-acetone (9:1–0:1, v/v) to give six fractions (4A–4F). Fraction 4A
(703.6 mg) was separated by silica gel CC using an isocratic mixture of
MC–acetone (5:1, v/v) to obtain compound 2 (2.0 mg).
sota, USA), while Mꢀ CSF was purchased from Prospec (East Brunswick,
New Jersey, USA). Antibodies (anti-c-Fos, anti-Src, anti- -tubulin, and
α
anti-β-actin) were purchased from Cell Signaling Technology (Beverly,
USA). Anti-CtsK antibodies were purchased from Santa Cruz Biotech-
nology (Santa Cruz, California, USA). Fluorescein Isothiocyanate (FITC)-
conjugated phalloidin was obtained from Invitrogen (Carlsbad, Cali-
fornia, USA), as were Alexa 488-conjugated anti-rabbit antibodies, and
′
4
,6-diamidino-2-phenylindole (DAPI). All solvents used for extraction
and isolation were of analytical grade and purchased from Duksan Pure
Chemicals Co., Ltd. (Kyunggi, South Korea). High-performance liquid
chromatography (HPLC) solvents were purchased from Burdick &
Jackson (Michigan, USA).
The water layer (2.93 kg) was eluted over HP-20 Diaion CC with
MeOH-H O (0:1–1:0, v/v) to produce four fractions (5A–5D). Fraction
5C (308.6 g) was continuously subjected to VLC with a gradient mixture
2
of MC–MeOH–H O (5:1:0.1–7:3:0.5, v/v/v) to yield 12 fractions
2
(
6A–6L). Fraction 6C (25.76 g) was applied to silica gel CC and eluted
2
.2. General experimental procedures
with a gradient mixture of MC–MeOH–H O (5:1:0.1–7:3:0.5, v/v/v) to
2
obtain ten fractions (7A–7 J). Fraction 7G (2.1 g) was chromatographed
Ultraviolet (UV) spectra were measured with
a
Thermo
on MCI gel CC with a gradient mixture of MeOH-H O (1:1–3:1, v/v) to
2
9
423AQA2200E UV spectrometer (Thermo Fisher Scientific, Massa-
yield seven fractions (8A–8H). Compounds 7 (4.1 mg) and 8 (2.0 mg)
chusetts, USA). Infrared (IR) spectra were measured with a JASCO FT/
IR-4100 spectrometer (Jasco, Tokyo, Japan). High-resolution electro-
spray ionization mass spectrometry (HR-ESI-MS) spectra were recorded
using a Thermo Scientific LTQ Orbitrap XL hybrid four transform mass
spectrometer (Thermo Fisher Scientific, Massachusetts, USA). High-
resolution fast atom bombardment mass spectrometry (HR-FAB-MS)
were isolated from fraction 8H (70.5 mg) by using a Sephadex LH-20 CC
and eluting with 100% MeOH. Fraction 7I (2.1 g) was separated using
MCI gel CC and a gradient mixture of MeOH-H O (1:2–1:0, v/v) to
2
produce compound 5 (3.5 mg) and 12 fractions (9A–9L). Fraction 9 K
(688.5 mg) was subjected to Sephadex LP-20 CC and eluted with MeOH
to yield compounds 13 (80.2 mg), 14 (205.7 mg), and 15 (12.8 mg).
Fraction 7H (13.9 g) was subjected to MCI gel CC with a gradient
1
data were recorded using a JMS-700 model (Jeol, Tokyo, Japan). H-
(
500 MHz), 13C-(125 MHz) NMR, and 2D-NMR spectra were recorded on
mixture of MeOH–H O (1:1–1:0, v/v) to give compound 9 (66.5 mg) and
2
a Bruker Avance Digital 500 MHz NMR spectrometer (Bruker, Karlsruhe,
Germany) (in ppm relative to tetramethylsilane (TMS) as an internal
standard, J in Hz at 294 K). Thin-layer chromatography (TLC) was
performed using glass plates pre-coated with silica gel 60F254 and RP-18
12 fractions (10A–10 M). Fraction 10E (853.5 mg) was continuously
separated by C-18 gel CC and eluted with an isocratic mixture of
MeOH–H O (2:1, v/v) to obtain compounds 6 (5.0 mg), 12 (25.5 mg),
and 11 (6.3 mg). Fraction 6D (7.9 g) was isolated by MCI gel CC with a
2
F
254s (Merck, Darmstadt, Germany). Compounds were detected under
gradient mixture of MeOH–H O (1:1–1:0, v/v) to produce 15 fractions
2
UV light and then visualized by spraying the plates with a 10% sulfuric
(11A–11O). Fractions 11 K (75.0 mg) and 11L (75.5 mg) were further
◦
acid reagent followed by heating to 110 C for 1 min. Column chro-
purified by using preparative HPLC and eluting with an isocratic mixture
matography (CC) was performed on Merck silica gel (60–200
Merck Lichroprep RP-18 gel (40–63 m), MCI gel (75–150 m), and
Diaion HP-20 (250–850 m) (Merck, Darmstadt, Germany). High-
μ
m),
of MeOH–H O (40%, 6 ml/min, 60 min) to yield compounds 23 (3.0 mg,
2
μ
μ
t = 25.5 min) and 24 (3.0 mg, t = 42.0 min). Fractions 6G–6I (230.0 g)
R
R
μ
were re-chromatographed on VLC silica gels and eluted with a gradient
performance liquid chromatography (HPLC) was performed with an
HPLC Water system (Waters, Middleton, USA) with a 1525 Binary pump,
a Water 2998 photodiode array detector, a YMC Pak ODS column (20 ×
mixture of EtOAc–MeOH–H O (9:1:0.2–0:1:0, v/v/v) to obtain five
2
fractions (12A–12E). Fraction 12C (45.3 g) was isolated using silica gel
CC and was eluted with an isocratic mixture of EtOAc–MeOH–H O
2
2
50 mm, 4
.3. Plant material
Citrus unshiu Makov. peels were purchased from a traditional market
μ
m), t
R
was measured in min.
(8.5:1.5:0.5, v/v/v) to yield compounds 10 (10.0 mg) and 22 (6.2 mg).
Citrusunshin A (7). Yellow powder; UV (λmax, nm, log
ε
) 256 (4.20),
1
2
280 (4.35), 366 (4.17); IR (
ν
max): 3373, 1714, 1657, 1565, 1448; H and
1
3
+
C NMR: see Table 1; HR-FAB-MS m/z 553.1560 [M] (calcd 553.1557
29
for C25H O14).
in Jeonju-si, Korea, in February 2017 and identified by Professor Byung
Sun Min at Daegu Catholic University (Daegu, South Korea). A voucher
specimen (21A-CU) was deposited at the Laboratory of Pharmacognosy
at the College of Pharmacy, Kyungpook National University, South
Citrusunshin B (8). Yellow powder; UV (λmax, nm, log
ε
) 260 (4.18),
1
275 (4.25), 360 (4.16); IR (
ν
max): 3357, 1725, 1651, 1565, 1448; H and
1
3
+
C NMR: see Table 1; HR-FAB-MS m/z 711.2132 [M+H] (calcd
39
711.2136 for C32H O18).
