close

Вход

Забыли?

вход по аккаунту

?

Involvement of CD44 in induction of matrix metalloproteinases by a COOH-terminal heparin-binding fragment of fibronectin in human articular cartilage in culture.

код для вставкиСкачать
ARTHRITIS & RHEUMATISM
Vol. 48, No. 5, May 2003, pp 1271–1280
DOI 10.1002/art.10951
© 2003, American College of Rheumatology
Involvement of CD44 in Induction of Matrix Metalloproteinases
by a COOH-Terminal Heparin-Binding Fragment of
Fibronectin in Human Articular Cartilage in Culture
Tadashi Yasuda,1 A. Robin Poole,2 Makoto Shimizu,1 Takefumi Nakagawa,1 Sohel M. Julovi,1
Hirokazu Tamamura,3 Nobutaka Fujii,3 and Takashi Nakamura1
effect. Treatment of cartilage with anti-CD44 antibody
or HSPG resulted in significant inhibition of HBFN-f–
stimulated MMP production. Preincubation with peptide V blocked binding of the anti-CD44 antibody to
chondrocytes in cartilage.
Conclusion. Interaction of the peptide V sequence
in HBFN-f with glycosaminoglycans, such as those in
CD44, plays an important role in HBFN-f–stimulated
MMP production in articular cartilage. Because CD44
is up-regulated in osteoarthritic and rheumatoid arthritic cartilage, the role of the interaction between
CD44 and HBFN-f in these pathologies should be of
relevance and should be studied further.
Objective. To investigate the mechanism of induction of matrix metalloproteinases (MMPs) by a 40-kd
COOH-terminal heparin-binding fibronectin fragment
(HBFN-f) containing III12–14 and IIICS domains in
human articular cartilage in culture.
Methods. Human articular cartilage was removed
from macroscopically normal femoral heads and cultured with HBFN-f. MMP secretion into conditioned
media was analyzed by immunoblotting (MMPs 1 and
13) and by gelatin zymography (MMPs 2 and 9). Type II
collagen cleavage by collagenase was monitored in culture by immunoassay. Involvement of specific peptidebinding domains in HBFN-f and the involvement of
CD44 were assessed with synthetic peptides and an
anti-CD44 antibody. Immunofluorescence histochemistry was performed using fluorescein isothiocyanate–
conjugated anti-CD44 antibody.
Results. HBFN-f stimulated production of MMPs
1, 2, 9, and 13 in association with type II collagen
cleavage by collagenase in human articular cartilage.
Peptide V (WQPPRARI) of HBFN-f, which can bind cell
surface heparan sulfate proteoglycan (HSPG), blocked
MMP induction by HBFN-f, while the scrambled peptide V (RPQIPWAR) had no effect. Peptide CS-1 of 25
amino acids in IIICS of HBFN-f caused no significant
The extracellular matrix of articular cartilage is
composed mainly of proteoglycans and collagens, and its
integrity provides the mechanical properties of cartilage
(1). Progressive destruction of cartilage, which results
from an imbalance between the anabolic and catabolic
processes, is a common feature of rheumatoid arthritis
(RA) and osteoarthritis (OA). Interleukin-1␤ (IL-1␤)
and tumor necrosis factor ␣ have been shown to promote cartilage degradation by stimulating the production of matrix metalloproteinases (MMPs) (2).
Fibronectin is a component of normal cartilage
matrix (3). It consists primarily of 3 types of homologous
repeating segments (designated I, II, and III). Fibronectin contains NH2-, gelatin-, cell-, and COOH-terminal
heparin-binding domains. The central cell-binding region has an RGD sequence in domain III10, which is
recognized by several cell surface integrin family members (4). Several sites in the heparin-binding domain that
lie COOH-terminal to the central cell-binding domain
also interact with the cell surface. Regions of domain
III12–14 support cell attachment with varying affinities
(5–7). The IIICS, or variable (V) region, contains the
1
Tadashi Yasuda, MD, PhD, Makoto Shimizu, MD, Takefumi Nakagawa, MD, Sohel M. Julovi, MD, Takashi Nakamura, MD,
PhD: Kyoto University Graduate School of Medicine, Kyoto, Japan;
2
A. Robin Poole, PhD, DSc: Joint Diseases Laboratory, Shriners
Hospital for Children, and McGill University, Montreal, Quebec,
Canada; 3Hirokazu Tamamura, PhD, Nobutaka Fujii, PhD: Kyoto
University Graduate School of Pharmaceutical Sciences, Kyoto, Japan.
Address correspondence and reprint requests to Tadashi
Yasuda, MD, PhD, Department of Orthopaedic Surgery, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin,
Sakyo-ku, Kyoto 606-8507, Japan. E-mail: tadyasu@kuhp.kyotou.ac.jp.
Submitted for publication September 5, 2002; accepted in
revised form January 24, 2003.
1271
1272
YASUDA ET AL
␣4␤1 integrin-binding sites, CS-1 and CS-5 (8–10). The
COOH-terminal heparin-binding domain of fibronectin
is known to bind CD44 (11), a principal hyaluronan
receptor (12).
Elevated levels of fibronectin fragments are
found in OA cartilage (13–15) and in OA synovial fluid
(15,16). The central cell-binding, NH2-terminal heparinbinding, and NH2-terminal gelatin-binding fragments of
fibronectin have been shown to stimulate proteoglycan
breakdown (17) and release of catabolic cytokines (18)
in cultured articular cartilage explants. In addition to
those fibronectin fragments, we recently found that the
40-kd COOH-terminal heparin-binding fibronectin fragment (HBFN-f) containing both the heparin-binding
III12–14 and IIICS domains can stimulate type II collagen cleavage by collagenase following proteoglycan degradation, in association with the production of MMPs 3
and 13 in bovine articular cartilage explant cultures (19).
