Chemico-Biological Interactions 194 (2011) 97105

Contents lists available at SciVerse ScienceDirect

Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint

Anthraquinone compounds from Morinda ofcinalis inhibit osteoclastic bone

resorption in vitro
Leilei Bao a,b, Luping Qin a, Lei Liu a,c, Yanbin Wu c, Ting Han a, Liming Xue a, Qiaoyan Zhang a,
a
Department of Pharmacogonosy, School of Pharmacy, Second Military Medical University, 325 Guohe Road, Shanghai 200433, China
b
411 Hospital of PLA, Shanghai 200434, China
c
Department of Pharmacy, Fujian University of Traditional Chinese Medicine, Fuzhou 350108, China

a r t i c l e i n f o a b s t r a c t

Article history: The root of Morinda ofcinalis has been claimed to have a protective effect against bone loss in sciatic neu-
Received 20 May 2011 rectomized and ovariectomized osteoporotic rats, and this protective effect is supposed to be attributed
Received in revised form 30 August 2011 to anthraquinone compounds in the plant. In the present study, we investigated the effects of three
Accepted 31 August 2011
anthraquinones isolated from M. ofcinalis, including 1, 3, 8-trihydroxy-2-methoxy-anthraquinone (1),
Available online 10 September 2011
2-hydroxy-1-methoxy-anthraquinone (2) and rubiadin (3) on bone resorption activity in vitro and the
mechanism on osteoclasts derived from rat bone marrow cells. Compound 1, 2 and 3 decreased the for-
Keywords:
mation of bone resorption pits, the number of multinucleated osteoclasts, and the activity of tartrate
Morinda ofcinalis
Anthraquinone
resistant acid phosphates (TRAP) and cathepsin K in the coculture system of osteoblasts and bone mar-
Osteoclast row cells in the presence of 1, 25-dihydroxyvitamine D3 and dexamethasone. They also enhanced the
Bone resorption apoptosis of osteoclasts induced from bone marrow cells with M-CSF and RANKL. In addition, Compound
Apoptosis 1, 2 and 3 improved the ratio of mRNA and protein expression of OPG and RANKL in osteoblasts, inter-
fered with the JNK and NF-jB signal pathway, and reduced the expression of calcitonin receptor (CTR)
and carbonic anhydrase/II (CA II) in osteoclasts induced from bone marrow cells with M-CSF and RANKL.
These ndings indicate that the anthraquinone compounds from M. ofcinalis are potential inhibitors of
bone resorption, and may also serve as evidence to explain the mechanism of the inhibitory effects of
some other reported anthraquinones on bone loss.
2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction have inhibitory effects on bone resorption of osteoclasts derived

Morinda ofcinalis How (Rubiaceae), locally known as bajitian,
is a small vein that grows widely in tropical and subtropical re-
gions. Morinda root has long been used in China and northeast Asia
as tonics for nourishing kidney, strengthening bone and enhancing
immunofunction in the treatment of impotence, menstrual disor-
ders, osteoporosis, diabetes and inammatory diseases such as
rheumatoid arthritis and dermatitis [1]. Morinda root preparations
could be administered to decrease bone loss and increase bone
mineral density in sciatic neurectomized and ovariectomized oste-
oporotic rats [2,3]. Morinda root contains polysaccharides, avone
glycosides, iridoid glycosides and anthraquinones such as rubiadin
and rubiadin-1-methyl ether [1,4]. Morinda polysaccharides are in-
volved in modulating bone formation and increasing proliferation
of osteoblasts and alkaline phosphate activity [5]. Iridoid glyco-
sides possess the anti-nociceptive and anti-inammatory activities
[6]. Anthraquinones existing in other plants of Genus Morinda
including Morinda citrifolia, Morinda lucida and Morinda morindoides
Corresponding author. Tel.: +86 21 81871303; fax: +86 21 81871305.
E-mail address: zqy1965@163.com (Q. Zhang).
from bone marrow cells [1,7], suggesting that M. ofcinalis root
may act as both a suppressor of bone resorption and an enhancer
of bone formation in vivo, and may have some favorable effects
for preventing and treating bone loss-related diseases.
To elucidate the mechanism of the inhibitory effects of Morinda
root on bone resorption, we further isolated anthraquinone com-
pounds and evaluated their effects on multinucleated osteoclasts
derived from bone marrow cells. Osteoclasts are multinucleated
cells (310 nuclei per cell) differentiating from hematopoietic pre-
cursors, and supposed to be involved in bone remodeling. Through
osteoclastic bone resorption, osteoclasts play both a crucial physi-
ological role in bone remodeling and a pathological role in diseases
involving abnormal bone lysis such as osteoporosis, rheumatoid
arthritis and periodontal bone erosion characterized by excess
activity of osteoclasts [8]. Osteoclasts are formed and differenti-
ated under the control of some cytokines, such as macrophage col-
ony stimulation factor (M-CSF), receptor activator of NF-jB ligand
(RANKL) and osteoprotegerin (OPG) [9]. The binding of RANKL to
its receptor RANK leads to recruitment of TNF receptor associated
factor 6 (TRAF6) to the cytoplasmic domain of RANK, and activates
the downstream targets of TRAF6 including transcription factors