2

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
Table 1
1
13
H (500 MHz) and C (125 MHz) NMR spectroscopic data for compounds 7–9, 11, and 12 (δ in ppm, J in Hz).
b)
b)
a)
b)
b)
Position
7
8
9
11
12
δ
H
(J in Hz)
δ
C
(mult, J in
δ
H
(J in Hz)
δ
C
(mult, J in
δ
H
(J in Hz)
δ
C
(mult, J in
δ
H
(J in Hz)
δ
C
(mult, J in
δ
H
(J in Hz)
C
δ (mult, J in
Hz) dC
Hz) dC
Hz) dC
Hz) dC
Hz) dC
2
3
4
5
6
7
8
9
159.1
135.2
180.0
150.0
137.4
154.2
134.3
146.4
108.3
122.7
114.2
148.6
151.7
116.3
124.1
159.4
135.0
179.8
149.9
137.5
154.3
134.3
146.3
108.2
122.3
114.1
148.8
152.4
116.5
124.4
153.8
135.5
172.3
143.6
137.5
151.0
146.3
147.4
114.3
122.6
112.3
148.1
151.2
111.4
121.8
158.0
135.1
179.0
160.0
102.0
163.7
94.6
158.2
134.4
179.0
161.1
97.0
6.35 d (2.0)
6.74 d (2.0)
6.53 s
159.0
125.1
149.2
105.9
124.8
113.9
148.6
151.1
116.1
122.9
148.5
107.5
122.7
114.4
148.5
151.6
116.2
124.2
1
1
2
3
4
5
6
0
′
′
′
′
′
′
8.01 d (2.1)
7.97 d (2.1)
7.83 d (2.0)
7.88 d (1.8)
7.93 d (2.0)
6.95 d (8.5)
7.74 dd
6.92 d (8.5)
7.75 dd
7.13 d (8.7)
7.69 dd (2.0,
8.7)
6.91 d (8.5)
7.63 dd (2.0,
8.5)
6.93 d (8.5)
7.90 dd (2.0,
8.7)
(
2.1, 8.5)
(2.1, 8.5)
5.44 d (7.5)
′
′
′
′
′
′
′
′
′
′
′
′
1
2
3
4
5
6
5.52 d (7.5)
103.3
76.0
78.0
71.6
78.6
62.6
103.4
75.6
75.8
71.7
77.9
64.5
5.42 d (7.3)
101.0
74.2
76.3
70.2
74.3
63.3
5.25 d (7.6)
102.8
d)
5.11 d (7.6)
103.5
d)
74.6
76.4
e)
e)
77.9
77.9
f)
f)
72.4
72.6
g)
g)
75.3
76.0
h)
h)
3.76 dd
2.2, 12.0)
.57 dd
5.7, 12.0)
4.28 dd
4.10 d (10.3)
4.22 ddd (1.5,
11.7, 23.3)
4.30 dd (2.4,
11.6)
65.2
4.49 dd (8.3,
10.1)
65.5
(
(2.1, 11.9)
c)
3
4.10 m
3.97 d (8.6)
(
′
′
′′
i)
i)
1
172.1
45.9
170.2
45.1
171.8
171.4
′′
c)
c)
c)
2
2.46 m
2.36 m
2.22 d (15.3)
47.0
2.24 m
46.5
2
.14 d (15.3)
′
′
′′
′′
3
4
70.6
46.2
68.8
45.6
70.4
47.4
70.4
48.2
c)
c)
c)
2.46 m
2.36 m
2.40 d (14.2)
.57 d (14.2)
2.24 m
2
2.54 d (13.8)
′
′
′
′
′
′
′
′
′′
′′
′′
′′
′′
′′
′′
′′
i)
i)
5
6
1
2
3
4
5
6
173.0
27.7
169.2
27.1
94.2
72.5
77.9
69.5
76.4
60.1
172.2
171.9
1.16 s
1.02 s
0.92 s
25.6
0.92 s
27.0
’
’
’
’
’
’
5.28 d (8.2)
5.25 d (7.6)
102.0
d)
5.85 d (7.5)
101.0
d)
75.9
76.3
e)
e)
77.8
78.2
f)
f)
72.9
72.1
g)
g)
75.0
75.4
h)
h)
4.45 m, 4.16
m
65.1
65.1
5
6
7
8
3
4
5
-OCH
-OCH
-OCH
3
3
3
3
3.84 s
3.94 s
4.02 s
3.81 s
3.84 s
3.84 s
61.9
61.7
61.6
62.1
55.6
55.7
3.92 s
3.99 s
57.2
57.1
3.92 s
4.11 s
3.97 s
3.98 s
61.5
62.2
62.6
56.7
3.92 s
4.12 s
3.98 s
3.97 s
61.5
62.3
62.6
56.7
-OCH
′
′
′
-OCH
-OCH
3
3
3.94 s
56.7
′′
-
3.60 s
51.9
OCH
3
a)
b)
c)
d-h)
In DMSO, In CD
3
OD, Overlapped,
Maybe interchangeable.
1
Citrusunshin C (9). Yellow powder; UV (λmax, nm, log
ε
) 258 (4.22),
272 (2.25), 358 (4.15); IR (
ν
max): 3383, 1723, 1651, 1562, 1430; H and
1
13
+
2
73 (4.23), 356 (4.16); IR (
ν
max): 3387, 1722, 1651, 1562, 1430; H and
C NMR: see Table 2; HR-FAB-MS m/z 845.2355 [M+H] (calcd
1
3
+
C NMR: see Table 1; HR-FAB-MS m/z 887.2823 [M+H] (calcd
45
845.2352 for C36H O23).
8
87.2821 for C39
H
51
O23).
Citrusunshin H (15). Yellow powder; UV (λmax, nm, log
ε
) 258 (4.23),
1
Citrusunshin D (11). Yellow powder; UV (λmax, nm, log
ε
) 255 (4.28),
275 (4.24), 358 (4.16); IR (
ν
max): 3387, 1722, 1652, 1562, 1430; H and
1
13
+
2
69 (4.35), 358 (4.19); IR (
ν
max): 3386, 1722, 1651, 1562, 1430; H and
C NMR: see Table 2; HR-FAB-MS m/z 859.2511 [M+H] (calcd
1
3
+
C NMR: see Table 1; HR-ESI-MS m/z 807.1990 [M+Na] (calcd
47
859.2508 for C37H O23).
8
40
07.1960 for C34H O21Na).
Citrusunshin E (12). Yellow powder; UV (λmax, nm, log
ε
) 258 (4.23),
1
2.5. Acid hydrolysis
2
73 (4.24), 356 (4.18); IR (
ν
max): 3377, 1722, 1652, 1562, 1430; H and
1
3
+
C NMR: see Table 1; HR-ESI-MS m/z 837.2078 [M+Na] (calcd
Each compound (1.5 mg) was refluxed in 5 ml of 2 N CF
3
COOH for 3
8
42
37.2065 for C35H O22Na).
◦
h at 95 C. After extraction with CH
2
Cl
2
(3 × 5 ml), the water layer was
Citrusunshin F (13). Yellow powder; UV (λmax, nm, log
ε
) 254 (4.21),
1
dried using an N2 stream, analyzed using TLC on a silica gel (CHCl :
3
2
80 (4.28), 358 (4.16); IR (
ν
max): 3382, 1722, 1651, 1562, 1430; H and
1
3
+
MeOH: H O, 8:5:1), and compared with an authentic sample. Mono-
2
C NMR: see Table 2; HR-FAB-MS m/z 815.2251 [M+H] (calcd
saccharide residue was obtained using preparative TLC with the same
8
43
15.2246 for C35H O22).
solvent. The optical rotation of sugar was measured and determined to
Citrusunshin G (14). Yellow powder; UV (λmax, nm, log
ε
) 258 (4.20),
α
D
be ᴅ-glucose (positive [ ] ) [10].
3

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
Table 2
1
13
H (500 MHz) and C (125 MHz) for compounds 13–15 (δ in ppm, J in Hz) in CD
3
OD.
