Immunology Letters 124 (2009) 102–110
Contents lists available at ScienceDirect
Immunology Letters
journa l homepage: www.e lsev ier .
Grape seed proanthocyanidin extract (GSPE) atte
collagen-induced arthritis
Mi-La Ch -Sil
Yun-Ju W un-
The Rheumatis ity of K
a r t i c l
Article history:
Received 23 M
Received in re
Accepted 3 Ma
Available onlin
Keywords:
Grape seed proanthocyanidin extract
(GSPE)
Oxidative stress
Autoimmune arthritis
Collagen-indu
Antioxidant
ntho
n-ind
intrap
l par
resis
and bone-marrow cells cultured with the receptor activator of nuclear factor B ligand (RANKL) and
macrophage colony-stimulating factor (M-CSF). Intracellular levels of hydrogen peroxide were deter-
mined using carboxy-dichlorodihydrofluorescein diacetate. GSPE treatment significantly attenuated the
severity of CIA in a dose-dependent manner and reduced the histology scores for synovial inflamma-
tion, cartilage erosion, bone erosion, and the number of TRAP+ osteoclasts. GSPE treatment significantly
1. Introdu
Reactive
(O2−), hydr
continuous
act as a ph
and are inv
ing prolifer
[1]. Howev
teins, DNA,
scavenged b
oxide dismu
reductase.N
∗ Correspon
Medicine, Coll
(RhRC), Catho
Korea, South K
E-mail add
1 These auth
0165-2478/$ –
doi:10.1016/j.i
ced arthritis (CIA) reduced the numbers of tumor necrosis factor alpha (TNF-�)- or interleukin 17 (IL-17)-producing cells
in the synovial tissue and the spontaneous production of TNF-� and IL-17 by splenocytes compared
with those in the control mice. The serum levels of type-II-collagen-specific IgG2a and plasma levels of
8-isoprostane in the GSPE-treated mice were significantly lower than those in the control mice. GSPE
dose-dependently suppressed osteoclastogenesis in vitro. GSPE significantly reduced hydrogen perox-
ide production by anti-CD3-monoclonal-antibody-stimulated CD4+ splenocytes. These results indicate
that intraperitoneal injection of GSPE attenuated CIA in mice. GSPE may be useful in the treatment of
rheumatoid arthritis.
© 2009 Elsevier B.V. All rights reserved.
ction
oxygen species (ROS) such as the superoxide anion
ogen peroxide (H2O2), and the hydroxyl radical, are
ly generated during normal cell metabolism. ROS can
ysiological defense system against microbial infection
olved in maintaining normal cellular functions, includ-
ation, apoptosis, and intracellular signal transduction
er, excessive amounts of ROS can damage lipids, pro-
and the extracellular matrix [2]. Therefore, ROS must be
y antioxidants. Enzymatic antioxidants include super-
tase, glutathione peroxidase, catalase, and thioredoxin
onenzymatic antioxidants includeglutathione, vitamin
ding author at: Division of Rheumatology, Department of Internal
ege of Medicine, Holy Family Hospital, Rheumatism Research Center
lic Research Institute of Medical Science, The Catholic University of
orea. Tel.: +82 32 340 7016; fax: +82 32 340 2669.
ress: rmin6403@hanmail.net (J.-K. Min).
ors contributed equally to this work.
A, vitamin C, and vitamin E [3]. Oxidative stress refers to the cellular
state in which the production of ROS is elevated and/or the levels of
antioxidants reduced. Oxidative stress is considered to be involved
in the pathogenesis of atherosclerosis, cancer, neurodegenera-
tive disorders, diabetes, ageing, and autoimmune rheumatological
diseases, including rheumatoid arthritis (RA), systemic lupus ery-
thematosus, and systemic sclerosis [3].