Thus, increased levels of fibronectin fragments are
thought to be involved in cartilage destruction in OA
and RA through the induction of cytokines and MMPs.
Currently, there is little information regarding
the mechanisms whereby MMPs are induced by fibronectin fragments. MMP production by the central
cell-binding fragment of fibronectin is probably mediated by ␣5␤1 integrin because the anti-␣5␤1 integrin
antibody and the RGD-containing peptide induce
MMP-1 and gelatinase in synovial fibroblasts (20). Recent studies using antisense oligonucleotides to an ␣5
integrin subunit have also shown the involvement of ␣5
integrin in cartilage chondrolysis induced by 29-kd NH2terminal heparin-binding and 50-kd NH2-terminal
gelatin-binding fragments in addition to the cell-binding
fragment of fibronectin (21). However, it remains unclear how other fibronectin fragments, including
HBFN-f, can induce MMP production in cartilage. In
this study, we investigated the mechanism of action of
HBFN-f on MMP production in human articular cartilage. We found that a specific amino acid sequence of
HBFN-f mediated the induction of collagenase and
gelatinase in human articular cartilage, and that this
involves binding of HBFN-f to CD44.
MATERIALS AND METHODS
Antibodies and reagents. Anti-human MMP-1 that
reacts with 53-kd and 51-kd bands of proenzyme (M4177),
anti-human MMP-2 that recognizes the 72-kd band (M4677),
anti-human MMP-9 that reacts with the 92-kd band (M5427),
and anti-human MMP-13 that recognizes the latent proenzyme
at 60 kd (M4052) were obtained from Sigma (St. Louis, MO).
Alkaline phosphatase–conjugated goat anti-rabbit IgG was
purchased from Southern Biotechnology (Birmingham, AL).
OS/37, a monoclonal anti-human CD44 antibody, was obtained
from Seikagaku Kogyo (Tokyo, Japan). Mouse IgG1 was
obtained from ICN Biomedicals (Aurora, OH). A 40-kd
proteolytic COOH-terminal heparin-binding fragment of human plasma fibronectin containing type III12–14 segments and
IIICS (HBFN-f) generated with ␣-chymotrypsin digestion was
obtained from Gibco BRL (Gaithersburg, MD). The purity of
the protein preparation was confirmed using the same method
as used in our previous studies (19). Recombinant human
IL-1␤ was purchased from R&D Systems (Minneapolis, MN).
Purified human MMPs 2 and 9 were obtained from Chemicon
(Temecula, CA). Heparan sulfate purified from porcine intestinal mucosa was purchased from Sigma. Synthetic peptides
CS-1 (DELPQLVTLPHPNLHGPEILDVPST), peptide V
(WQPPRARI), and the scrambled peptide V (RPQIPWAR)
were obtained from Takara Shuzo (Kusatsu, Japan). The
sequence for the scrambled peptide V is identical to that used
in previous studies (22,23).
Articular cartilage explant culture. Adult human articular cartilage was obtained from the femoral head after
replacement surgery for femoral neck fracture. No significant
arthritic changes such as fibrillation, were identified macroscopically in the articular cartilage. The cartilage was placed in
a 24-well plate (⬃80 mg/well) and kept in 2 ml of serum-free
Dulbecco’s modified Eagle’s medium (DMEM) containing 100
units/ml penicillin, 100 ␮g/ml streptomycin, and 10 mM
HEPES (all from Gibco BRL) in a humidified 5% CO2
atmosphere at 37°C. In this study, a serum-free cartilage
explant culture system was used because chondrocytes in
cartilage explants under serum-free conditions are still viable
40 days after treatment with fibronectin fragments (24), and
serum supplementation decreases the chondrolytic activity of
fibronectin fragments (17). The cartilage was precultured for 2
days, and medium was changed on day 0. Thereafter, medium
was replaced every 4 days. Fresh HBFN-f or IL-1␤ was added
(starting on day 0) at each medium change, as indicated below.
Assays of endotoxin levels in 100 nM HBFN-f solution with the
endotoxin assay kit (Sigma) showed minimal levels of detection
of 6–8 ng/ml (19), significantly lower than the concentrations
(1.0 ␮g/ml) that are required to alter cartilage metabolism
(25). In some experiments, heparan sulfate was added with or
without HBFN-f beginning on day 0. In another set of experiments, following preincubation with peptide V, the scrambled
peptide V, or CS-1, or following preincubation with antibody
OS/37 or subclass-matched mouse IgG1 for 2 hours, articular
cartilage was coincubated with HBFN-f from day 0. The
cartilage explants and conditioned media were harvested on
days 4, 8, and 12, and stored at ⫺20°C.
Immunoblot analysis. Conditioned media were heated
with sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer at 80°C for 20 minutes.
Proteins were separated by SDS-PAGE under reducing conditions, and then transferred onto nitrocellulose membranes
(Schleicher & Schuell, Dassel, Germany). Gel loading was
standardized based on the DNA contents of the cartilage
explants. Membranes were blocked in phosphate buffered
saline (PBS) containing 5% nonfat dry milk and 0.1% Tween
20 and incubated with the first antibody (concentration
1:1,000) overnight at 4°C. After incubation with alkaline
phosphatase–conjugated second antibody (concentration
CD44 IN MMP INDUCTION BY FIBRONECTIN FRAGMENT
1273
1:1,000) for 3 hours at room temperature, immunoreactive
bands were visualized using BCIP and nitroblue tetrazolium.