Position
13a
13b
14a
14b
(J in Hz)
15a
(J in Hz)
15b
δ
H
(J in Hz)
δ
C
(mult,
δ
H
(J in Hz)
δ
C
(mult,
δ
H
(J in Hz)
δ
C
(mult,
δ
H
δ
C
δ
H
δ
C
(mult,
δ
H
(J in Hz)
C
δ (mult,
dC
J in Hz)
dC
J in Hz)
dC
J in Hz)
dC
J in Hz)
dC
J in Hz)
dC
dC
dC
dC
dC
2
3
4
5
6
7
8
9
158.5
135.3
179.4
150.3
100.3
158.9
129.2
158.0
105.6
123.0
114.2
148.4
151.0
116.2
124.0
158.5
135.3
179.4
150.3
100.3
158.9
129.2
158.0
105.6
123.0
114.2
148.4
151.0
116.2
124.0
158.3
135.0
179.7
146.5
133.0
152.5
129.4
149.5
105.1
123.1
114.2
148.4
151.0
116.2
124.1
158.6
135.0
179.7
146.5
133.0
152.5
129.4
149.5
105.1
123.1
114.2
148.4
151.0
116.2
124.1
159.1
135.2
179.8
149.8
137.4
154.3
134.3
146.3
108.2
122.9
114.2
148.4
151.2
116.2
124.2
159.1
135.2
179.8
149.8
137.4
154.3
134.3
146.3
108.2
122.9
114.2
148.4
151.2
116.2
124.2
6.18 s
6.18 s
1
1
2
3
4
5
6
0
ʹ
ʹ
7.97 d (2.0)
7.97 d (2.0)
7.97 d (2.0)
7.97 d (2.0)
7.96 d (2.0)
7.96 d (2.0)
ʹ
ʹ
ʹ
6.94 d (8.5)
7.74 dd
6.94 d (8.5)
7.74 dd
6.95 d (8.5)
7.74 dd
6.95 d (8.5)
7.74 dd
6.93 d (8.5)
7.72 dd
6.93 d (8.5)
7.72 dd
ʹ
(
2.0, 8.5)
(2.0, 8.5)
5.36 d (7.3)
(2.0, 8.5)
5.35 d (7.3)
(2.0, 8.5)
5.35 d (7.3)
(2.0, 8.5)
5.40 d (7.3)
(2.0, 8.5)
5.40 d (7.3)
′
′
′
′
′
′
′
′
′
′
′
′
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
3
6
7
8
5.36 d (7.3)
103.9
104.0
103.9
103.9
75.7
77.9
71.6
75.9
64.8
172.2
46.1
70.7
46.1
172.2
27.8
98.2
76.2
77.8
71.6
74.7
64.5
56.7
61.2
103.6
103.6
c)
d)
c)
d)
c)
d)
c)
d)
e)
f)
c)
d)
c)
d)
3.00–4.30
75.7
3.00–4.30
75.7
3.17–4.30
75.7
3.17–4.30
3.11–4.38
75.6
3.11–4.38
75.6
e)
e)
e)
e)
e)
77.9
77.9
77.9
77.8
77.9
f)
f)
f)
f)
f)
71.6
71.6
71.6
71.6
71.6
g)
g)
g)
g)
h)
g)
g)
75.9
75.8
75.8
75.8
75.9
h)
h)
h)
h)
h)
64.8
64.8
64.8
64.8
64.8
ʹʹʹ
172.2
46.0
70.7
46.1
172.5
27.8
94.0
172.2
46.1
70.7
46.1
172.2
46.1
70.7
46.1
172.2
46.1
70.7
46.1
172.4
46.0
70.7
46.1
172.4
27.8
98.2
c)
c)
c)
c)
c)
c)
ʹʹʹ
2.52 m
2.52 m
2.51 m
2.51 m
2.49 m
2.49 m
ʹʹʹ
c)
c)
c)
c)
c)
c)
ʹʹʹ
2.52 m
2.52 m
2.51 m
2.51 m
2.49 m
2.49 m
i)
i)
i)
ʹʹʹ
172.2
172.2
172.2
ʹʹʹ
1.20 s
1.20 s
27.8
98.2
1.20 s
27.8
94.0
1.20 s
1.17 s
27.8
94.0
1.17 s
ʹʹʹʹ
ʹʹʹʹ
ʹʹʹʹ
ʹʹʹʹ
ʹʹʹʹ
ʹʹʹʹ
ʹ-OCH
5.11 d (3.6)
4.52 d (7.8)
5.11 d (3.6)
4.52 d (7.8)
5.11 d (3.6)
4.51 s (7.8)
c)
d)
c)
d)
c)
d)
c)
d)
c)
d)
c)
d)
3.00–4.30
73.7
3.00–4.30
76.2
3.17–4.30
73.7
3.17–4.30
3.11–4.38
73.7
3.11–4.38
76.2
e)
e)
e)
e)
e)
e)
75.2
77.8
75.2
75.2
77.7
f)
f)
f)
f)
f)
f)
71.9
71.7
71.9
71.9
71.7
g)
g)
g)
g)
h)
g)
g)
70.6
74.7
70.6
70.6
74.7
h)
h)
h)
h)
h)
64.5
64.5
64.5
64.4
64.4
3
3.98 s
56.7
3.98 s
3.94 s
56.7
3.98 s
3.92 s
56.7
61.2
3.87 s
3.81 s
3.97, s
3.92, s
4.11, s
3.96, s
56.7
61.5
62.3
62.7
3.97, s
3.92, s
4.11, s
3.96, s
56.7
61.5
62.3
62.7
-OCH
-OCH
-OCH
3
3
3
3.94 s
d-h)
62.1
62.1
3.96 s
62.1
3.85 s
62.1
c)
Overlapped,
Maybe interchangeable.
2
.6. Isolation of BMMs and in vitro osteoclast differentiation assay
Billerica, MA, USA). TRAP-positive multinucleated cells with more than
five nuclei were defined as osteoclasts [11].
Bone marrow cells were isolated from the femurs and tibias of 6-
week-old, male mice from the Institute of Cancer Research (DBL,
2.7. Bone resorption and cytotoxicity assays
Emseong, Chungbuk, Korea). Each mouse weighed 22–25 g. All mice
◦
4
were housed in a temperature-controlled (24 ± 2 C) colony room with
BMMs (5 × 10 cells/well) were seeded in OsteoAssay Surface 96-
5
5 ± 10% humidity under a 12 h light and dark cycle. All mice were
well plates (Corning Incorporated, Corning, NY, USA) in -MEM sup-
α
given ad libitum access to a standard chow diet and water. Bone marrow
cells were cultured in -Minimal Essential Medium (MEM) (HyClone; GE
Healthcare Life Sciences, Logan, UT, USA) containing 10% FBS, 100 U/
ml penicillin, 100
plemented with 10% FBS, 1% penicillin and streptomycin, 30 ng/ml
Mꢀ CSF, 100 ng/ml RANKL, and various concentrations of the positive
control (Maslinic Acid (MA)), and compounds 1–24. The culture ran for
7 days, and the medium was replaced every 2 days. The differentiated
BMMs were washed with tap water, and images of the resorption pit
surfaces were captured using a model H550L microscope (Nikon Cor-
poration, Tokyo, Japan), and quantified via Image J software (Java
1.6.0_20 (64-bit); NIH, Bethesda, MD, USA).
α
μg/ml streptomycin, and 10 ng/ml Mꢀ CSF (Prospec-
Tany TechnoGene Ltd., East Brunswick, NJ, USA) overnight in a hu-
◦
midified incubator with 5% CO
2
at 37 C. The floating cells were har-
vested and maintained for 3 days with 30 ng/ml Mꢀ CSF. The cells that
adhered to the culture dish were characterized as BMMs and used for
4
subsequent experiments. The BMMs (5 × 10 cells/well) were cultured
Cytotoxicity was measured using a 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyl tetrazolium bromide (MTT)-based assay. The BMMs were
cultured in 96-well plates (1 × 104 cells/well) and supplemented with
30 ng/ml Mꢀ CSF in the presence of the indicated concentrations of
DRG. After 72 h, 0.5 mg/ml of MTT was added to each well for 3 h. At
the end of the incubation, the insoluble formazan products were dis-
solved in dimethyl sulfoxide, and absorbance at 540 nm was determined
[11].
in 96-well plates and maintained with RANKL (100 ng/ml) and Mꢀ CSF
(
30 ng/ml) for 7 days in the presence or absence of the indicated com-
pounds. The medium was replaced every 2 days. RAW264.7 cells (5 ×
4
1
0 cells/mL) were seeded onto 96-well plates and incubated with the
test compounds in the presence of RANKL (100 ng/mL). The compounds
and culture media were replaced every two days for a total of four days.
At the end of incubation, cells were fixed for 15 min in 3.7% formalin,
permeabilized with 0.1% Triton X-100, and then stained for TRAP with
an acid phosphatase leukocyte kit (Sigma-Aldrich; EMD Millipore,
4

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
2
.8. Western blot analysis
cells were incubated with an Alexa Fluor 488 goat anti-rabbit secondary
antibody (1:250 dilution) for 4 h at room temperature, then washed,
stained with DAPI, and mounted. Confocal images were acquired using
an OLYMPUS FV1000 inverted laser scanning confocal microscope
equipped with an external argon laser, HeNe laser Green, and HeNe laser
Red. Images were captured at the colony midsection using a UPLSAPO
60X NA1.35 oil immersion objective (OLYMPUS) [12].