RA is a chronic inflammatory autoimmune disease character-
ized by synovitis, bone destruction with pannus formation, and the
degradation of the articular cartilage. The cause of RA is unknown,
but several studies have suggested that oxidative stress is asso-
ciated with the pathogenesis of RA. Epidemiological studies have
demonstrated an inverse correlation between the dietary intake
of antioxidants and the incidence of RA [4]. Inverse associations
between serum antioxidant levels and the risk of developing RA
havebeen reported [5]. Thedisease activity also correlates inversely
with antioxidant levels and positively with the presence of oxida-
tive stress in patients with RA [6,7]. Increased oxidative enzyme
activity has been demonstrated, together with reduced enzymatic
and nonenzymatic antioxidant levels, in RA sera and synovial flu-
ids [8]. Oxidative damage to proteins, lipids, DNA, cartilage, and
see front matter © 2009 Elsevier B.V. All rights reserved.
mlet.2009.05.001
o1, Yu-Jung Heo1, Mi-Kyung Park, Hye-Jwa Oh, Jin
oo, Ji-Hyeon Ju, Sung-Hwan Park, Ho-Youn Kim, J
m Research Center, Catholic Research Institute of Medical Science, The Catholic Univers
e i n f o
arch 2009
vised form 29 April 2009
y 2009
e 14 May 2009
a b s t r a c t
To examine whether grape seed proa
dant has therapeutic effect on collage
arthritis. Mice were treated with an
Clinical, histological, and biochemica
genesis were determined by tartrate-
com/ locate /
nuates
Park,
Ki Min ∗
orea, Seoul, South Korea
cyanidin extract (GSPE) which is known to act as an antioxi-
uced arthritis (CIA) in mice, an animal model of rheumatoid
eritoneal injection of GSPE (10, 50, or 100mg/kg) or saline.
ameters were assessed. The effects of GSPE on osteoclasto-
tant acid phosphatase (TRAP) staining of the inflamed joints
M.-L. Cho et al. / Immunology Letters 124 (2009) 102–110 103
extracellular collagen has been demonstrated in patients with RA
[9]. Some drugs used to treat RA such as methotrexate, etancercept
and infliximab are known to play essential roles as antioxidative
agents [1]. Lipid peroxidation markers such as serum malondi-
aldehyde a
collagen-in
[10,11]. The
tempol, �-l
of green tea
strated in m
Grape se
lipids, prot
dins are the
and are hi
or trimers
proanthocy
activity tha
vitamin E, a
such as anti
vasodilator
It has b
oxidation-r
atheroscler
[21]. Thebe
lipoprotein
chloasma h
However, li
rheumatoid
potentials i
on the sev
cal, and bio
attenuates
type-II-coll
oxidative st
vitro.
2. Materia
2.1. Animal
DBA/1J m
housed in p
(Ralston Pu
imental pro
Research Et
2.2. Prepara
Bovine t
Kang of the
in its native
described p
2.3. Inducti
CIA was
cations [26
injection of
Freund’s ad
base of the
intraderma
sified in inc
into thehin
tical Compa
intraperiton
after the se
different doses of GSPE (10, 50, or 100mg/kg) five times per 2 days
for 2.5 weeks. The control mice were injected with saline. Blood
samples were collected from all treated and control mice 8 weeks
after the primary immunization and stored at −70 ◦C until use.
inica
seve
ers. T
verit
mu
with
dem
t or
sal b
tarsa
tire l
sum
coll
ed. T
mea
trol
istolo
ind l
hydr
hen
sec
mati
infla
nfiltr
ining
keni
esenc
ted w
ined
term
ing cr
sing
re ex
e ero
eros
mag
ous a
in th
becu
tex,
, 4 =
ular b
rtical
nd m
he re
e-res
com
the
ema
were
eterm
.All h
d obs
easu
seru
eas
nd urine isoprostane are reported to be elevated in
duced arthritis (CIA) compared with those in controls
beneficial effects of antioxidants, including vitamin E,
ipoic acid, N-acetylcysteine, the polyphenolic fraction
, and (−)-epigallocatechin-3-gallate, have been demon-
ice with CIA [12–17].
ed extract is a natural plant constituent and contains
eins, carbohydrates, and polyphenols. Proanthocyani-
most abundant phenolic compounds in grape seeds,
gh-molecular-weight polymers comprised of dimers
of (+)-catechin and (−)-epicatechin [18]. Grape seed
anidin extract (GSPE) has more powerful antioxidative
n other well-known antioxidants, including vitamin C,
nd gallic acid [19]. GSPE has various biological functions
bacterial, antiviral, anti-inflammatory, antiallergic, and
y actions [20].
een suggested that GSPE is an effective therapy for
elated diseases in animal models, including tumors,
osis, gastric ulcer, cataract, and diabetic retinopathy
neficial effects ofGSPE in reducingoxidized low-density
s and postprandial oxidative stress and improving
ave been demonstrated in human clinical trials [22–24].