0009-2797/$ - see front matter 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.cbi.2011.08.013
98 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105
R5 O R1
R2
R4 R3
O methasone from bone marrow cells in a coculture system with
R1 R2 R3 R4 R5
1 OH OCH3 OH H OH
2 OCH3 OH H H H disarticulated from Wistar rats aged 3 days, and the ends were re-
3 OH CH3 OH H H
Fig. 1. Chemical structures of anthraquinone compound 1, 2 and 3 isolated from
root of Morinda ofcinalis. 1: 1, 3, 8-trihydroxy-2-methoxy-anthraquinone; 2:
2-hydroxy-1-methoxy-anthraquinone; 3: rubiadin.
such as nuclear factor kappa B (NF-jB), activator protein-1 (AP-1)
and nuclear factor of activated T cells (NFAT), as well as the mito-
gen-activated protein kinases (MAPK) including p38 MAP kinases,
c-Jun N-terminal kinases (JNK/SAPK), extracellular-signal regu-
lated kinases (ERK), and phosphatidyl-inositol-3-kinase (PI3K)/
Akt. These signaling cascades induce and activate osteoclastogenic
transcription factors such as c-Fos and NFATC1 [1013]. Osteopro-
tegerin (OPG), which is a soluble decoy receptor of RANKL, nega-
tively regulates osteoclast differentiation and bone resorption.
Mature osteoclast formation is also associated with the expression
of differentiation markers, including tartrate resistant acid phos-
phates (TRAP), cathepsin K, calcitonin receptor (CTR) and carbonic
anhydrase II (CA II) [14]. Modulating these differentiation-related
signaling pathways has been considered as a therapeutic strategy
for the treatment of skeletal diseases.
There are few reports concerning the inhibitory effect of Morin-
da root on bone resorption. In the present study, we used three
anthraquinone compounds including 1, 3, 8-trihydroxy-2-meth-
oxy-anthraquinone (1), 2-hydroxy-1-methoxy-anthraquinone (2)
and rubiadin (3) (Fig. 1) isolated from Morinda root to examine
their effects on osteoclast differentiation-related signaling path-
ways, in an attempt to gain new insights into the mechanism of
Morinda root in bone resorption suppression and the possibility
of using it in the prevention and treatment of excessive bone
resorption-related diseases.
2. Materials and methods
2.1. Reagents
Reagents used in this study included a-modied minimum
essential medium (a-MEM) and fetal bovine serum (FBS) (Gibco,
US); 1, 25-dihydroxyvitamin D3, dexamethasone, trypsin and
coomassie brilliant blue G-250 (Sigma, US); recombinant rat
M-CSF (40028) and RANKL (40030) (Peprotech EC, US); OPG
(sc11383), RANKL (sc9073) and b-actin (sc81178) antibody (Santa
Cruz, US); anti-phospho-Akt (9275), anti-phospho-ERK (3371),
anti-phospho-JNK (9251), anti-phospho-p38 (9211) and anti-
phospho-I-jB (9245) (Cell Signaling Technology, Beverly, MA);
PrimeScript RT Reagent Kit (Perfect Real Time DRR037S) and
SYBR Premix Ex Taq (Perfect Real Time DRR041A) (Bao Biotech
Company, China); NucBuster Protein Extraction Kit (Merck, Nova-
gen, Germany); Cathepsin K Activity Kit (Biovision, US); Annexin
V-FITC Apoptosis Detection Kit (Keygentec, China); Diolamine,
potassium sodium tartrate, disodium 4-nitrophenylphosphate,
Triton X-100 and 4-nitrophenol of domestic AR grade. Anthraqui-
none compounds from M. ofcinalis How root were isolated and
identied in our laboratory as described previously [7]. The chem-
ical structures of the anthraquinone compounds are shown in
Fig. 1.
2.2. Cell culture
Osteoclasts induced with 1, 25-dihydroxyvitamine D3 and dexa-
osteoblasts. Primary osteoblastic cells were prepared from neona-
tal rat calvarial osteoblasts according to the literature [15]. Induc-
tion and culture of osteoclasts are as follows. Briey, the femur was
moved. Bone marrow cells were ushed out using a 1-mL syringe.
Primary osteoblastic cells (1  105/mL) and bone marrow cells
(1  106/mL) were cocultured in a-MEM medium containing 10%
FBS, 1, 25-dihydroxyvitamine D3 (10 nmol/L) and dexamethasone
(100 nmol/L) at 37 C in a humidied atmosphere of 5% CO2. Prior
to plating the cells, cover glasses (5  5 mm) or dental slices
(40 lm thick) were put into culture dishes. The formation of osteo-
clast-like multinucleated osteoclasts (MNCs) was conrmed by the
staining of tartrate resistant acid phosphates (TRAP) and resorption
pit formed on dental slices. All cultures were maintained at 37 C in
a humidied atmosphere of 5% CO2.
Osteoclasts induced with M-CSF and RANKL from bone marrow
cells 5  107 bone marrow cells were cultured in a 6-well plate
containing a-MEM supplemented with 10% FBS, 100 U/mL penicil-
lin, 100 lg/mL streptomycin and 5 ng/mL M-CSF in a humidied
atmosphere of 5% CO2 for 24 h. Non-adherent cells were collected
and cultured in a-MEM medium containing 10% FBS and 50 ng/mL
M-CSF for 3 days. Cells remaining on the bottom of the wells were
considered as bone marrow-derived macrophages (BMM), and cul-
tured in a-MEM containing 10% FBS, 50 ng/mL M-CSF and 100 ng/
mL RANKL. The culture medium was replaced for fresh medium
every 3 days. After 6 days, cells differentiated into mature osteo-
clasts. This experiment was approved by the Bioethic Committee
of the Second Military Medical University (Shanghai, China), and
the procedures of the experiment were strictly performed accord-
ing to the generally accepted international rules and regulations.
2.3. Determination of bone resorption pits
The primary osteoblast and bone marrow cell suspension was
seeded into the wells of 96-well culture plates with a sterilized
dental slice. After 24-h culture in a-MEM medium containing
10% FBS, 1, 25-dihydroxyvitamine D3 (10 nM) and dexamethasone
(100 nM), cells were treated with or without the test compounds
for 12 days. Dental slices were ultrasonicated in 1 mol/L NH4OH
to remove adherent cells and stained with 0.1% toluidine blue solu-
tion. Resorption pits were observed under a microscope, 20 vision
elds of each dental slice were randomly chosen to measure the pit
area with image analysis software (Leica Q550IW, Germany). The
sum of the resorption pit area of 20 random vision elds was cal-
culated as resorption pit area of each dental slice. The inhibitory ef-
fects of anthraquinone compounds on bone resorption were
expressed as the resorption pit area of dental slice treated with test
compounds/resorption pit area of control  100%.
2.4. Counting of positive MNCs
Primary osteoblasts and bone marrow cells were placed on 96-