Cells were lysed in a lysis buffer [150 mM NaCl, 50 mM Tris-HCl (pH
.4), 1 mM EDTA, 1% Nonidet P-40, 5 mM sodium orthovanadate, and
7
protease inhibitor cocktail (BD Biosciences, Franklin Lakes, NJ, USA)].
Equal amounts of proteins were separated by SDS-polyacrylamide gel
electrophoresis, followed by Western blot analysis. Western blots were
incubated overnight with the indicated primary antibodies (1:1,000
dilution), and anti-rabbit or anti-mouse secondary antibodies conju-
gated to horseradish peroxidase (1: 2,000 dilution) were used to visu-
alize signals using an enhanced chemiluminescence system
2.10. Statistical analysis
All experimental data are presented as the mean ± standard error of
the mean from at least three independent experiments. Statistical sig-
nificance was assessed by one-way analysis of variance and Tukey’s
posthoc corrections for multiple comparisons to examine differences
between the two groups using SPSS version 14.0 (SPSS Inc., Chicago, IL,
USA). P values < 0.05 were considered statistically significant.
(
ThermoFisher Scientific, Rockford, IL, USA). The band intensity was
quantified using Image J software (NIH, Bethesda, MD, USA) [12].
2
.9. Immunofluorescence
Cells were rinsed once in phosphate-buffered saline (PBS), fixed in
fresh 4% paraformaldehyde for 20 min at room temperature, and per-
meabilized in 0.5% TritonX-100. Nonspecific sites were blocked using a
PBS solution containing 1% goat serum. Cells were then incubated with
an anti-NFATc1 (1:200 dilution) antibody. After four washes with PBS,
Fig. 1. Chemical structures of isolated compounds from C. unshiu peels.
5

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
3
. Results and discussions
(C-3), 180.0 (C-4), 150.0 (C-5), 137.4 (C-6), 154.2 (C-7), 134.3 (C-8),
′
′
′
1
46.4 (C-9), 108.3 (C-10), 122.7 (C-1 ), 114.2 (C-2 ), 148.6 (C-3 ), 151.7
′
′
′
3
.1. Structural determination of compounds 1–24
(C-4 ), 116.3 (C-5 ), and 124.1 (C-6 ) and four methoxy groups at δ
C
61.5
′
(
6-OCH
3
), 62.2 (7-OCH
3
), 62.6 (8-OCH
3
), and 56.7 (3 -OCH
3
) were
Using repeated CC (silica gel, MCI, Sephadex LH-20, preparative
observed (Table 1). The location of methoxy groups at C-6, C-7, C-8, and
′
HPLC), 24 compounds, including eight new compounds (7–9, 11–15),
were isolated from the C. unshiu peels (Fig. 1). The chemical structures of
the known compounds were identified by comparison of their spectro-
scopic data with those reported previously in the literature, and
C-3 was confirmed by the heteronuclear multiple bond correlations
(HMBCs) between δ
134.3 (C-8), between δ
and δ
moiety was assumed from the signals between δ
NMR spectrum together with a set of characteristic signals at δ
H
3.92 and δ
C
137.4 (C-6), between δ
H
3.97 and δ
C
′
H
3.98 and δ
C
148.6 (C-3 ), and between δ
H
4.11
C
154.2 (C-7), respectively (Fig. 2). In addition, a glucopyranosyl
′
′
1
included: dimethylmikanin (1) [13], quercetogetin (3,3 ,4 ,5,6,7-hex-
H
3.28 and 5.52 in the H
103.3
′
′
amethoxyflavone) (2) [14], 3,3 ,4 ,5,7,8-hexamethoxyflavone (3) [15],
C
′
′
′′
′′
′′
′′
′′
13
3
-methoxynobiletin (3,5,6,7,8,3 ,4 -heptamethoxyflavone) (4) [15],
(C-1 ), 78.6 (C-5 ), 78.0 (C-3 ), 76.0 (C-2 ), and 62.6 (C-6 ) in the
C
rutin (5) [16], isorhamnetin 7-O-rutinoside (6) [17], limocitrunshin
NMR spectrum (Table 1). Furthermore, β-form of glucopyranosyl moiety
was deduced by the large coupling constant J = 7.5 Hz of anomeric
proton [24]. After acid hydrolysis, the sugar unit was determined to be
ᴅ-glucose, which was identified by comparison of TLC data and optical
21.0
′
(
10) [18], 6-demethoxytangeretin (16) [19], 5,6,7,4 -tetramethoxy-
flavone (17) [20], tangeretin (18) [19], 6-demethoxynobiletin
′ ′
(
3 ,4 ,5,7,8-pentamethoxyflavone) (19) [19], nobiletin (20) [19],
′ ′ ′ ′ ′′
,7,8,2 ,3 ,4 ,5 -heptamethoxyflavone (21) [21], vitexin 2 -O-xyloside
5
rotation [
α
]
+ 50.8 (c 0.2, H
2
O) with those of authentic samples
(
22) [22], diosmetin 8-C-glucoside (23) [23], and diosmetin 6-C-
D
glucoside (24) [23]. The structure elucidation of the new compounds
[10,25,26]. The location of a glucopyranosyl moiety at C-3 of aglycone
′′
(
7–9, 11–15) was established according to their derivatives (1–6 and
was confirmed by the HMBC between H-1 and C-3 (Fig. 2). Thus,
′
′
1
0) together with the 1D, 2D NMR and HR-MS spectra.
compound 7 was identified as 5,4 -dihydroxy-6,7,8,3 -tetramethoxy-
flavone 3-O-β-ᴅ-glucoside and named citrusunshin A.
Compound 7 was isolated as a yellow powder, and its molecular
formula was determined as C25
H
28
O
14 by the HR-FAB-MS ion at m/z
Compound 8 was a yellow powder and was assigned the molecular
+
+
5
53.1560 [M+H] (calcd for C25
H
29
O14, 553.1557) (Fig. S1.9, Sup-
formula C32
(calcd for C32
Its IR spectrum showed an absorption band for hydroxyl (3357 cm ),
H
38
O
18 from its HR-FAB-MS ion at m/z 711.2132 [M+H]
porting Information). UV absorption bands at 256, 280, and 366 nm
H
39
O
18, 711.2136) (Fig. S1.18, Supporting Information).
ꢀ 1
suggested that compound 7 is a flavonol derivative. Its IR spectrum
ꢀ 1
ꢀ 1
ꢀ 1
ꢀ 1
indicated the presence of hydroxyl (3373 cm ), carbonyl (1714 cm ),
carbonyl (1725 cm ), conjugated carbonyl (1651 cm ), and aromatic
ꢀ
1
ꢀ 1
(1565, 1448 cm ) groups. The 1H and 13C NMR data of compound 8
ꢀ 1
conjugated carbonyl (1657 cm ), and aromatic (1565, 1448 cm
)
1
groups. The H NMR spectrum of 7 displayed signals of an ABM coupling
system at δ 8.01 (1H, d, J = 2.1 Hz), 7.74 (1H, dt, J = 2.1, 8.5 Hz), and
were similar to those of compound 7, except for the additional signals of
′
′′
′′′
two carbonyl groups at δ
quaternary carbon at δ
C
172.1 (C-1 ), 173.0 (C-5 ); an oxygenated
H
1
′′′
6
.95 (d, 1H, J = 8.5 Hz). Additionally, the H NMR spectrum also
C
70.6 (C-6 ); two methylene groups at δ
H
2.46
′
′′
′′′
exhibited four methoxy protons at δ
3
H
3.92 (6-OCH
3
), 3.97 (8-OCH
3
),
(4H, m) and δ
and δ
C
45.9 (C-2 ), 46.2 (C-4 ); a tert-methyl at δ
H
1.16 (3H, s)
′
13
′′′
.98 (3 -OCH
3
), and 4.11 (7-OCH
3
) (Table 1). In the C NMR spectrum,
C
27.7 (C-6 ) which were determined to be a 3-hydroxy-3-methyl-
the characteristic signals of a flavone skeleton at δ
C
159.1 (C-2), 135.2
glutaroyl (HMG) moiety (Tables 1). Further confirmation was deduced
1
1
Fig. 2. Key H– H COSY and HMBC correlations of 7–9 and 11–15.