The presence of heparan sulfate itself had no significant effect
on immunoblotting (data not shown). Protein band intensity
was evaluated by densitometry using NIH Image (National
Institutes of Health, Bethesda, MD).
Gelatin zymography. Serum-free conditioned media
were collected from human articular cartilage explant culture.
Samples were prepared in nondenaturing loading buffer and
separated on a 10% SDS–polyacrylamide gel impregnated with
1 mg/ml of gelatin. The amount of sample applied was
determined, as above, by the DNA content in the cartilage
explants. After electrophoresis, gels were washed with 50 mM
Tris HCl (pH 7.5) containing 0.1M NaCl and 2.5% Triton
X-100, and then rinsed with 50 mM Tris HCl (pH 7.5). To
visualize proteinases, gels were incubated at 37°C in an incubation buffer containing 50 mM Tris HCl (pH 7.5), 200 mM NaCl,
10 mM CaCl2, and 0.02% NaN3. Gels were subsequently fixed
and stained in Coomassie blue fixation solution. The areas of
gelatin degraded by gelatinase were measured on the scanned
digital image of stained gels. The presence of heparan sulfate
itself had no significant effect on gelatin zymography (data not
shown).
Extraction and assay for cleavage of type II collagen
by collagenase. Cartilage explants were digested to extract
cleaved type II collagen as previously described (19,26).
Briefly, the harvested cartilage explant was incubated overnight
with 1.0 mg/ml ␣-chymotrypsin at 37°C to cleave and solubilize
denatured collagen. After inhibition of ␣-chymotrypsin activity
with N-tosyl-L-phenylalanine-chloromethyl ketone (Sigma), the
samples were centrifuged, and the supernatants were recovered.
The Col2-3/4Cshort epitope (hereafter referred to as Col2-3/4C)
generated by cleavage of type II collagen by collagenase (26)
was measured in ␣-chymotrypsin extracts by immunoassays.
The release of the Col2-3/4C epitope into media was also
measured by immunoassay (19,27,28). From the measurement of
Col2-3/4C in ␣-chymotrypsin extracts and medium, the total
amount of epitope present in both the tissue and medium for the
period between the last medium change and the harvest was
calculated. The remaining explant residues were digested overnight with 1.0 mg/ml proteinase K at 56°C for DNA assay
(see below).
Immunofluorescence histochemistry. After blocking
with 1% bovine serum albumin for 24 hours, articular cartilage
slices from the femoral head were incubated with fluorescein
isothiocyanate (FITC)–conjugated antibody OS/37 (Seikagaku
Kogyo) at 5 ␮g/ml or subclass-matched FITC-conjugated
mouse IgG1 (KPL, Gaithersburg, MD) at 5 ␮g/ml for 24 hours
at 37°C. In some experiments, after preincubation with peptide
V at 100 ␮M for 1 hour, articular cartilage was then incubated
with FITC-conjugated OS/37 at 5 ␮g/ml. Thereafter, following
an extensive wash with DMEM, cartilage slices were subjected
to cryostat sectioning at 6 ␮m and fixed with 4% paraformaldehyde in PBS for 20 minutes. After counterstaining with propidium iodide (KPL), the sections were examined by confocal
microscopy (FluoView; Olympus, Tokyo, Japan).
Assay for DNA. DNA content was measured in proteinase K digests of articular cartilage explants, as previously
described (29).
Statistical analysis. Comparisons between 2 groups
were performed by Student’s t-test. P values less than 0.05 were
considered significant.
RESULTS
Induction of collagenase and gelatinase by
HBFN-f in human articular cartilage culture. Consistent with our previous study, which showed that
HBFN-f can induce significant type II collagen degradation in bovine articular cartilage by day 12 under
serum-free conditions (19), HBFN-f stimulated the production of collagenases (MMPs 1 and 13) and gelatinases (MMPs 2 and 9) in human articular cartilage
under serum-free conditions. Control cultures without
treatment secreted basal levels of MMP-1, with barely
detectable levels of MMP-13 production, as shown by
Western blotting of culture media. When cartilage was
incubated with HBFN-f at 1, 10, and 100 nM for 4 days,
the fragment induced MMPs 1 and 13 only at 100 nM
(Figure 1A). Immunoblot analysis demonstrated that
HBFN-f at 100 nM also induced MMPs 2 and 9, while
gelatin zymography revealed that the fragment caused
increased secretion of latent and active forms of MMPs
2 and 9 (Figure 1B).
Treatment with 100 nM HBFN-f from days 0 to
12 resulted in enhanced secretion of MMPs 1 and 13
into media during days 0–4, 4–8, and 8–12 (Figure 2A).