well plates containing cover glasses and cultured in a-MEM med-
ium containing 10% FBS, 1, 25-dihydroxyvitamine D3 (10 nM) and
dexamethasone (100 nM) for 24 h. Cells were then treated with
or without the test compounds for 10 days, and stained for TRAP
 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105
with a TRAP Staining Kit (No.387A-1KT; SigmaAldrich). Cells with
three or more nuclei were counted as osteoclast-like multinucle-
ated osteoclasts under a microscope.
2.5. Assay for TRAP activity
Primary osteoblasts and bone marrow cells in a-MEM medium
containing 10% FBS, 1, 25-dihydroxyvitamine D3 (10 nM) and dexa-
methasone (100 nM) were placed on 96-well culture dishes, cul-
tured for 6 days, and then treated with or without the test
compounds for 48 h. TRAP was determined as follows: cells were
washed twice with PBS, and then added with 20 lL 0.1% Triton
X-100 to lyse the cells at room temperature. After 15 min, 100 lL
substrate solution (0.4 g disodium 4-nitrophenylphosphate and
2.0 g potassium tartrate, dissolved in 200 mL of deionized water,
adjusting pH to 3.5 with 1 mol/L HCl) was added and incubated
at 37 C for 30 min. To terminate the reaction, 100 lL 1 mol/L
NaOH was added to each well. UV absorbance was measured at
405 nm. At the same time, positive cells for TRAP were counted.
TRAP activity was expressed as nanomoles p-nitrophenol per min-
ute per 100 osteoclasts.
2.6. Determination of cathepsin K activity
Primary osteoblasts (2  106) and bone marrow cells (2  107)
in a-MEM medium containing 10% FBS, 1, 25-dihydroxyvitamine
D3 (10 nM) and dexamethasone (100 nM) were placed on a 24-well
culture plate, cultured for 7 days, and then treated with or without
the test compounds for 48 h. Cells were collected by centrifugation,
lysed in 50 lL chilled CK cell lysis buffer, incubated on ice for
10 min, and then vortexted for 5 min. CK reaction buffer (50 lL)
and 10 mM CK substrate Ac-LR-AFC (2 lL) were added to the
sample. The mixture solution was incubated at 37 C for 2 h. For a
plate-reading set-up, the sample was transferred to a 96-well plate.
Fluoresce intensity was assayed in a uorometer equipped with a
400-nm excitation lter and a 505-nm emission lter.
2.7. Assay of osteoclast cytotoxicity
Osteoclastic cells (1  105) induced with M-CSF and RANKL
were seeded onto the 96-well plates for 8 days, and then treated
with or without the test compounds for 48 h. Cytotoxicity of the
test compounds on osteoclasts was evaluated by MTT assay.
2.8. Flow cytometry quantication of apoptosis
Apoptosis was assessed using Annexin V, a protein that binds to
phosphatidyl serine (PS) residues that are exposed on the surface
of apoptotic cells [16]. Osteoclastic cells (1  106) in the culture
system of M-CSF and RANKL were seeded onto the 6-well plates
for 8 days, during which cells were treated with test compounds
for 1, 2 and 6 days. After treatment, cells were digested with tryp-
sin without EDTA, washed twice with PBS (pH 7.4), and resus-
pended in staining buffer containing 1 lg/mL PI and 0.025 lg/mL
Annexin V-FITC. Double labeling was performed at room tempera-
ture for 10 min in the dark. Cells were immediately analyzed by
ow cytometry.
2.9. Real time PCR analysis
For real time PCR analysis, cells from rat neonatal calvarial oste-
oblasts or osteoclasts in the culture system of M-CSF and RANKL
were cultured in a-MEM medium containing 10% FBS with or with-
out the test compounds on 60-mm culture dishes for 24 h. Total
RNA was extracted with TRIzol reagent (Life Technologies, MD,
USA) according to the manufacturers protocol. The concentration
99
of total RNA was calculated from the absorbance at 260 and
280 nm with a BioPhotometer (Thermo Electron, USA). Primers
were chosen with an on-line primer design program (Rozen and
Skaletsky, 2000). Quantitative real time PCR was performed with
the Applied Biosystems 7300 instrument (Applied Biosystems) in
triplicates using standard curve method. cDNA was synthesized
by using the Superscript II (Invitrogen). The thermal cycling condi-
tion was one time 5 min incubation at 95 C followed by 40 cycles
of 15 s at 95 C and 1 min at 60 C. The PCR primer sequences used
are as follows: OPG, 50 -TGGAGATCGAATTCTGCTTG-30 and 50 -TC
AAGTGCTTGAGGGCATAC-30 ; RANKL, 50 -GGTCGGGCAATTCTGAAT
T-30 and 50 -GGGAATTACAAAGTGCACCAG-30 ; calcitonin receptor,
50 -TTCGCCATAGAGGAGATAA-30 and 50 -CTCGTCCAGGTTCAAAGA-
30 ; carbonic anhydrase II, 50 -ATCCTTGCTCCCTTCT-30 and 50 -CAG-
TTGTCCACCATCAGT-30 ; MPP-9, 50 -CCCACTTACTTTGGAAACG-30
and 50 -GAAGATGAATGGAAATACGC-30 ; c-Fos, 50 -CGAAGGGAAAG-
GA ATAAGA-30 and 50 -GTCCAGGGAGGTCACAGA-30 ; NFATC1, 50 -A
AAAGGTGGAGAAGATTGG-30 and 50 -CGATGTTAGGGTGATTGAG-30 .
All reactions were run in triplicate and analyzed by the 244CT
method [17]. b-actin was used as an internal standard gene.
2.10. Western blot
Cells from rat neonatal calvarial osteoblasts or osteoclasts in the
culture system of M-CSF and RANKL were treated with or without
the test compounds for 24 h. Cells were lysed in a buffer containing
20 mM TrisHCl, 150 mM NaCl, 1% Triton X-100, and protease and
phosphates inhibitors. The lysates (3040 mg) were separated by
10% SDS PAGE and transferred to a polyvinylidene diuoride mem-
brane. After blocking with 5% skim milk, the membrane was
probed with anti-OPG and RANKL for osteoblasts or anti-phospho
Akt, ERK, JNK, p38 and Ij-B for osteoclasts. The same membrane
was stripped and reprobed. Chemiluminescent signals were de-
tected with a Gel Doc 2000 luminescent image analyzer (Wealtec
Dolphin-doc, US).
140
0.1mol/L 1mol/L 10mol/L
120
Bone resorption pit area (% control)
100
80
60 ***
***
***
***
***
*** ***
40
20
0
C 1 2 3
Fig. 2. Inhibitory effects of anthraquinone compound 1, 2 and 3 from Morinda
ofcinalis on osteoclastic bone resorption. Primary osteoblasts and bone marrow
cells were cocultured with a sterilized dental slice in a-MEM medium in the
presence of 1, 25-dihydroxyvitamine D3 (10 nM) and dexamethasone (100 nM), and
treated with or without the test compounds for 12 days. Resorption pits were
observed under a microscope and the pit area was quantitated with image analysis
software. Data were presented as mean standard deviation. The experiments were
repeated 3 times in ve replicate samples (n = 5) P < 0.05, P < 0.01, P < 0.001
compared with control.
100 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105