6

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
′
′′
′′′
′′′
′′′
′′′
′′′
′′′
by HMBCs from H-2 to C-1 , C-3 , C-4 , and C-6 , from H-4 to C-
data analyses, compound 11 was determined to be isorhamnetin-7-O-
′
′′
′′′
′′′
′′′
′′′
′′′
′′′
2
, C-3 , C-5 , and C-6 , from H-6 to C-2 , C-3 , and C-4 (Fig. 2).
In addition, an upfield methoxy group at δ 3.60 (3H, s) and δ 51.9
) was also observed in the H and C NMR spectra of 8. The
[β-ᴅ-glucopyranosyl-(6
6)]-β-ᴅ-glucopyranoside and named citrusunshin D.
→
5)-(3S)-3-hydroxy-3-methylglutarate(1
→
H
13
C
′
′′
1
(
5 –OCH
3
Compound 12 was isolated as a yellow powder. Its HR-ESI-MS
′
′′
+
location of this methoxy at C-5 was confirmed by an HMBC from
spectrum exhibited an ion at m/z 837.2078 [M+Na] (calcd 837.2065
′
′′
′′′
′′
5
–OCH
3
to C-5 . The downfield chemical shift of C-6 (δ
C
64.5) in 8
62.6) suggests that the HMG
moiety is at position C-6 . This was also confirmed by an HMBC from H-
42
for C35H O22Na), which consistent with a molecular formula of
′
′
1
13
compared to the C-6 of glucose in 7 (δ
C
35 42
C H O22 (Fig. S1.45, Supporting Information). The H, C NMR, and
′
′
distortionless enhancement by polarization transfer spectra (DEPT) of
′′
′′′
′′′
6
to C-1 (Fig. 2). The S-configuration for C-3 of HMG was deter-
compound 12 revealed signals of a flavone aglycone, two β-glucopyr-
1
13
mined as a natural product since naturally occurring HMG esters formed
anosyl units, and an HMG moiety (Table 1). The H and C NMR data of
compound 12 were almost identical to those of compound 11, except for
the addition of a methoxy group at C-5 and the absence of a singlet
via the acylation of the hydroxyl group with (S)-HMG-CoA possess an
(
S)-configuration at the C-3 stereogenic carbon [27,28]. These data led
′
us to conclude that the structure of compound 8 was 5,4 -dihydroxy-
proton H-8. In addition, the HMBC spectrum indicated a linkage be-
′
′′
6
,7,8,3 -tetramethoxyflavone
3-O-[(3S)-3-hydroxy-3-methyl-methyl-
tween H-1 (δ
H
5.11) and C-8 (δ
H
125.1), which suggested that the po-
5.11) was located at C-8, instead
of at C-7 in compound 11 (Fig. 2). Furthermore, the HMBCs between
glutaroyl (1 → 6)]-β-ᴅ-glucoside and named citrusunshin B.
sition of this β-glucopyranosyl unit (δ
H
Compound 9 was isolated as a yellow powder. Its molecular formula
′
′
was established as C39
H
50
O
23 by the HR-FAB-MS ion at m/z 887.2823
23, 887.2821) (Fig. S1.27, Supporting In-
methoxy groups δ 3.92 (5-OCH ) and 3.99 (3 -OCH ) with C-5 and C-3
H 3 3
+
[
M+H] (calcd for C39
H
51
O
confirmed the location of these groups. The connection of two glucose
formation). Its IR spectrum showed absorption bands at 3386, 1722,
651, 1562, and 1430, which suggested that hydroxy, carbonyl, and
moieties via an HMG moiety was also determined by the HMBC cross-
′
′
′′′
′′′
′′′
1
peaks of H-6 /C-1 and H-6 ’/C-5 (Fig. 2). The positions of the
remaining functional groups were based on the HMQC and HMBC
spectra (Figs. S1.43 and S1.44, Supporting Information). Therefore,
compound 12 was elucidated to be 3,7,4ʹ-trihydroxy-5,3ʹ-dimethyl-
1
13
aromatic groups were present. The H NMR and C NMR spectra
exhibited a flavonol skeleton, two sugar units, an HMG moiety, and six
methoxy groups (Table 1). The 1D NMR data of compound 9 were
similar to those of compound 8 except for the replacement of a methoxy
flavone-8-O-[β-ᴅ-glucopyranosyl-(6
→
5)-(3S)-3-hydroxy-3-methyl-
′
′′
group at the C-5 (δ
C
51.9) of HMG moiety by a glucose unit (δ
H
5.28, δ
C
glutarate-(1 → 6)-β-ᴅ-glucopyranoside and named citrusunshin E.
Compound 13 was isolated as a yellow powder and observed in the
1H and 13C NMR spectra as a mixture of anomeric forms (13a, 13b), with
9
4.2, 72.5, 77.9, 69.5, 76.4, 60.1) in compound 9 (Fig. 2). This was also
′
′′
confirmed by the HMBC of the anomeric proton H-1 ’ (δ
H
5.28) with C-
(171.4) (Fig. 2). Acid hydrolysis of compound 9 produced ᴅ-glucose
using TLC (CHCl O = 8:5:1, R 0.30) and optical rotation
–MeOH–H
+50.8 (c 0.2, H O) in comparison with those of authentic samples
10,25,26]. The orientation of the glucopyranosyl moieties was deduced
to be β, according to the large coupling constant of the anomeric protons
at δ
′
′′
5
a ratio of ~1:1. The molecular formulas of 13a and 13b were established
+
3
2
f
as C35
H
42
O
22 by the HR-FAB-MS ion at m/z 815.2251 [M+H] (calcd for
2
D
1.0
1
35 43
C H O22, 815.2246) (Fig. S1.54, Supporting Information). The H
[
α
]
2
NMR spectrum of compounds 13a and 13b was similar to those of the
co-isolated limocitrunshin (10), except for the addition of two anomeric
[
protons δ
H
5.11 (1H, d, J = 3.6 Hz) and 4.52 (1H, d, J = 7.8 Hz)
H
5.42 (1H, d, J = 7.3 Hz) and 5.28 (1H, d, J = 8.2 Hz). Furthermore,
(
Table 2). These anomers were determined to be
α
- and β-ᴅ-glucopyr-
a comparison of the NMR data of compounds 8 and 9 also indicated the
′
anosyl based on their coupling values (13a: J = 3.6 Hz and 13b: J = 7.8
presence of two extra methoxy groups δ
C-5 and C-4 in 9. The HMBC correlations between these methoxy groups
H
3.84 (6H, 5-OCH
3
, 4 -OCH
3
) at
′
Hz). In both cases, the locations of additional
α
- and β-ᴅ-glucopyranosyl
′
′′
′′′
′
were identified at C-5 of limocitrunshin by the HMBC between H-6 ’
and C-5 and C-4 further confirmed their locations (Fig. 2). Thus, com-
′
′′
′′′
′
′
and C-5 (Fig. 2) as well as the evidence of the downfield shift of C-6 to
64.4 (Table 2). Therefore, the structure of compound 13 was deter-
pound 9 was defined as 5,6,7,8,3 ,4 -hexamethoxy-flavonone-3-O-[β-ᴅ-
glucopyranosyl(6 → 5)-3(S)-3-hydroxy-3-methylglutarate-(1 → 6)]-β-ᴅ-
glucopyranoside and named citrusunshin C.
δ
C
mined to be limocitrunshin-6-O-glucopyranoside and named citrusun-
shin F.