ttle is known about the effects of GSPE on CIA and
arthritis. To examine whether GSPE has therapeutic
n the treatment of RA, we examined the effects of GSPE
erity of CIA in DBA/1 mice, using clinical, histologi-
chemical parameters. We found that GSPE treatment
the severity of CIA in mice, reducing serum levels of
agen-specific IgG2a, inflammatory cytokine production,
ress, bone destruction in vivo, and osteoclastogenesis in
ls and methods
s
ice (SLC, Inc., Shizuoka, Japan), 6–8 weeks old, were
olycarbonate cages and fed with standard mouse chow
rina, St Louis, MO, USA) and water ad libitum. All exper-
cedures were examined and approved by the Animal
hics Committee of the Catholic University of Korea.
tion of type II collagen
ype II collagenwas kindly providedbyProfessorAndrew
University of Tennessee. Type II collagen was extracted
form from fetal calf articular cartilage and purified as
reviously [25].
on of CIA and GSPE treatment
induced as previously described, with minor modifi-
]. Briefly, male DBA/1J mice were given an intradermal
100�g of bovine type II collagen emulsified in complete
juvant (1:1,w/v; Arthrogen-CIA, Redmond,WA) into the
tail. Two weeks later, the mice were given a booster
l injection of 100�g of bovine type II collagen emul-
omplete Freund’s adjuvant (1:1, v/v; Difco, Detroit, MI)
dpaw.GSPEwaskindlyprovidedbyHanlimPharmaceu-
ny (Seoul, Korea). GSPE, dissolved in saline, was given
eally to three different groups (10, 50, or 100mg/kg)
condary immunization. The mice were injected with
2.4. Cl
The
observ
and se
mary im
of 0–4
0=no e
the foo
the tar
to the
the en
as the
type II
exclud
12. The
the con
2.5. H
A h
fied in
were t
stained
Inflam
0=no
some i
of the l
3 = thic
and pr
infiltra
determ
age de
follow
ited to
3=mo
of bon
(0=no
at low
numer
cation,
the tra
the cor
fication
trabec
ing co
bone a
file of t
Tartrat
with a
ing to
with h
nuclei
were d
al. [30]
blinde
2.6. M
The
were m
l assessment of arthritis
rity of arthritis was determined by three independent
hemicewere observed three times aweek for the onset
y of joint inflammation for up to 8 weeks after the pri-
nization. The severityof arthritiswasassessedonascale
the following criteria, as described previously [27]:
a or swelling, 1 = slight edema and erythema limited to
ankle, 2 = slight edema and erythema from the ankle to
one, 3 =moderate edema and erythema from the ankle
l bone, and 4=edema and erythema from the ankle to
eg. The arthritic score for each mouse was expressed
of the scores of three limbs. The hind paw into which
agen+ incomplete Freund’s adjuvant was injected was
he highest possible arthritis score for a mouse was thus
n arthritis index was used to compare the data among
and experimental groups.
gical assessment of arthritis
eg of each mouse was fixed with 10% formalin, decalci-
ochloric acid, and embedded in paraffin. The sections
stained with hematoxylin and eosin (H&E). The H&E-
tions were scored for inflammation and bone erosion.
on was scored according to the following criteria [28]:
mmation, 1 = slight thickening of the lining layer or
ating cells in the sublining layer, 2 = slight thickening
layer plus some infiltrating cells in the sublining layer,
ng of the lining layer, influx of cell in the sublining layer,
e of cells in the synovial space, and 4= synovium highly
ith many inflammatory cells. Cartilage damage was
with safranin-O staining. The extent of cartilage dam-
ined by safranin-O staining was scored according to the
iteria [28]: 0 =no destruction, 1 =minimal erosion, lim-
le spots, 2 = slight to moderate erosion in a limited area,
tensive erosion, and 4=general destruction. The extent
sion was expressed using a scoring system from 0 to 5
ion, 1 = small areas of resorption not readily apparent
nification, in the trabecular or cortical bone, 2 =more
reas of resorption, not readily apparent at low magnifi-
e trabecular or cortical bone, 3 =obvious resorption of
lar and cortical bone, without full-thickness defects in
loss of some trabeculae, lesions apparent at low magni-
full-thickness defects in the cortical bone and marked
one loss,without distortion of theprofile of the remain-
surface, and 5= full-thickness defects in the cortical
arked trabecular bone loss, with distortion of the pro-
maining cortical surface), as previously described [29].