2.11. Statistical analysis bone resorption pits on the dental slices at concentration of 1 and
10 lmol/L, while compound 3 showed inhibitory effects within the
The experiments were repeated three times in ve replicate dose range of 0.110 lmol/L. Compound 1, 2 and 3 reduced the
samples. Data were expressed as mean standard deviation. dental resorption activity of osteoclasts to 35%, 30% and 38% of
One-way ANOVA followed by Dunnetts t-test was used for statis- control respectively at the concentration of 1 lmol/L.
tical analysis (SPSS 13.0 software, SPSS, USA), and the level of sig-
nicance was set at P < 0.05 (), P < 0.01 (), P < 0.001 ().
3.2. Inhibiting formation and differentiation of osteoclasts

3. Results As shown in Fig. 3A, after 10-day treatment with the anthraqui-
none compounds in the coculture system of primary osteoblasts
3.1. Inhibiting osteoclastic bone resorption and bone marrow cells, the number of TRAP-positive multinucle-
ated cells reduced signicantly. Compound 1, 2 and 3 decreased
The effects of anthraquinone compound 1, 2 and 3 from M. of- the number of osteoclast to 74%, 76% and 82% of control respec-
cinalis on osteoclastic bone resorption are shown in Fig. 2. A tively at the concentration of 10 lmol/L.
homogenous surface was seen in the untreated dental slices and TRAP activity was directly related with osteoclastic bone
was eroded by mature osteoclasts, forming resorption pits, which resorption. As shown in Fig. 3B, after 48-h treatment, TRAP activity
were readily identied by toluidine blue staining. After 12-day was signicantly suppressed by compound 3, but not by com-
treatment, osteoclasts produced large numbers of resorption pound 1 and 2. The compound 3 inhibited the TRAP activity to
lacunae on the dental slices of the control group. Anthraquinone 81%, 80% and 76% of control respectively at the concentration of
compound 1 and 2 decreased signicantly the area of osteoclastic 0.1, 1 and 10 lmol/L.

A 0.1mol/L 1mol/L 10mol/L B 120 0.1mol/L 1mol/L 10mol/L

120
TRAP activity ( % control)

100
Number of osteoclasts ( % control )

100 ** **
*
**
***
***
*** ***
** 80 ***

***
80
60
60
40
40

20
20

0 0
C 1 2 3 C 1 2 3

0.1mol/L 1mol/L 10mol/L 0.1mol/L 1mol/L 10mol/L

C 120 D
1.2
Cathepsin K activity ( % control )

100 **
***

*** ***
***
80
Osteoclast viability (OD)

*** 0.8
***

60

40 0.4

20

0 0
C 1 2 3 C 1 2 3

Fig. 3. Inhibitory effects of anthraquinone compound 1, 2 and 3 from Morinda ofcinalis on osteoclasts. (A) Osteoclasts in the coculture system of primary osteoblasts and
bone marrow cells were cultured for 24 h, and then treated with or without the test compounds for 10 days. Cells were stained for TRAP, and those possessing three or more
nuclei were counted as osteoclast-like multinucleated osteoclasts under a microscope. (B) Osteoclasts in the coculture system of primary osteoblasts and bone marrow cells
were cultured for 6 days, and then treated with or without the test compounds for 48 h. TRAP activity was measured by p-nitrophenyl sodium phosphate assay. (C) Cells were
treated as described in TRAP assay, and the activity of cathepsin K was measured by a uorometer equipped with a 400 nm excitation lter and 505 nm emission lter. (D)
Osteoclastic cells induced with M-CSF and RANKL were cultured in 96-wells plates for 8 days, and then treated with or without the test compounds for 48 h; osteoclast
viability was determined by MTT assay. Data were presented as mean standard deviation. The experiments were repeated 3 times in ve replicate samples (n = 5) P < 0.05,

P < 0.01, P < 0.001 compared with control.