IL-1␤ at 2 ng/ml also induced these collagenases; data
from days 8 to 12 are shown in Figure 2A. MMP-1 levels
(mean ⫾ SD) induced by HBFN-f reached a plateau by
day 4, whereas MMP-13 levels increased with time, 20 ⫾
13% (n ⫽ 3) during days 0–4 and 85 ⫾ 5% (n ⫽ 3)
during days 4–8 compared with the protein band intensity during days 8–12 (calculated as 100%). While basal
levels of MMP-2 (4 ⫾ 1%, n ⫽ 3) were observed in
control cultures, HBFN-f enhanced levels of proMMP-2
and active MMP-2 (50 ⫾ 8% and 80 ⫾ 7%, n ⫽ 3) during
days 0–4 and 4–8, respectively, compared with the levels
during days 8–12 (calculated as 100%) (Figure 2B). Similarly, HBFN-f induced a significant secretion of both
proMMP-9 and active MMP-9 with time, while no detectable MMP-9 was found in control cultures (Figure 2B).
HBFN-f–increased levels of MMP-9 were 20 ⫾ 9% and
50 ⫾ 8% during days 0–4 and 4–8, respectively (n ⫽ 3),
when the levels during days 8–12 were calculated as 100%.
Induction of the release of collagenase-generated
cleavage epitope of type II collagen by HBFN-f in human
articular cartilage culture. HBFN-f at 100 nM induced
an enhanced release into medium of the Col2-3/4C
epitope generated by collagenase cleavage of type II
collagen during days 0–8 (Figure 3). This release de-
1274
YASUDA ET AL
to bind other receptors. Therefore, the CS-1 synthetic
peptide was tested to determine the effect of its sequence in HBFN-f–stimulated MMP induction. Treatment with 10 ␮M CS-1 alone resulted in a slightly
enhanced secretion of MMP-1, compared with the
control (Figure 4). When CS-1 was used in conjunction
with HBFN-f, secretion of MMP-1 was enhanced.
HBFN-f–induced levels of MMP-1 in the presence of
10 ␮M CS-1 were 110 ⫾ 18%, 120 ⫾ 23%, and 140 ⫾
27% during days 0–4, 4–8, and 8–12, respectively (n ⫽
3), compared with HBFN-f–induced levels in the absence of CS-1 during days 8–12 (calculated as 100%).
However, CS-1 at 10 ␮M enhanced secretion of MMP-1
only, with no effect on MMPs 13, 2, or 9 until day 12.
Treatment with lower (0.1 and 1 ␮M) or higher (100
␮M) concentrations of CS-1 peptide yielded similar
results (data not shown).
Effects of peptide V derived from HBFN-f on
HBFN-f–stimulated MMP induction in human articular
cartilage explant culture. Peptide V derived from the
III14 repeat of HBFN-f has a known heparin-binding
activity and promotes focal adhesion formation (31).
Figure 1. Induction of matrix metalloproteinases (MMPs) by
COOH-terminal heparin-binding fibronectin fragment (HBFN-f) in
human articular cartilage explant culture. A, HBFN-f at 1, 10, or 100
nM was added from day 0. Secreted levels of MMPs 1 and 13
during days 0–4 in conditioned media were analyzed by immunoblotting. B, HBFN-f at 100 nM was added from day 0. Secreted levels
of MMPs 2 and 9 in conditioned media during days 4–8 were evaluated by immunoblotting and gelatin zymography. Purified MMPs 2
and 9 were used as positive controls. Two separate experiments were
performed, with similar results.
pends upon secondary cleavage of the denatured ␣ chain
bearing the epitope that involves MMP activity (28). Levels
of Col2-3/4C were similar in cultures with or without
treatment with 100 nM HBFN-f on days 4 and 8. Consequently, HBFN-f caused increased generation of the Col23/4C epitope with time during days 0–8 in human articular
cartilage culture. Our previous studies, using the specific
inhibitor of MMP-13, suggest that collagenase 3 is a major
player in type II collagen cleavage caused by HBFN-f (19).
Enhancement of MMP-1 induction by the CS-1
domain of HBFN-f in human articular cartilage explant
culture. Binding of HBFN-f to the ␣4␤1 integrin involves the CS-1 sequence in the IIICS domain of
HBFN-f (8). Although normal chondrocytes scarcely
express ␣4␤1 integrin (30), CS-1 may have the potential
Figure 2. Time course of HBFN-f–induced MMP production in explant culture. Articular cartilage was incubated with HBFN-f at 100
nM from day 0. Control cultures contained no additives.
Interleukin-1␤ (IL-1␤) at 2 ng/ml was used as a positive control for day
12. Conditioned media were collected on days 4, 8, and 12. A, Secreted
levels of MMP-1 and MMP-13 in conditioned media were analyzed by
immunoblotting. B, Secreted levels of MMPs 2 and 9 in conditioned
media were evaluated by gelatin zymography. Three separate experiments were performed, with similar results. Each lane represents a
4-day accumulation of material and not the total accumulation from
day 0 because conditioned media were collected every 4 days. See
Figure 1 for other definitions.
CD44 IN MMP INDUCTION BY FIBRONECTIN FRAGMENT
1275
This prompted an examination of whether peptide V
could influence MMP induction by HBFN-f in human
articular cartilage (Figure 5). Incubation of cartilage
Figure 4. Effects of CS-1 peptide on HBFN-f–induced MMP production in explant culture. CS-1 derived from the IIICS domain of
HBFN-f was added at 10 ␮g/ml from day 0 in the presence or absence
of HBFN-f at 100 nM. Control cultures (C) contained no additives.
Conditioned media were collected on days 4, 8, and 12. Secreted levels
of MMPs 1 and 13 in conditioned media were analyzed by immunoblotting. Secreted levels of MMPs 2 and 9 in conditioned media were
evaluated by gelatin zymography. Three separate experiments were
performed, with similar results. Each lane represents a 4-day accumulation of material and not the total accumulation from day 0. See
Figure 1 for other definitions.