istant acid phosphatase (TRAP) stainingwas performed
mercial kit (cat. no. 387-A, Sigma, St Louis, MO), accord-
manufacturer’s instructions, omitting counterstaining
toxylin. TRAP+ multinucleate cells with three or more
considered osteoclasts. The numbers of osteoclasts
ined according to the method described by Bendele et
istological assessmentsweremadeby two independent
ervers.
rement of type-II-collagen-specific antibodies
m levels of type-II-collagen-specific IgG2a and IgG1
ured by enzyme-linked immunosorbent assay (ELISA),
104 M.-L. Cho et al. / Immunology Letters 124 (2009) 102–110
as previously described, with minor modifications [31]. Briefly,
microtiter plateswere coatedwith type II collagen (4�g/mL in PBS)
at 4 ◦C overnight, followed by a blocking step for 30min at room
temperature. Serum samples were then diluted 1:8000 in Tris-
buffered sa
0.5% Tween
which the
IgG2a and I
Quantitatio
tively. Stand
in serial dil
arbitrary un
The absorba
reader oper
2.7. Cell pre
Eightwe
were collec
spleens wer
ammonium
and centrifu
resuspende
(Corning, N
CD4+ T cells
of >90%.
2.8. Measur
supernatant
Isolated
� and IL-17
Antibodies
biotinylated
ture and de
(Sigma) wa
of cytokine
standard cu
murine TNF
mined with
2.9. Immun
Joint sp
staining 8 w
chemical st
previously,
joints were
bone decalc
deparaffiniz
Endogenou
Immunohis
peroxidase
kit (Vector
instructions
serum for
had been w
� antibody
Biotechnolo
ulins were u
washed, bio
The slides w
tain ABC ki
the diamin
The slides
mounted.
2.10. Osteoclastogenesis
Murine osteoclasts were generated as previously described,
with minor modifications [32]. Briefly, bone-marrow cells (BMCs)
ormal 6-week-old mice were cultured overnight and the
herent BMCswere collected. TheBMCswere culturedwith�-
um essential medium containing 10% fetal bovine serum for
in the presence of 30ng/mL macrophage colony-stimulating
(M-CSF) and 50ng/mL receptor activator of nuclear factor
d (RANKL), with or without various concentrations of GSPE.
tainingwasperformed. TRAP+ cellswith three ormorenuclei
efined as osteoclasts and the numbers of osteoclasts were
d.
easurement of 8-isoprostane
use plasma sampleswere obtained 8weeks after the primary
ization and stored at −70 ◦C until analysis. The concen-
of 8-isoprostane in the plasma was measured using a
ercially available immunoassay kit (Cayman Chemical Co.,
bor, MI), according to the instructions provided by the man-
rer.
easu
boxy
lar
o det
ncub
ere
son,
son).
tatist
resu
s of d
nces
ults
fects
hritis
mic
epen
mic
hange
rolmi
issolve
diffe
immu
g/kg
S. The
betwe
rthriti
A repr
line (pH 8.0) containing 1% bovine serum albumin and
-20, and incubated in the microtiter plates for 1h, after
plates were washed five times. The concentrations of
gG1 were measured using mouse IgG2a and IgG1 ELISA
n Kits (Bethyl Laboratories, Montgomery, TX), respec-
ard serum from arthritic mice was added to each plate
utions, and a standard curve was constructed to assign
its to the levels anti-type-II-collagen IgG2a and IgG1.
nce values were determined with an ELISA microplate
ating at 450nm.
paration and culture
eks after the primary immunization, themouse spleens
ted for cell preparation and washed twice with PBS. The
e minced and the red blood cells were lysed with 0.83%
chloride. The cells were filtered through a cell strainer
ged at 1300 rpm at 4 ◦C for 5min. The cell pellets were
d in RPMI 1640 medium and plated in 24-well plates
Y, USA) at a concentration of 1×106 cells/well. Splenic
were isolated by magnetic bead separation to a purity
ement of TNF-˛ and interleukin 17 (IL-17) in culture
s
splenocyteswere cultured for 72h. The amounts of TNF-
in the culture supernatants were measured by ELISA.