 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105 101

Knowing that another possible mechanism of the inhibitory ef-
fect on bone resorption of mature osteoclasts was that these
anthraquinone compounds could reduce bone matrix degradation
by inhibiting cathepsin K (CK) activity [18], we further evaluated
the effects on cathepsin K activity of osteoclasts in a cell-free en-
zyme assay using synthetic substrate Ac-LR-AFC and recombinant
cathepsin K. As shown in Fig. 3C, compound 1 inhibited osteoclas-
tic cathepsin K activity in a dose-dependent manner within the
dose range of 0.110 lmol/L, and compound 2 and 3 showed inhib-
itory effects at the concentration of 1 and 10 lmol/L. Compound 1,
2 and 3 decreased osteoclastic cathepsin K activity to 64%, 77% and
75% of control respectively at the concentration of 10 lmol/L.
Cytotoxic effects of anthraquinone compound 1, 2 and 3 were
then analyzed. Osteoclasts induced from bone marrow cells with
M-CSF and RANKL were treated with increased amounts of com-
pound 1, 2 and 3 for 48 h to examine cell viability by the colorimet-
ric MTT assay. As shown in Fig. 3D, anthraquinone compound 1, 2
and 3 did not cause any cytotoxic effect on the total cell population
(more than 95% cells were induced into TRAP-positive cells) at con-
centration of 0.1, 1 and 10 lmol/L, indicating that the inhibitory ef-
fects of these compounds on TRAP and CK activity of osteoclasts
were not caused by decreasing cell viability.
3.3. Induction on apoptosis of osteoclast
signicantly, especially the early-term apoptosis rate. After 6-day
treatment with compound 1, 2 and 3, the overall apoptosis rate
of osteoclasts reached 37.3%, 33.4% and 35.5% respectively versus
10.3% in the control group, and the early apoptosis rate was
35.1%, 30.6% and 32.9% respectively versus 7.2% in the control
group, suggesting that the anthraquinone compounds increased
cell apoptosis in cultured osteoclasts.
3.4. Enhancing the ratio of OPG and RANKL in osteoblasts
In the coculture system, osteoblasts produce OPG and RANKL to
modulate the differentiation of osteoclast progenitors, of which
OPG inhibits osteoclast formation, and RANKL are required for
osteoclast development [19]. As shown in Fig. 5, treatment with
anthraquinone compound 1, 2 and 3 at concentration of 1 lm/L
for 24 h signicantly increased the expression of OPG, and de-
creased the expression of RANKL, enhanced the ratio of expression
levels of OPG and RANKL in osteoblasts, reaching 1.32, 1.23 and
1.45 times of control respectively for mRNA levels (Fig. 5A), and
1.48, 1.09 and 1.99 times of control respectively for protein levels
(Fig. 5B and C). These results indicate that compound 1, 2 and 3
inhibited osteoclastic bone resorption through mediating the pro-
duction of OPG and RANKL of osteoblasts in a coculture model
system.

The effects of anthraquinone compound 1, 2 and 3 from 3.5. Inhibiting expression of differentiation-related genes of osteoclast
M. ofcinalis on apoptosis of osteoclasts are shown in Fig. 4. After
treatment with compound 1, 2 and 3 for 1, 2 and 6 days at the con- Knowing that osteoclast differentiation is associated with up-
centration of 1 lm/L, the apoptosis rate of osteoclasts increased regulation of specic genes in response to RANKL, we subsequently

A
PI intensity

Annexin V intensity

B
apoptosis percentage (%)

Fig. 4. Effects of anthraquinone compound 1, 2 and 3 from Morinda ofcinalis on the apoptosis rate of osteoclasts. Osteoclasts induced from bone marrow cells by M-CSF and
RANKL were treated with 1 lM of compound 1, 2 and 3 for 1, 2 and 6 days, and cell apoptosis was evaluated by ow cytometry, Annexin V and PI double-staining. (A) The ow
cytometric dot plots of osteoclasts treated with 1 lM of compound 1, 2 and 3 for 6 days. Horizontal axis shows Annexin V intensity, and vertical axis represents PI intensity.
The lines divide each plot into four quadrants: left lower quadrant: living cells; right lower quadrant: early apoptotic cells; left upper quadrant: necrotic cells; and right upper
quadrant: late apoptotic cells. (B) The apoptosis percentage of osteoclasts treated with 1 lM of compound 1, 2 and 3 for 1, 2 and 6 days.
102 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105

A 1.8 A 1.4

Relative mRNA expression of CAII

***
Ratio of mRNA expression of

1.6 ***
*
1.2
1.4 *
OPG and RANKL

1 *
1.2 **
0.8
1
0.8 0.6

0.6 0.4
0.4 0.2
0.2
0
0 C 1 2 3
C 1 2 3
B 1.4

Relative mRNA expression of CTR

B OPG 1.2 *
RANKL 1

0.8
-actin
0.6
C 1 2 3 **
**
0.4
C 2.5
Ratio of protein expression of

***
0.2

2
OPG and RANKL

0
*** C 1 2 3

1.5 Fig. 6. Effects of anthraquinone compound 1, 2 and 3 from Morinda ofcinalis on

*
mRNA expression of carbonic anhydrase II (CA II, A) and calcitonin receptor (CTR, B)
of osteoclasts. Osteoclasts induced from bone marrow cells with M-CSF and RANKL
1 were treated with anthraquinone compound 1, 2 and 3 for 24 h and assessed for
mRNA expression of CA II and CTR by real time PCR. b-actin was used as an internal
reference. Data were presented as mean standard deviation. The experiments
were repeated 3 times in ve replicate samples (n = 5) P < 0.05, P < 0.01,
0.5
P < 0.001 compared with control.