Figure 3. Heparin-binding fibronectin fragment (HBFN-f)–induced
cleavage of type II collagen by collagenase in explant culture. The
collagenase-generated cleavage epitope in type II collagen was measured by enzyme-linked immunosorbent assay in media and in cartilage (the latter following proteolysis of collagen to release epitope).
HBFN-f at 100 nM was added from day 0. Control cultures contained
no additives. Cartilage explants and conditioned media with (open
bars) or without (solid bars) treatment with HBFN-f were harvested on
days 4 and 8. Two separate experiments were performed, with similar
results. Values are the mean and SD of 4 determinations. ⴱ ⫽ P ⬍ 0.05
versus control.
with 10 ␮M peptide V alone had no effect on MMP
production, while HBFN-f induced MMPs 1, 2, 9, and
13. Peptide V at 10 ␮M completely suppressed HBFNf–stimulated production of MMPs 2, 9, and 13 to control
levels, and suppressed HBFN-induced MMP-1 production by 81 ⫾ 4%, 79 ⫾ 7%, and 58 ⫾ 4% during days
0–4, 4–8, and 8–12, respectively (n ⫽ 3). A significant
difference was found between the secreted levels of the
individual MMPs from HBFN-f–treated cultures in the
presence and absence of peptide V (P ⬍ 0.05 by t-test).
In contrast to peptide V, the scrambled peptide V at 10
␮M had no significant effect on HBFN-f–stimulated
MMP production (Figure 6A). In addition, lower concentrations of peptide V (1 ␮M) failed to suppress
HBFN-f–induced MMP (Figure 6B).
Effects of heparan sulfate on HBFN-f–stimulated
MMP production. Cell surface heparan sulfate has been
implicated in cell adhesion to several sites in the
1276
YASUDA ET AL
Involvement of CD44 in HBFN-f–induced MMP
production in human articular cartilage. CD44 is expressed on a variety of cell types, including chondrocytes. While the principal ligand for CD44 is hyaluronan
(12), HBFN-f is also known to bind CD44 (11). Because
antibody OS/37 has been shown to block CD44 binding
to its ligand (34), an attempt was made to clarify the
Figure 5. Suppression of HBFN-f–stimulated production of MMPs 1,
2, 9, and 13 by peptide V in explant culture. Peptide V derived from
III14 domain of HBFN-f was added at 10 ␮M from day 0 in the
presence or absence of HBFN-f at 100 nM. Control cultures (C)
contained no additives. Conditioned media were collected on days 4, 8,
and 12. Secreted levels of MMPs 1 and 13 in conditioned media were
analyzed by immunoblotting. MMPs 2 and 9 secreted into media were
evaluated by gelatin zymography. Three separate experiments were
performed, with similar results. Each lane represents a 4-day accumulation of material and not the total accumulation from day 0. See
Figure 1 for other definitions.
III12–14 domains of HBFN-f, and peptide V can bind
heparan sulfate (6,7,31–33). Thus, 1–100 ␮g/ml of heparan sulfate proteoglycan (HSPG) was added to cartilage
explant cultures to ascertain whether HSPG ligation
with the HSPG-binding sites of HBFN-f including peptide V could block HBFN-f–induced MMP. Coincubation of articular cartilage with 100 ␮g/ml HSPG in the
presence of 100 nM HBFN-f resulted in complete reduction of HBFN-f–stimulated secretion of MMPs 2, 9, and
13, with decreased secretion of MMP-1 by 80 ⫾ 9% (n ⫽
3) (Figure 7). There was a significant difference between
the secreted levels of the individual MMPs from HBFNf–treated cultures in the presence and absence of HSPG
(P ⬍ 0.05 by t-test).
Based on the experiments using peptide V (Figures 5 and 6) and HSPG (Figure 7), we concluded that
HBFN-f–stimulated MMP production in human articular cartilage involves the peptide V domain of HBFN-f.
Figure 6. A, Effects of scrambled peptide V on HBFN-f–stimulated
MMP production. Human articular cartilage was incubated with
HBFN-f at 100 nM from day 0 in the presence or absence of peptide
V or the scrambled peptide V at 10 ␮M. Secreted levels of MMPs 1, 2,
9, and 13 during days 4–8 in the conditioned media were analyzed by
immunoblotting. B, Dose-dependent effects of peptide V on HBFN-f–
stimulated MMP production. HBFN-f at 100 nM was added from day
0 in the presence or absence of peptide V at 1 or 10 ␮M. Secreted
levels of MMPs 1 and 13 during days 4–8 in the conditioned media
were analyzed by immunoblotting. Three separate experiments were
performed, with similar results. See Figure 1 for definitions.
CD44 IN MMP INDUCTION BY FIBRONECTIN FRAGMENT
1277
binding of HBFN-f mediated via the same binding site for
HSPG in HBFN-f (11). In addition, HSPG and CSPG
interact with the same site within the III14 repeat of
HBFN-f, which contains the peptide V sequence (37).
Thus, ligation of peptide V with GAGs on CD44 is likely
the reason why the peptide blocked the access of antibody
OS/37 to CD44, which could result in the inhibition of
HBFN-f–stimulated MMP production in articular cartilage.