Compound 11 was obtained as a yellow powder. Its molecular for-
Compound 14 was isolated as a yellow powder. Its molecular for-
mula was determined as C34
H
40
O
21 by the HR-ESI-MS ion at m/z
40
H O21Na, 807.1960) (Fig. S1.36,
+
mula was established as C36
H
44
O
22 by the HR-FAB-MS ion at m/z
8
07.1990 [M+Na] (calcd for C34
+
8
45.2355 [M+H] (calcd for C36
H O22, 845.2352) (Fig. S1.63, Sup-
45
Supporting information). Its IR spectrum suggested the presence of hy-
ꢀ 1
ꢀ 1
porting Information). Compound 14 was also determined to be a
droxy (3386 cm ), carbonyl (1722, 1651 cm ), and aromatic (1562,
1
ꢀ
mixture of
of compound 14 was very similar to that of compound 13, except for an
extra methoxy group at δ 3.92. The position of this methoxy at C-6 was
α- and β-anomers of glucose (Table 2). The H NMR spectrum
1
430 cm 1) groups. The 1H NMR spectrum showed signals for five ar-
H
omatic protons corresponding to the meta coupled protons δ 6.35 (1H,
d, J = 2.0 Hz, H-6) and 6.74 (1H, d, J = 2.0 Hz, H-8) on the A ring and
H
′
confirmed by the HMBC from 6-OCH3 (δH 3.92) to C-6 (δC 133.0)
(Fig. 2). Thus, the structure of compound 14 was elucidated as 6-
methoxy-limocitrunshin (5 → 6)-O-glucopyranoside and named cit-
rusunshin G.
the ABM coupling system δ
H
7.88 (1H, d, J = 1.8 Hz, H-2 ), 7.63 (1H, dd,
′
′
′
J = 1.8, 8.5 Hz, H-6 ), and 6.91 (1H, d, J = 8.5 Hz, H-5 ) on the B ring of
the flavonoid. A methoxy group at δ 3.94 (3H, s, 3 -OCH ) was also
H 3
observed (Table 1). The 13C spectrum revealed signals for 34 carbons, of
Compound 15 was isolated as a yellow powder. Its molecular for-
which 16 signals were ascribed to isorhamnetin, including a carbonyl
′
mula was established as C37H46O by the HR-FAB-MS ion at m/z
group (δ
C
179.0, C-4) and methoxy group (δ
C
56.7, 3 -OCH
3
). The
23
+
8
59.2511 [M+H] (calcd for C37
47
H O23, 859.2508) (Fig. S1.72, Sup-
remaining signals were attributed to two glucopyranosyl units and an
HMG moiety, which were verified by HMBCs. After acid hydrolysis, the
sugar units were confirmed to be ᴅ-glucose by the same method used for
compounds 8 and 9. The β-configuration for the two sugar moieties was
also determined based on the large coupling constant (J = 7.6 Hz) of
porting Information). The 1D NMR spectra of compound 15 also dis-
played the presence of both the - and β- glucose anomers. (Table 2). The
H NMR spectrum of compound 15 was very similar to that of compound
4 except for an extra a methoxy group at δ 4.11. The position of this
methoxy at C-7 was confirmed by the HMBC from 7-OCH (δ 4.11) to C-
α
1
1
H
′
′
′′′
anomeric protons at δ
H
5.25 (2H, H-1 , H-1 ’). The position of one of the
3
H
7
C
(δ 154.3) (Fig. 2). Thus, the structure of 15 was elucidated as 4-dehy-
droxy-6,7-dimethoxy-limocitrunshin (5 → 6)-O-glucopyranoside and
named citrusunshin H.
β-ᴅ-glucopyranosyl moieties at C-7 was confirmed by the HMBC be-
′
′
′′
tween H-1 (δ
H
5.25) and C-7 (δ
C
163.7). The HMBC of H-6 (δ
171.8) and between H-6 ’ (δ
C
(δ 172.2) suggested that these glucose units were connected via an
H
4.30 and
′
′′
′′′
3
.97) with C-1 (δ
C
H
4.45 and 4.16) and C-
′
′′
5
HMG moiety (Fig. S1.34, Supporting Information). Based on the above
7

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
3
.2. Anti-osteoclastogenic effect assay
concentrations of compounds 1–4. The number of osteoclasts decreased
in the presence of compounds 1–4 (Fig. 3A and B). Maslinic acid was
used as a positive control [29]. Maslinic acid also inhibited RANKL-
induced osteoclastogenesis with the IC50 value of 1.2 ± 0.6 µM
(Fig. S3.2, Supporting Information). Next, the effect of these compounds
(1–4) on cell viability of BMMs was determined. None of the active
compounds significantly decreased cell viability of BMMs, even at a
To evaluate the anti-osteoclastogenic effect of the isolated com-
pounds 1–24, we used two standard osteoclast differentiation in vitro
models, RAW264.7 cells model and BMMs model. For primary
screening, RAW264.7 cells were stimulated with RANKL in the presence
of 10
μM of compounds 1–24. Among them, four compounds, including
′
′
dimethylmikanin (1), quercetogetin (3,3 ,4 ,5,6,7-hexamethoxyflavone)
concentration of 30 M. Instead, compound 3 considerably increased the
μ
′
′
(
2),
3,3 ,4 ,5,7,8-hexamethoxyflavone
(3),
3-methoxynobiletin
cell viability of BMMs at concentrations used in this study, suggesting
that the anti-osteoclastogenic activity of compounds (1–4) was not
mediated by decreasing cell viability of BMMs (Fig. 3C).
′
′
(
3,5,6,7,8,3 ,4 -heptamethoxyflavone) (4), inhibited RANKL-induced
formation of TRAP-positive multinucleated osteoclasts (Fig. S3.2, Sup-
porting Information). To further confirm the anti-osteoclastogenic ef-
fects of compounds 1–4, we employed BMMs model. BMMs were
stimulated with Mꢀ CSF and RANKL in the presence of various
Next, we compared the anti-osteoclastogenic effects of these com-
pounds (1–4) in BMMs. Among them, compound 4 showed the most
potent inhibitory effect on RANKL-induced TRAP-positive
Fig. 3. Effects of active flavonols on RANKL-induced osteoclastogenesis in BMMs (A and B). Flavonols inhibit RANKL-induced differentiation of BMMs into oste-
oclasts. (C) Cell viability was measured by the MTT assay. Data are presented as the mean ± SE (*P < 0.05, **P < 0.01, versus vehicle-treated control; n = 3).
8

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
multinucleated osteoclasts. Thus, compound 4 was subsequently tested
for the dose-dependent inhibition of RANKL-induced osteoclast forma-
tion. As shown in Fig. 4A and B, compound 4 exhibited inhibitory effects
in a dose-dependent manner. Therefore, we further investigated the
effects of compound 4 in more detail.
an MTT assay was conducted. Compound 4 did not exhibit cytotoxicity
in BMMs at a concentration of up to 30 M (Fig. 4C). To determine the
μ
inhibitory effect of compound 4 during the early or late stages of
osteoclastogenesis, we tested the effects of compound 4 on RANKL-
induced osteoclast precursor differentiation by treating RANKL-
stimulated BMMs with compound 4 at different times from day 0 to
day 3. Compound 4 extremely inhibited RANKL-induced osteoclast
differentiation on day 1 but did not effectively inhibit osteoclast dif-
ferentiation on day 3 (Fig. 4E), which suggests that compound 4 inhibits
RANKL-induced osteoclast differentiation during an early stage of
differentiation.
By examining whether compound 4 could inhibit actin-ring forma-
tion, an essential characteristic of mature osteoclasts. Stimulation of
BMMs with RANKL and Mꢀ CSF resulted in the formation of an actin-
ring structure, as evidenced by FTTC-phalloidin staining, however, the
size and density of the actin-ring structure significantly decreased in a
concentration-dependent manner in the presence of compound 4
(
Fig. 4D). These results indicate that compound 4 suppressed RANKL-
In osteoclast precursors, c-Fos is known to play an important role
during the early stage of osteoclast differentiation via a RANKL-
mediated pathway [30]. To investigate the mechanisms by which
compound 4 suppressed RANKL-induced osteoclast formation, the
induced formation of mature osteoclasts.
To evaluate whether the inhibition of osteoclast differentiation and
formation by compound 4 was caused by its potential cytotoxic effects,
Fig. 4. Compound 4 inhibits RANKL-induced osteoclastogenesis in BMMs. (A and B) Compound 4 inhibits RANKL-induced differentiation of BMMs into osteoclasts in
a dose-dependent manner. (C) Cell viability was measured using the MTT assay. (D) Compound 4 suppresses RANKL-induced actin-ring formation. (E) Compound 4
inhibits RANKL-induced osteoclast formation during an early stage of differentiation. Data are presented as the mean ± SE (*P < 0.01, versus vehicle-treated control;
n = 3).