0
C 1
Fig. 5. Effects of anthraquinone compound 1, 2 and 3 from Morinda ofcinalis on
expression of OPG and RANKL of osteoblasts. Primary osteoblastic cells from
neonatal rat calvaria were treated with or without the test compounds for 24 h. (A)
The ratio of mRNA expression of OPG and RANKL were determined by real time PCR.
b-actin was used as an internal reference. (B) Protein expression of OPG and RANKL
was determined by Western blotting. (C) The ratio of protein expression of OPG and
RANKL. Data were presented as mean standard deviation. The experiments were
repeated 3 times in ve replicate samples (n = 5) P < 0.05, P < 0.01, P < 0.001
compared with control.
investigated the effect of anthraquinone compounds on the expres-
sion of carbonic anhydrase II (CA II), calcitonin receptor (CTR), ma-
trix metalloproteinase 9 (MMP-9), c-Fos and NFACT1 using real
time PCR analysis. As shown in Fig. 6, treatment with anthraqui-
none compound 1, 2 and 3 at the concentration of 1 lm/L for
24 h suppressed the mRNA expression levels of carbonic anhydrase
II (Fig. 6A) and calcitonin receptors (Fig. 6B), but did not affect the
mRNA expression levels of mpp-9, c-Fos and NFACT1 (data not
shown), indicating that the anthraquinone compounds regulated
osteoclast differentiation through modulating the expression of
calcitonin receptors and carbonic anhydrase II.
3.6. Interference with RANKL-induced JNK and I-jB activation
Knowing that RANKL stimulation increases phosphorylation of
Akt, ERK, JNK and p38 in osteoclasts [20], and activation of the
NF-jB pathway requires the phosphorylation of I-jB [21], we next
2 3 p- ERK
p- SARK / JNK
p- P38
p- AKT
p- IB
- actin
C 1 2 3
Fig. 7. Effects of anthraquinone compound 1, 2 and 3 from Morinda ofcinalis on
RANKL-induced signaling pathways. Osteoclasts induced from bone marrow cells
by M-CSF and RANKL were treated with compound 1, 2 and 3 for 24 h and assessed
for phosphorylation of JNK, ERK, p38 MAPK and Akt, and the degradation of IjBa. b-
actin was used as an internal reference.
determined whether the anthraquinone compounds affected the
signaling pathways involving these kinases, and found that treat-
ment with anthraquinone compound 1, 2 and 3 blocked phosphor-
ylation of JNK and I-jB of osteoclasts induced by RANKL, and
increased phosphorylation of p38, ERK and Akt (Fig. 7). These
results suggest that the anti-osteoclastogenic effect of anthraqui-
none compound 1, 2 and 3 was caused by disturbing I-jB and
JNK signaling pathways.
4. Discussion
It was found in this study that the three anthraquinone
compounds from M. ofcinalis How were potent inhibitors of
 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105
osteoclastogenesis in RANKL-stimulated bone marrow-derived
macrophages. These inhibitory effects also led to the decrease of
osteoclast-specic genes like CA II and CTR. In addition, the three
anthraquinone compounds inhibited the resorption capacity of
osteoclasts by suppressing NF-jB and JNK signaling pathways
and modulating the ratio of OPG/RANKL of osteoblasts.
In our previous study, we found that the ethanol extract of
Morinda root signicantly decreased serum DPD/Cr and TRAP
levels in ovariectomized rats, suggesting that the extracts had a
potential effect of inhibiting bone resorption [2]. Further investiga-
tion found that anthraquinone compounds, such as physicion,
rubiadin-1-methyl ether, 2-hydroxy-1-methoxy-anthraquinone,
1,2-dihydroxy-3-methyl-anthraquinone, 1, 3, 8-trihydroxy-2-
methoxy-anthraquinone, 2-hydroxymethyl-3-hydroxy-anthraqui-
none, 2-methoxy-anthraquinone inhibited osteoclastic bone
resorption [7], and 1, 3, 8-trihydroxy-2-methoxy-anthraquinone
(1), 2-hydroxy-1-methoxy-anthraquinone (2) and rubiadin (3)
isolated in later research showed stronger inhibitory effects on
osteoclasts and similar action characteristics. In the present study,
we found that treatment of the osteoclasts with anthraquinone
compound 1, 2 and 3 at the concentration of 0.110 lmol/L
reduced the formation of resorption pits in a dose-dependent
manner. Osteoclastic bone resorption is mediated by the formation
of new osteoclasts and the resorption activity of osteoclasts. As
depicted in Fig. 3, exposure of osteoclasts to these compounds
signicantly reduced the number of TRAP-positive multinucleated
cells and the activity of TRAP, and the inhibitory effect of com-
pound 1, 2 and 3 on osteoclast formation was stronger than their
effect on TRAP activity, indicating that these compounds reduced
bone resorption mainly by inhibiting the formation of multinucle-
ated osteoclasts induced from bone marrow cells, and the effect on
osteoclastic differentiation and maturation was less signicant.
Treatment cells with a cytotoxic compound can result in a
variety of cell fates, such as decrease in cell viability, cell apoptosis
or necrosis. In present study, MTT assay (Fig. 3D) showed that
treatment of osteoclast with anthraquinone compounds did not
decrease the osteoclast viability, and ow cytometry analysis
(Fig. 4A) indicated that these compounds also did not cause the sig-
nicant increase of the number of necrotic osteoclast (see the left
upper quadrant of Fig. 4A), but induced osteoclast apoptosis (see
the right upper and right lower quadrant of Fig. 4A and B). The
investigation on the effects of estradiol and resveratrol on osteo-
clastic apoptosis have yield similar results to those reported here
[2223]. The inconsistent of results of MTT assay and ow cytom-
etry analysis may be because that the genes involved in apoptosis
regulation are more sensitive to the stimulation of anthraquinone
compounds than that of enzyme related with cell viability, such
as succinate dehydrogenase in osteoclast, and treatment of
osteoclast with anthraquinone compounds at concentration of
0.110 lmol/L for 48 h is not enough to cause signicant change
of osteoclast viability. May be, treatment of osteoclast with anthra-
quinone compounds at higher dose for a longer time will cause a
signicant decrease in cell viability.
The development of osteoclasts is controlled by cytokines
synthesized by osteoblast, such as receptor activators of NF-jB
ligand (RANKL), osteoprotegerin (OPG) and tumor necrosis factor