DISCUSSION
Figure 7. Suppression of HBFN-f–stimulated production of MMPs 1,
2, 9, and 13 by heparan sulfate proteoglycan (HSPG) in explant
culture. HBFN-f at 100 nM was added from day 0 in the presence or
absence of HSPG at 100 ␮g/ml. Control cultures contained no
additives. Secreted levels of MMPs 1, 2, 9, and 13 during days 0–4 in
conditioned media were analyzed by immunoblotting. Three separate
experiments were performed, with similar results. See Figure 1 for
other definitions.
involvement of CD44 in MMP induction stimulated by
HBFN-f in articular cartilage by use of the antibody. In
contrast to control IgG1, antibody OS/37 blocked
HBFN-f–stimulated production of MMPs 2, 9, and 13 by
99 ⫾ 2%, 99 ⫾ 2%, and 95 ⫾ 5%, respectively (n ⫽ 3)
(Figure 8). In addition, the antibody partially inhibited
HBFN-f–induced MMP-1 by 50 ⫾ 8% (n ⫽ 3) (Figure
8). A significant difference was found between the
secreted levels of the individual MMPs from HBFN-f–
treated cultures in the presence and absence of the
anti-CD44 antibody (P ⬍ 0.05 by t-test). Analysis by
fluorescence microscopy revealed that FITC-conjugated
antibody OS/37 localized CD44 in association with chondrocytes (Figure 9A), indicating that occupancy of CD44
by antibody OS/37 on chondrocytes can block HBFN-f–
induced MMP production.
The inhibitory effect of antibody OS/37 on
HBFN-f–induced MMP production (Figure 8) appeared
to be similar to that of peptide V (Figure 6) and HSPG
(Figure 7). When articular cartilage was incubated with
FITC-conjugated OS/37 following preincubation with
excessive amounts of peptide V, antibody OS/37 failed to
bind CD44 on chondrocytes (Figure 9B). Glycosaminoglycans (GAGs) can attach to the membrane proximal portion of the extracellular domain of CD44 (35). Chondroitin
sulfate proteoglycan (CSPG) has been found on the standard isoform of CD44 (36), and it is required for the
Degradation products of fibronectin are of interest as amplifiers or catalysts in diseased joints, including
those in RA and OA (38), because of their ability to
stimulate MMP induction (17,19,20) and cartilage destruction (17,19). In this study, we demonstrated that
HBFN-f can stimulate type II collagen cleavage by
collagenase in human articular cartilage explant culture
under serum-free conditions, in association with enhanced production of collagenases (MMPs 1 and 13)
and gelatinases (MMPs 2 and 9). These findings extend
the results of our previous study, which showed that
Figure 8. Suppression of HBFN-f–stimulated production of MMPs 1,
2, 9, and 13 by anti-CD44 antibody OS/37 in explant culture. Articular
cartilage was incubated with HBFN-f at 100 nM from day 0, with or
without pretreatment with antibody OS/37 or control IgG1. Secreted
levels of MMPs 1, 2, 9, and 13 during days 0–4 in conditioned media
were analyzed by immunoblotting. Three separate experiments were
performed, with similar results. See Figure 1 for definitions.
1278
Figure 9. Suppression of binding of OS/37 to CD44 on chondrocytes
by peptide V. A, Following blocking with bovine serum albumin (BSA),
articular cartilage was incubated with fluorescein isothiocyanate
(FITC)–conjugated antibody OS/37 or subclass-matched FITCconjugated mouse IgG1. B, After blocking with BSA, articular cartilage was preincubated with or without peptide V and then incubated
with FITC-conjugated antibody OS/37. Bar ⫽ 50 ␮m; in A and 200 ␮m
in B.
HBFN-f can induce type II collagen degradation caused
mainly by MMP-13 in bovine articular cartilage under
serum-free conditions (19).
The observation that injection of fibronectin fragments into rabbit knee joints induces depletion of cartilage proteoglycan is consistent with the pathophysiologic
significance of the fragments (39). Elevated levels of
these fragments are present in OA cartilage (15) and in
the synovial fluid of OA and RA patients (15,16). In OA
synovial fluids, ⬃1 ␮M of 100–200-kd fibronectin fragments have been found (16). Since the levels of fibronectin fragments in OA cartilage have been suggested to be
similar to those in OA synovial fluids (15), the contents of
fibronectin fragments may reach 100 nM in OA cartilage,
YASUDA ET AL
comparable to the concentration used in the present study.
Thus, fibronectin fragments could play an important role in
cartilage destruction in arthritis. Although fibronectin isoforms containing III12–14 and IIICS domains are present
in human femoral head cartilage (40), the presence of
HBFN-f in diseased articular cartilage or synovial fluid in
the current study remains unclear.
The CD44 gene has 19 exons, 12 of which may be
alternatively spliced to produce a number of different
isoforms (41). Restricted expression of CD44 isoforms
and posttranslational glycosylation of the parent protein
provide diverse functions of CD44. Of the CD44 isoforms, CD44H is commonly expressed in human articular chondrocytes (42). Although CD44H is predominant,
messenger RNA containing the V3 exon of CD44 is also
found in chondrocytes (42). The diversity of CD44 is
further amplified by the differential use of GAG attachment sites on its extracellular domain. While CSPG is
attached to the membrane proximal portion of the
external domain of CD44H (35), HSPG can bind CD44
at V3 in the membrane proximal extracellular domain of
CD44v (43). CSPG and HSPG use identical or overlapping binding sites in the repeats III13 and III14 of
HBFN-f (11,37). In our study, the observed suppression
of HBFN-f–stimulated MMP production by peptide V
suggests that the peptide V domain, a binding site of
HBFN-f for cell surface HSPG, is required for HBFN-f–
activated MMP induction in human articular cartilage.