9

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
inhibitory effect of 4 on RANKL-induced c-Fos expression was examined.
While BMMs were stimulated, which led to an increase in the expression
of c-Fos, RANKL-induced c-Fos expression was significantly decreased in
a dose-dependent manner (Fig. 5A), indicating that compound 4 sup-
pressed RANKL-induced osteoclastogenesis via modulation of the c-Fos
signaling pathway.
3.3. The structure–activity relationship (SAR) analysis
In this study, the SAR analysis was valuable guidance to identify and
understand new bioactive components from C. unshiu peels. 24 isolated
compounds, including fifteen flavonols (1–15) and nine flavones
(16–24), were evaluated for their anti-osteoclastogenesis effects
(Fig. S3.2, Supporting information). We realized that flavonols inhibited
RANKL-induced osteoclast formation without causing cytotoxicity,
whereas flavones did not show any inhibitory effects at the concentra-
NFATc1 is a master transcription factor of osteoclast formation
following RANKL stimulation. The enhanced expression of NFATc1 was
associated with an efficient induction of mature osteoclasts [31]. A
previous study demonstrated that c-Fos signaling is important for the
activity of the osteoclast differentiation transcription factor NFATc1.
Since compound 4 suppressed RANKL-induced c-Fos expression, we
designed an assay to evaluate the effect of compound 4 on RANKL-
induced NFATc1 expression. As shown in Fig. 5B, the expression level
of NFATc1 decreased dose-dependently. We next determined the effects
of compound 4 on the expression of osteoclastogenesis-related target
genes, such as CtsK and c-Src, which are downstream genes of the
NFATc1 pathway. The expression level of these genes was attenuated
during RANKL-induced osteoclastogenesis (Fig. 5C). These results
demonstrate that compound 4 suppressed RANKL-induced NFATc1
activation by decreasing c-Fos signaling.
tions used in this study. We also found that the OCH
flavonols played an indispensable role in the inhibition of osteoclast
formation (Figs. 1 and 3). When the OCH group at C-3 was replaced by
3
group at the C-3 of
3
other groups, such as OH, glucopyranosyl, or derivatives of glycopyr-
anosyl, compounds 5–15 did not show any inhibitory effects, whereas
compounds 1–4 containing an OCH
3
group at C-3 significantly inhibited
3
RANKL-induced osteoclastogenesis. The importance of the OCH group
at the C-3 of flavonoids was further confirmed by comparison of the
inhibitory effects of multimethoxy flavones (16–21) and multimethoxy
flavonols (1–4). The absence of an OCH
3
group at C-3 in compounds
16–21 resulted in no inhibition of RANKL-induced osteoclast differen-
tiation. In addition, among four active flavonols (1–4), the main out-
comes show that the number of OCH
3
and the position of OCH
3
groups at
groups
C-8 are crucial. For example, compound 4 contained seven OCH
3
Fig. 5. Compound 4 inhibits RANKL-induced c-Fos and NFATc1 activation. (A) RAW264.7 cells were incubated with RANKL in the presence of indicated con-
centrations of compound 4. The expression level of c-Fos was determined by Western blotting. (B) RAW264.7 cells were stimulated with RANKL for 48 h in the
presence or absence of 30 M compound 4, and then fixed and stained with an anti-NFATc1 antibody, followed by an Alexa Fluor 488-conjugated secondary
μ
antibody. DAPI was used to stain nuclei. Scale bar: 10 µm. (C) RAW264.7 cells were stimulated with RANKL (100 ng/ml) in the presence of indicated concentrations
of compound 4 for 48 h. The expression levels of CtsK and c-Src were determined by Western blotting.
1
0

T.O. Vu et al.
Bioorganic Chemistry 107 (2021) 104613
with OCH
ability, while compounds 2 and 3 containing six OCH
exhibited weaker inhibition (Fig. 3). We later found that the OCH
3
group at C-8, and showed the highest anti-osteoclastogenesis
groups, and
group
[9] J.H. Lyu, H.-T. Lee, Effects of dried Citrus unshiu peels on gastrointestinal motility
in rodents, Arch. Pharm. Res. 36 (5) (2013) 641–648, https://doi.org/10.1007/
s12272-013-0080-z.
10] Q.-M. Ngo, H.-S. Lee, V.T. Nguyen, J.A. Kim, M.H. Woo, B.S. Min, Chemical
constituents from the fruits of Ligustrum japonicum and their inhibitory effects on T
cell activation, Phytochemistry 141 (2017) 147–155, https://doi.org/10.1016/j.
phytochem.2017.06.001.
11] P.T. Tran, D.H. Park, O. Kim, S.H. Kwon, B.S. Min, J.H. Lee, Desoxyrhapontigenin
inhibits RANKL-induced osteoclast formation and prevents inflammation-mediated
bone loss, Int. J. Mol. Med. 42 (2018) 569–578, https://doi.org/10.3892/
ijmm.2018.3627.
12] P.T. Tran, T.-M. Ngo, S. Lee, O. Kim, H.N.K. Tran, C. Hwangbo, B.S. Min, J.-H. Lee,
Identification of anti-osteoclastogenic compounds from Cleistocalyx operculatus
flower buds and their effects on RANKL-induced osteoclastogenesis, J. Funct. Foods
3
3
[
[
[
[
at C-8 appeared to enhance the inhibitory effect on osteoclast formation.
Compound 3 showed higher activity than compound 2, both having a
similar number of OCH
3
groups except for the methoxyl group at C-8 in
at C-8,
compound 3, instead of at C-6 in compound 2. Without OCH
3
compounds 1 and 2 displayed similar inhibitory effect on RANKL-
induced osteoclast differentiation. The above results indicate that the
presence of the OCH
crucial role in improving anti-osteoclastogenesis activity and that the
number of OCH groups in the structure of the flavonols promoted the
3
groups at the C-3, C-8 of the flavonoids played a
6
0 (2019) 103388, https://doi.org/10.1016/j.jff.2019.05.044.
3
13] T. H o` rie, Y. Ohtsuru, K. Shibata, K. Yamashita, M. Tsukayama, Y. Kawamura, 13C
inhibition of RANKL-induced osteoclast differentiation.
NMR spectral assignment of the A-ring of polyoxygenated flavones,
Phytochemistry 47 (5) (1998) 865–874, https://doi.org/10.1016/S0031-9422(97)
In conclusion, we described eight new flavonols (7–9, and 11–15),
together with sixteen known compounds (1–6, 10, and 16–24), all iso-
lated from C. unshiu peels. Among them, compounds 1–4 inhibited
RANKL-induced osteoclastogenesis in a dose-dependent manner. In
addition, compound 4 inhibited actin-ring formation during the early
stage of differentiation but had no cytotoxic effects. At the molecular
level, compound 4 suppressed RANKL-induced c-Fos expression, and
subsequently inhibited NFATc1 activation, as well as the expression of
NFATc1 target genes, including CtsK and c-Scr, in a dose-dependent
manner. These results suggest that compound 4 may potentially have
a use as a therapeutic compound for the treatment or prevention of
osteoclast-related diseases, such as osteoporosis.
0
0629-8.
[
[
14] S. Li, C.-Y. Lo, C.-T. Ho, Hydroxylated Polymethoxyflavones and Methylated
Flavonoids in Sweet Orange (Citrus sinensis) Peel, J. Agric. Food Chem. 54 (12)
(2006) 4176–4185, https://doi.org/10.1021/jf060234n.
15] D.K. Kim, K.T. Lee, J.S. Eun, O.P. Zee, J.P. Lim, S.S. Eum, S.H. Kim, T.Y. Shin, Anti-
allergic components from the peels ofCitrus unshiu, Arch. Pharm. Res. 22 (6)
(
1999) 642–645, https://doi.org/10.1007/BF02975340.
[16] M.-A. Beck, H. H ¨a berlein, Flavonol glycosides from Eschscholtzia californica,
Phytochemistry 50 (2) (1999) 329–332, https://doi.org/10.1016/S0031-9422(98)
0
0503-2.