a (TNF-a) [14]. RANKL provides a signal to osteoclast progenitors
through the membrane-anchored receptor activator of NF-jB
(RANK) to activate osteoclast differentiation and function. OPG
blocks the interaction between RANKL and the RANK receptor. In
other words, OPG inhibits osteoclastogenesis while RANKL sup-
ports bone resorption of osteoclast [19]. Therefore, bone remodel-
ing can be assessed by the relative ratio of OPG to RANKL. In our
study, as shown in Fig. 5A, B and C, the compound 1, 2 and 3
up-regulated the ratio of mRNA and protein expression of OPG to
RANKL, suggesting that these compounds may act on
103
osteoblast-like cells to alter the OPG/RANKL ratio and inhibit
osteoclastogenesis.
During the process of bone resorption, osteoclast precursors dif-
ferentiate into mature osteoclasts in the presence of RANKL and M-
CSF. Mature osteoclasts secrete hydrogen ions (H+) and proteases
such as cathepsin K and matrix metalloproteinase (MMP)-9 from
the rufed border to dissolve the inorganic and organic compo-
nents of bone. H+ is produced via carbonic anhydrase II (CA II) from
CO2 and H2O in the cytoplasm and secreted extracellularly by H+-
ATPase [24]; acidosis has been shown to increase CA-II mRNA
expression and vacuolar pump H+-ATPase activity of osteoclast
[25]. The inhibition of anthraquinone compounds on CA-II may
contribute to the decreased bone resorption of osteoclasts. Calcito-
nin receptor (CTR) is an osteoclast-specic gene. Calcitonin binding
to its receptor results in contraction of mature osteoclasts and de-
creased bone resorption [26]. Under acidic conditions, CTR mRNA
expression increased, and calcitonin treatment decreased CTR
mRNA expression [25]. Like treatment with calcitonin, treatment
with anthraquinone compounds decreased CA II and CTR mRNA
expression, indicating that the three anthraquinone compounds di-
rectly act on the CA II and calcitonin receptor to inhibit bone
resorption of osteoclasts.
Treatment with anthraquinone compound 1, 2 and 3 induced
the apoptosis of osteoclasts. The regulation of osteoclast apoptosis
contains general apoptosis machinery, and has several features
that differ from those in other cell types. NF-jB, which is one of
the essential transcription factors for osteoclastogenesis, promotes
osteoclast survival by inducing anti-apoptotic genes [26]. JNK acti-
vation was shown to be required for efcient osteoclast differenti-
ation from bone marrow monocytes, and to protect bone marrow
monocytes from RANKL-induced apoptosis during osteoclast dif-
ferentiation [27]. Studies showed that inhibiting the NF-jB or
JNK pathway alone did not increase the basal apoptosis, while
simultaneous inhibition of both pathways markedly enhanced
apoptosis of osteoclasts [28]. In our study, anthraquinones signi-
cantly inhibited the phosphorylation of NF-jB and JNK induced by
RANKL, indicating that the inhibitory effect of the anthraquinones
on osteoclastic bone resorption and the induction on apoptosis
may result from the disturbance in the signal transduction by
RANKL for the activation of NF-jB and JNK.
In the present study, we elucidated the stimulatory activity of
the anthraquinone compounds from M. ofcinalis How on osteo-
clasts, and explored the mechanism of inhibiting bone resorption.
Our ndings are consistent with other reported anthraquinones
with anti-osteoporotic activity. Emodin could increase bone
formation through activating the mRNA expression of bone
morphogenetic protein (BMP-2), alkaline phosphates and the min-
eralization in the differentiation process of mouse osteoblastic
MC3T3-E1 [29]. Diacerein, a drug for arthritis, may inhibit osteo-
clastic bone destruction by inhibiting RANKL expression and
improving OPG expression in MC3T32-E1 cells [30]. Two newly
synthesized anthraquinone molecules have been shown to in-
crease the trabecular bone volume and trabecular thickness in
studies in vivo [31]. These results together with our present study
suggest the therapeutic potentials of anthraquinones in the
treatment of bone resorption-related diseases, and provide the
initiative in the early drug discovery and development.
In summary, we have demonstrated for the rst time that
anthraquinones from M. ofcinalis possess an inhibitory effect on
osteoclastogenesis from primary precursor cells and bone resorp-
tion activity. The molecular mechanisms for this inhibitory effect
are summarized in Fig. 8. The anthraquinones from M. ofcinalis
decreased the formation and differentiation of osteoclasts by
improving the ratio of OPG to RANKL in osteoblasts, enhancing
osteoclastic apoptosis through disturbing JNK signal pathway and
transcription factors such as NF-jB, reducing bone resorption by
104 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105
Bone marrow cells Osteoblast
osteoblast
Osteoclastic precusor cells
OPG, RANKL
NF-B
JNK
Apoptosis
Calcitonin
Osteoclast
osteoclast
receptor
Fig. 8. Schematic representation of the molecular mechanism of the inhibitory effects of anthraquinones from Morinda ofcinalis on bone resorption. The anthraquinones
decreased the formation and differentiation of osteoclasts by improving the ratio of OPG and RANKL in osteoblasts, enhancing osteoclastic apoptosis through disturbing JNK
signal pathway and transcription factors such as NF-jB, reducing bone resorption by modulating expression of CA II and CTR of osteoclasts, and nally decreasing bone loss.
modulating expression of CA II and CTR of osteoclasts, and nally
decreasing bone loss. Although additional experiments are needed
to conrm the efcacy of these compounds in treating disease con-
ditions in vivo, they could possibly be used in the development of a
therapeutic drug for disorders associated with bone loss.
Conict of interest
There are no conicts of interest among the authors.
Acknowledgement
This study was supported by the National Natural Science Foun-
dation of China (NO. 90709023).
References
[1] M.Y. Wang, B.J. West, C.J. Jensen, D. Nowichi, C. Su, A. Palu, G. Anderson,
Morinda citrifolia (Noni): A literature review and recent advances in Noni
research, Acta Pharmacologica Sinica 23 (2002) 11271141.