The results from our inhibition studies using HSPG are
consistent with this idea. Thus, the results using the antiCD44 antibody indicate that HBFN-f may directly bind
GAGs on CD44 through the peptide V sequence. The
characterization of GAGs on CD44 on chondrocytes that
can interact with peptide V is under investigation.
The pathologic roles of CD44 in arthritis are not
fully understood. CD44 is expressed by different cells in
the RA synovium (44). Anti-CD44 treatment in other
studies has been shown to result in a reduction of tissue
swelling and leukocyte infiltration in a murine arthritis
model (45), as well as in the inhibition of cartilage
invasion by RA synovial fibroblasts (46). Immunohistochemical studies have demonstrated up-regulation of
CD44 on chondrocytes in articular cartilage from patients with OA (47) and RA (48). Recent studies suggest
that up-regulated CD44 on chondrocytes plays a role in
increased internalization and degradation of hyaluronan
(49). However, this is the first evidence that MMP
induction by fragments of fibronectin may be mediated
by CD44. In addition to IL-1␣, fibronectin fragments have
been shown to enhance CD44 expression in bovine articular chondrocytes (50). Thus, increased generation of
CD44 IN MMP INDUCTION BY FIBRONECTIN FRAGMENT
1279
HBFN-f in OA and RA may up-regulate CD44 on OA and
RA chondrocytes, further supporting activation of MMP
production by HBFN-f. Based on the results of the present
study, CD44-directed therapy may therefore help prevent
cartilage destruction by HBFN-f in OA and RA. Whether
other fragments of fibronectin can stimulate MMP production through a similar mechanism remains to be clarified.
The present study cannot exclude the possibility
that the action of HBFN-f on chondrocytes may involve
other receptors besides CD44. While fibronectin can
bind several integrins and other cell surface protein
ligands (51), fragments of fibronectin could interact with
more than one receptor on chondrocytes. Because receptors such as integrins can work together cooperatively, blocking any one of the receptors may inactivate
fibronectin fragments.
Although the results of the present study indicate
the minor role of the ␣4␤1 integrin in HBFN-f–induced
MMP, CS-1 did stimulate MMP-1 production when used
with HBFN-f (Figure 4). The interaction between CS-1
and chondrocytes may alter signaling events by blocking
the binding of the natural ligand to its receptor, because
one of the proposed mechanisms for the action of
fibronectin fragments is that their binding to a fibronectin receptor or to other matrix components near the
receptor perturbs the binding of native fibronectin to
␣5␤1 integrin (21). When rabbit synovial fibroblasts bind
to the central RGD-containing fragments of fibronectin
via ␣5␤1 integrin, expression of MMPs 1 and 3 is
induced (20). In addition, antisense oligonucleotides to
␣5 integrin inhibit chondrolysis induced by 29-kd NH2terminal heparin-binding and 50-kd NH2-terminal
gelatin-binding fragments of fibronectin (21). Thus, both
cell-binding and non–cell-binding fragments of fibronectin could individually operate through ␣5␤1 integrin.
Integrin ␣5␤1 is the primary receptor involved in the
assembly of dimeric fibronectin into the extracellular
matrix (52). The I1–5 repeats of NH2-terminal heparinbinding fragment of fibronectin block the assembly of
fibronectin into fibrils, and fibronectin dimers lacking
these domains fail to be incorporated into fibrils (53–
56). Since the III12–14 repeats of HBFN-f have also
been shown to contribute to fibronectin fibrillogenesis
(57), HBFN-f may interfere with fibronectin assembly
and indirectly alter ␣5␤1 signaling.
Vol. 1. 14th ed. Baltimore: Lippincott Williams & Wilkins; 2001. p.
226–84.
Arend WP, Dayer J-M. Inhibition of the production and effects of
interleukin-1 and tumor necrosis factor ␣ in rheumatoid arthritis.
Arthritis Rheum 1995;38:151–60.
Burton-Wurster N, Lust G. Fibronectin in cartilage. In: Carson S,
editor. Fibronectin in health and disease. Boca Raton (FL): CRC
Press; 1989. p. 243–55.
Yamada KM. Adhesive recognition sequences. J Biol Chem
1991;266:12809–12.
McCarthy JB, Chelberg MK, Mickelson DJ, Furcht LT. Localization and chemical synthesis of fibronectin peptides with melanoma
adhesion and heparin binding activities. Biochemistry 1988;27:
1380–8.
Drake SL, Klein DJ, Mickelson DJ, Oegema TR, Furcht LT,
McCarthy JB. Cell surface phosphatidylinositol-anchored heparan
sulfate proteoglycan initiates mouse melanoma cell adhesion to a
fibronectin-derived, heparin-binding synthetic peptide. J Cell Biol
1992;117:1331–41.
Iida J, Skubitz AP, Furcht LT, Wayner EA, McCarthy JB.
Coordinate role for cell surface chondroitin sulfate proteoglycan
and ␣4␤1 integrin in mediating melanoma cell adhesion to fibronectin. J Cell Biol 1992;118:431–44.
Wayner EA, Garcia-Pardo A, Humphries MJ, McDonald JA,
Carter WG. Identification and characterization of the T lymphocyte adhesion receptor for an alternative cell attachment domain
(CS-1) in plasma fibronectin. J Cell Biol 1989;109:1321–30.
Mould AP, Humphries MJ. Identification of a novel recognition
sequence for the ␣4␤1 in the COOH-terminal heparin-binding
domain of fibronectin. EMBO J 1991;10:4089–95.