[
[
[
17] M.S. Hifnawy, A.MK. Mahrous, A.A. Sleem, R.MS. Ashour, Comparative chemical
and biological studies of leaves and pollens of Phoenix canariensis hort. ex Chabaud,
Future Med. Chem. 9 (9) (2017) 871–880, https://doi.org/10.4155/fmc-2017-
0
010.
18] H.J. Eom, D. Lee, S. Lee, H.J. Noh, J.W. Hyun, P.H. Yi, K.S. Kang, K.H. Kim,
Flavonoids and a Limonoid from the Fruits of Citrus unshiu and Their Biological
Activity, J. Agric. Food Chem. 64 (38) (2016) 7171–7178, https://doi.org/
Declaration of Competing Interest
The authors have no conflicts of interest to declare.
Acknowledgments
1
0.1021/acs.jafc.6b03465.s001.
19] H. Nagase, N. Omae, A. Omori, O. Nakagawasai, T. Tadano, A. Yokosuka,
Y. Sashida, Y. Mimaki, T. Yamakuni, Y. Ohizumi, Nobiletin and its related
flavonoids with CRE-dependent transcription-stimulating and neuritegenic
activities, Biochem. Biophys. Res. Commun. 337 (4) (2005) 1330–1336, https://
doi.org/10.1016/j.bbrc.2005.10.001.
[
[
[
20] S.H. Lee, B.H. Moon, Y.H. Park, E.J. Lee, S. Hong, Y.H. Lim, Methyl substitution
1
13
This research was supported by the National Research Foundation of
Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning
effects on H and C NMR data of methoxyflavone, Bull. Korean Chem. Soc. 29
(
2008) 1793–1796.
21] R.J. Ferracin, M.F. das G.F. da Silva, J.B. Fernandes, P.C. Vieira, Flavonoids from
the fruits of Murraya paniculata, Phytochemistry 47 (3) (1998) 393–396, https://
doi.org/10.1016/S0031-9422(97)00598-0.
(
(
NRF-2018R1D1A1B07045287) and by the Korea government (MSIT)
No. NRF-2020R1A5A2017323). We are grateful to the Korea Basic
22] R.S. Compagnone, A.C. Suarez, S.G. Leitao, F. Delle Monache, Flavonoids,
benzophenones and a new euphane derivative from Clusia columnaris Engl, Rev.
Bras. Farmacogn. 18 (2008) 6–10.
Science Institute (KBSI) for mass spectrometric measurements.
Appendix A. Supplementary material
[23] Y. Miyake, K. Yamamoto, Y. Morimitsu, T. Osawa, Isolation of C-glucosylflavone
from lemon peel and antioxidative activity of flavonoid compounds in lemon fruit,
J. Agric. Food Chem. 45 (12) (1997) 4619–4623, https://doi.org/10.1021/
jf970498x.
The supporting information including IR, UV, 1D and 2D NMR, HR-
MS of compounds (7–9, and 11–15), 1D NMR of known compounds
[24] H. Nakagawa, Y. Takaishi, N. Tanaka, K. Tsuchiya, H. Shibata, T. Higuti, Chemical
constituents from the peels of citrussudachi, J. Nat. Prod. 69 (8) (2006)
(
1–6, 16–24), and anti-osteoclastogenesis effect of compounds 1–24 to
1
177–1179, https://doi.org/10.1021/np060217s.s001.
this article can be found online. Supplementary data to this article can be
found online at https://doi.org/10.1016/j.bioorg.2020.104613.
[
[
25] S.P.B. Ovenden, M. Cobbe, R. Kissell, G.W. Birrell, M. Chavchich, M.D. Edstein,
Phenolic glycosides with antimalarial activity from Grevillea “Poorinda Queen”,
J. Nat. Prod. 74 (1) (2011) 74–78, https://doi.org/10.1021/np100737q.
26] N. Zhang, W.-X. Huang, G.-Y. Xia, M.B. Oppong, L.-Q. Ding, P. Li, F. Qiu, Methods
for determination of absolute configuration of monosaccharides, Chinese Herbal
Med. 10 (1) (2018) 14–22, https://doi.org/10.1016/j.chmed.2017.12.009.
References
[
[
1] J.T. Lin, J.M. Lane, Osteoporosis: A Review, Clin. Orthop. Relat. Res. 425 (2004)
[27] Y. Hattori, K.-H. Horikawa, H. Makabe, N. Hirai, M. Hirota, T. Kamo, A refined
method for determining the absolute configuration of the 3-hydroxy-3-methylglu-
taryl group, Tetrahedron Asymmetry 18 (10) (2007) 1183–1186, https://doi.org/
1
26–134.
2] L.J. Raggatt, N.C. Partridge, Cellular and molecular mechanisms of bone
remodeling, J. Biol. Chem. 285 (33) (2010) 25103–25108, https://doi.org/
1
0.1016/j.tetasy.2007.05.013.
1
0.1074/jbc.R109.041087.
[28] S. Song, X. Zheng, W. Liu, R. Du, L. Bi, P. Zhang, 3-Hydroxymethylglutaryl flavonol
glycosides from a mongolian and tibetan medicine, Oxytropis racemosa, Chem.
Pharm. Bull. 58 (12) (2010) 1587–1590, https://doi.org/10.1248/cpb.58.1587.
[29] C. Li, Z. Yang, Z. Li, Y.u. Ma, L. Zhang, C. Zheng, W. Qiu, X. Wu, X. Wang, H. Li,
J. Tang, M. Qian, D. Li, P. Wang, J. Luo, M. Liu, Maslinic acid suppresses
osteoclastogenesis and prevents ovariectomy-induced bone loss by regulating
RANKL-mediated NF-κB and MAPK signaling pathways, J Bone Miner Res 26 (3)
(2011) 644–656, https://doi.org/10.1002/jbmr.242.
[
[
3] G.D. Roodman, Cell biology of the osteoclast, Exp. Hematol. 27 (8) (1999)
1
229–1241, https://doi.org/10.1016/S0301-472X(99)00061-2.
4] M. Asagiri, H. Takayanagi, The molecular understanding of osteoclast
differentiation, Bone 40 (2) (2007) 251–264, https://doi.org/10.1016/j.
bone.2006.09.023.
[
5] D.B. Logar, R. Komadina, J. Pre ˇz elj, B. Ostanek, Z. Tro ˇs t, J. Marc, Expression of
bone resorption genes in osteoarthritis and in osteoporosis, J. Bone Miner. Metab.
2
5 (4) (2007) 219–225, https://doi.org/10.1007/s00774-007-0753-0.
[30] J.H. Kim, N. Kim, Signaling pathways in osteoclast differentiation, Chonnam Med.
J. 52 (1) (2016) 12–17, https://doi.org/10.4068/cmj.2016.52.1.12.
[
[
6] J.P. Bilezikian, L.G. Raisz, T.J. Martin, In Principles of Bone Biology, 4 th John
Bilezikian, J., Martin, T. J, Clemens, T., Rosen, C., Elsevier, 2020, pp. 111–131.
7] Y.-C. Oh, W.-K. Cho, Y.H. Jeong, G.Y. Im, M.C. Yang, Y.-H. Hwang, J.Y. Ma, Anti-
Inflammatory Effect of Citrus Unshiu Peel in LPS-Stimulated RAW 264.7
[31] Y.-Y. Choo, P.T. Tran, B.-S. Min, O. Kim, H.D. Nguyen, S.-H. Kwon, J.-H. Lee,
Sappanone A inhibits RANKL-induced osteoclastogenesis in BMMs and prevents
inflammation-mediated bone loss, Int. Immunopharmacol. 52 (2017) 230–237,
https://doi.org/10.1016/j.intimp.2017.09.018.
Macrophage Cells, Am. J. Chin. Med. 40 (03) (2012) 611–629, https://doi.org/
1
0.1142/S0192415X12500462.
[
8] D.W. Lim, Y. Lee, Y.T. Kim, Preventive effects of Citrus unshiu peel extracts on bone
and lipid metabolism in OVX rats, Molecules 19 (1) (2014) 783–794, https://doi.
org/10.3390/molecules19010783.
1
1