[2] N. Li, L.P. Qin, T. Han, Y.B. Wu, Q.-Y. Zhang, H. Zhang, Inhibitory effects of
Morinda ofcinalis extract on bone loss in ovariectomized rats, Molecules 14
(2009) 20492061.
[3] B. Seo, S.K. Ku, E.M. Cha, J.H. Park, J.D. Kim, H.Y. Choi, H.S. Lee, Effect of
Morindae Radix extracts on experimental osteoporosis in sciatic
neurectomized mice, Phytother. Res 19 (2005) 231238.
[4] Y.J. Yang, H.Y. Shu, Z.D. Min, Anthraquinones isolated from Morinda ofcinalis
and Damnacanthus indicus, Acta Pharmaceutica Sinica 27 (1992) 358364.
[5] N. Li, H.M. Wang, S.H. Guo, X. Lin, L.P. Zheng, L. Wang, Effects of Morinda root
polysaccharide and its extract on the activity of osteoblast cultured in vitro,
Journal of Clinical Rehabilitative Tissue Engineering Research 23 (2007) 4570
4572. Chineese.
[6] J. Choi, K.T. Lee, M.Y. Choi, J.H. Nam, H.J. Jung, S.K. Park, H.J. Park,
Anti-nociceptive anti-inammatory effect of monoterpene isolated from the
root of Morinda ofcinalis, Biological and Pharmaceutical Bulletin 28 (2005)
19151918.
[7] Y.B. Wu, C.J. Zheng, L.P. Qin, L.N. Sun, T. Han, L. Jiao, Q.-Y. Zhang, J.Z. Wu,
Anti-osteoporotic activity of anthraquinones from Morinda ofcinalis on
osteoblasts and osteoclasts, Molecules 14 (2009) 573583.
[8] G.D. Roodman, Advances in bone biology: the osteoclast, Endocrine Reviews 17
(1996) 308332.
R5 O R1
R2
R4 R3
O
Carbonic
anhydrase II
Bone
resorption
[9] S. Kwan Tat, M. Padrines, S. Theoleyre, D. Heymann, Y. Fortun, IL-6, RANKL,
TNF-a/IL-1: interrelations in bone resorption pathology, Cytokine and Growth
Factor Reviews 15 (2004) 4960.
[10] Z.H. Lee, H.H. Kim, Signal transduction by receptor activator of nuclear factor
kappa B in osteoclasts, Biochemical and Biophysical Research Communications
305 (2003) 211214.
[11] B.R. Wong, D. Besser, N. Kim, J.R. Arron, M. Vologodskaia, H. Hanafusa, Y. Choi,
TRANCE, a TNF family member, activates Akt/PKB through a signaling complex
involving TRAF6 and c-Src, Molecular Cell 4 (1999) 10411049.
[12] H. Takayanagi, S. Kim, T. Koga, H. Nishina, M. Isshiki, H. Yoshida, Induction and
activation of the transcription factor NFATc1 (NFAT2) integrate RANKL
signaling in terminal differentiation of osteoclasts, Developmental Cell 3
(2002) 889901.
[13] N. Ishida, K. Hayashi, M. Hoshijima, T. Ogawa, S. Koga, Y. Miyatake, Large scale
gene expression analysis of osteoclastogenesis in vitro and elucidation of
NFAT2 as a key regulator, Journal of Biological Chemistry 277 (2002)
4114741156.
[14] W.J. Boyle, W.S. Simonet, D.L. Lacey, Osteoclast differentiation and activation,
Nature 423 (2003) 337342.
[15] Z.D. Liu, H.S. Zang, Y.P. Ou-yang, Biological study of growth pattern of
newborn rat calvarial osteoblastic cell in vitro, Acta Anatomic Sin 26 (1995)
157159.
[16] I. Vermes, C. Haanen, H. Steffens-Nakken, C. Reutelingsperger, A novel assay
for apoptosis: ow cytometric detection of phosphatidyl serine expression on
early apoptotic cells using uoresce in labelled annexin V, Journal of
Immunological Methods 184 (1995) 3951.
[17] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using real
time quantitative PCR and the 2-44CT method, Methods 25 (2001) 402408.
[18] P. Garnero, O. Borel, I. Byrjalsen, M. Ferreras, F.H. Drake, M.S. McQueney, The
collagenolytic activity of cathepsin K is unique among mammalian
proteinases, Journal of Biological Chemistry 273 (1998) 3234732352.
[19] S. Khosla, The OPG/RANKL/RANK system, Endocrinology 142 (2001)
50505055.
[20] C.K. Park, Y. Lee, E.J. Chang, M.H. Lee, J.H. Yoon, J.H. Ryu, Bavachalcone inhibits
osteoclast differentiation through suppression of NFATc1 induction by RANKL,
Biochemical Pharmacology 75 (2008) 21752182.
[21] H.H. Kim, J.H. Kim, H.B. Kwak, H. Huang, S.H. Han, H. Ha, Inhibition of
osteoclast differentiation and bone resorption by tanshinone IIA isolated
from Salvia miltiorrhiza Bunge, Biochemical Pharmacology 67 (2004)
16471656.
[22] D. Saintier, V. Khanine, B. Uzan, H.K. Ea, M.C. de Vernejoul, M.E. Cohen-Solal,
Estradiol inhibits adhesion and promotes apoptosis in murine osteoclasts
in vitro, Journal of Steroid Biochemistry and Molecular Biology 99 (2006)
165173.
[23] X. He, G. Andersson, U. Lindgren, Y. Li, Resveratrol prevents RANKL-induced
osteoclast differentiation of murine osteoclast progenitor RAW 264.7 cells
through inhibition of ROS production, Biochemistry and Biophysics Research
Communication 401 (2010) 356362.
 L. Bao et al. / Chemico-Biological Interactions 194 (2011) 97105
[24] J.M. Lion, R. Mentaverri, S. Rossard, N. Jullian, B. Courtois, J. Courtois,
Oligogalacturonic acid inhibit bone resorption and collagen degradation
through its interaction with type I collagen, Biochemical Pharmacology 78
(2009) 14481455.
[25] D.M. Biskobing, D. Fan, Acid pH increases carbonic anhydrase II and calcitonin
receptor mRNA expression in mature osteoclasts, Calcied Tissue International
67 (2000) 178183.
[26] T. Miyazaki, H. Katagiri, Y. Kanegae, H. Takayanagi, Y. Sawada, A. Yamamoto,
Reciprocal role of ERK and NF-kappaB pathways in survival and activation of
osteoclasts, Journal of Cell Biology 148 (2000) 333342.
[27] T. Thalhamer, M.A. McGrath, M.M. Harnett, MAPKs and their relevance to
arthritis and inammation, Rheumatology 47 (2008) 409414.
[28] M. Matsumoto, T. Sudo, T. Saito, H. Osada, M. Tsujimoto, Involvement of p38
mitogen-activated protein kinase signaling pathway in osteoclastogenesis
105
mediated by receptor activator of NF-jB ligand (RANKL), Journal of Biological
Chemistry 275 (2000) 3115531161.
[29] S.U. Lee, H.K. Shin, Y.K. Min, S.H. Kim, Emodin accelerates osteoblast
differentiation through phosphatidylinositol 3-kinase activation and bone
morphogenetic protein-2 gene expression, International Journal of
Immunopharmacology 8 (2008) 741747.
[30] L. Wang, Y.J. Mao, W.J. Wang, Inhibitory effect of diacerein on osteoclastic bone
destruction and its possible mechanism of action, Acta Pharmaceutica Sinica
41 (2006) 555560.
[31] L. Savarino, D. Benetti, N. Baldini, C. Tarabusi, M. Greco, R. Aloisi, A preliminary
in vitro and in vivo study of the effects of new anthraquinones on neutrophils
and bone remodeling, Journal of Biomedical Materials Research 75 (2005)
324332.