directed against mouse TNF-� and IL-17 and against
anti-mouse TNF-� and IL-17 were used as the cap-
tection antibodies, respectively. Alkaline phosphatase
s used for the chromogenic reaction. The amounts
s present in the test samples were determined from
rves constructed with serial dilutions of recombinant
-� and IL-17 (R&D Systems). The absorbance was deter-
an ELISA microplate reader at 405nm.
ohistochemical analysis of TNF-˛ and IL-17
ecimens were collected for immunohistochemical
eeks after the primary immunization. Immunohisto-
aining for TNF-� and IL-17 was performed as described
with a minor modification [31]. Briefly, whole ankle
fixed in 4% paraformaldehyde, decalcified in EDTA
ifier, and embedded in paraffin. The tissue slides were
ed with xylene and rehydrated in a gradient of ethanol.
s peroxidase activity was quenched with 3% H2O2.
tochemistry was performed with the avidin–biotin
complex (ABC) technique using the Vectastain ABC
Laboratories, USA), according to the manufacturer’s
. Nonspecific binding sites were blocked with normal
30min at room temperature. After the tissue slides
ashed, they were incubated with primary anti-TNF-
(R&D Systems) and anti-IL-17 antibody (Santa Cruz
gy) overnight at 4 ◦C. Isotype-matched immunoglob-
sed as the negative controls. After the slides had been
tinylated secondary antibody was applied for 30min.
ere incubated with an avidin–biotin complex (Vectas-
t) for 1h. The final color product was developed using
obenzidine chromogen (Dako, Carpinteria, CA, USA).
were counterstained with Mayer’s hematoxylin and
from n
nonad
minim
5 days
factor
B ligan
TRAP s
were d
counte
2.11. M
Mo
immun
tration
comm
Ann Ar
ufactu
2.12. M
Car
Molecu
used t
then i
cells w
Dickin
Dickin
2.13. S
The
Group
Differe
3. Res
3.1. Ef
Art
tion of
dose-d
treated
Fig. 1. C
thecont
GSPE, d
to three
ondary
50, 100m
with PB
by GSPE
for the a
control.
rement of intracellular H2O2
-dichlorodihydrofluorescein diacetate (H2DCFDA;
Probes), an oxidation-sensitive fluorescent dye, was
ect H2O2. The cells were washed twice with PBS and
ated with 5�M H2DCFDA for 10min at 37 ◦C. The
analyzed with a FACSCalibur flow cytometer (Becton
San Jose, CA, USA) using Flowjo software (Becton
ical analysis
lts are expressed as means± S.D. (or means± S.E.M.).
ata were compared using the Mann–Whitney U-test.
were considered statistically significant at P<0.05.
of GSPE treatment on arthritis scores
developed about 3 weeks after the primary immuniza-
e with type II collagen (Fig. 1). There was a significant
dent reduction of themean arthritis scores of theGSPE-
e compared with those of the control mice (Fig. 1).
s in the arthritis scores of GSPE-treated mice compared with those of
ce. CIAwas inducedaspreviouslydescribed,withminormodifications.
d in phosphate-buffered saline (PBS), was given intraperitoneally
rent groups (10, 50, or 100mg/kg), (n=6 mice/group) after sec-
nization. The mice were injected with different doses of GSPE (10,
) five times per 2 days for 2.5 weeks. The control mice were injected
re was a significant dose-dependent reduction in the arthritis scores
en days 24 and 35. Values are means± S.E.M.s (n=5–6). Mean values
s scores are indicated on the graph. *P<0.01 comparedwith the saline
esentative result of at least three independent experiments is shown.
M.-L. Cho et al. / Immunology Letters 124 (2009) 102–110 105
Fig. 2. Histolo tive hi
(100×), safran mati
described in S phs ar
result of at lea
3.2. Effects
The effe
assessed by
immunizati
In the CIA
tory cells, sy
cartilage an
reduction in
or 100mg/k
arthritis sco
but this wa
destr
Fig. 3. Effects
each group) w
*P<0.01 versu
gical assessment of arthritis in mice with CIA after treatment with GSPE. Representa
in-O (200×), and TRAP (100×) (a). The effects of GSPE treatment on synovial inflam
ection 2 (b). The results are expressed as the mean (±S.E.M.) scores. Photomicrogra
st three independent experiments is shown.
of GSPE treatment on histological findings tilage
ct of GSP