Mould AP, Komoriya A, Yamada KM, Humphries MJ. The CS5
peptide is a second site in the IIICS region of fibronectin
recognized by the integrin ␣4␤1: inhibition of ␣4␤1 function by
RGD peptide homologues. J Biol Chem 1991;266:3579–85.
Jalkanen S, Jalkanen M. Lymphocyte CD44 binds the COOHterminal heparin-binding domain of fibronectin. J Cell Biol 1992;
116:817–25.
Aruffo A, Stamenkovic I, Melnick M, Underhill CB, Seed B. CD44
is the principal cell surface receptor for hyaluronate. Cell 1990;61:
1303–13.
Miller DR, Mankin HJ, Shoji H, D’Ambrosia RD. Identification
of fibronectin in preparations of osteoarthritic human cartilage.
Connect Tissue Res 1984;12:267–75.
Jones KL, Brown M, Ali SY, Brown RA. An immunohistochemical study of fibronectin in human osteoarthritic and disease free
articular cartilage. Ann Rheum Dis 1987;46:809–15.
Homandberg GA, Wen C, Hui F. Cartilage damaging activities of
fibronectin fragments derived from cartilage and synovial fluid.
Osteoarthritis Cartilage 1998;6:231–44.
Xie DL, Meyers R, Homandberg GA. Fibronectin fragments in
osteoarthritic synovial fluid. J Rheumatol 1992;19:1448–52.
Homandberg GA, Meyers R, Xie DL. Fibronectin fragments cause
chondrolysis of bovine articular cartilage slices in culture. J Biol
Chem 1992;267:3597–604.
Homandberg GA, Hui F, Wen C, Purple C, Bewsey K, Koepp H,
et al. Fibronectin-fragment-induced cartilage chondrolysis is associated with release of catabolic cytokines. Biochem J 1997;321:
751–7.
Yasuda T, Poole AR. A fibronectin fragment induces type II
collagen degradation by collagenase through an interleukin1–mediated pathway. Arthritis Rheum 2002;46:138–48.
Werb Z, Tremble PM, Behrendtsen O, Crowley E, Damsky CH.
Signal transduction through the fibronectin receptor induces collagenase and stromelysin gene expression. J Cell Biol 1989;109:
877–89.
Homandberg GA, Costa V, Ummadi V, Pichika R. Antisense
oligonucleotides to the integrin receptor subunit ␣5 decrease
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
REFERENCES
1. Poole AR. Cartilage in health and disease. In: Koopman WJ,
editor. Arthritis and allied conditions: a textbook of rheumatology.
21.
1280
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
fibronectin fragment mediated cartilage chondrolysis. Osteoarthritis Cartilage 2002;10:381–93.
Huebsch JC, McCarthy JB, Diglio CA, Mooradian DL. Endothelial cell interactions with synthetic peptides from the carboxylterminal heparin-binding domains of fibronectin. Circ Res 1995;
77:43–53.
Chon JH, Chaikof EL. Soluble heparin-binding peptides regulate
chemokinesis and cell adhesive forces. Am J Physiol Cell Physiol
2001;280:C1394–402.
Xie DL, Hui F, Homandberg GA. Fibronectin fragments alter
matrix protein synthesis in cartilage tissue cultured in vitro. Arch
Biochem Biophys 1993;307:110–8.
Morales TI, Wahl LM, Hascall VC. The effect of bacterial
lipopolysaccharides on the biosynthesis and release of proteoglycans from calf articular cartilage cultures. J Biol Chem 1984;259:
6720–9.
Billinghurst RC, Dahlberg L, Ionescu M, Reiner A, Bourne R,
Rorabeck C, et al. Enhanced cleavage of type II collagen by
collagenases in osteoarthritic articular cartilage. J Clin Invest
1997;99:1534–45.
Dahlberg L, Billinghurst RC, Manner P, Nelson F, Webb G,
Ionescu M, et al. Selective enhancement of collagenase-mediated
cleavage of resident type II collagen in cultured osteoarthritic
cartilage and arrest with a synthetic inhibitor that spares collagenase 1 (matrix metalloproteinase 1). Arthritis Rheum 2000;43:
673–82.
Billinghurst RC, Wu W, Ionescu M, Reiner A, Dahlberg L, Chen
J, et al. Comparison of the degradation of type II collagen and
proteoglycan in nasal and articular cartilages induced by interleukin-1 and the selective inhibition of type II collagen cleavage by
collagenase. Arthritis Rheum 2000;43:664–72.
Labarca C, Paigen K. A single, rapid, and sensitive DNA assay
procedure. Anal Biochem 1980;102:344–52.
Ostergaard K, Salter DM, Petersen J, Bendtzen K, Hvolris J,
Andersen CB. Expression of ␣ and ␤ subunits of the integrin
superfamily in articular cartilage from macroscopically normal and
osteoarthritic human femoral heads. Ann Rheum Dis 1998;57:
303–8.
Woods A, McCarthy JB, Furcht LT, Couchman JR. A synthetic
peptide from the COOH-terminal heparin-binding domain of
fibronectin promotes focal adhesion formation. Mol Biol Cell
1993;4:605–13.
Lories V, Cassiman JJ, Van den Berghe H, David G. Differential
expression of cell surface heparan sulfate proteoglycans in human
mammary epithelial cell and fibroblasts. J Biol Chem 1992;267:
1116–22.
Giuseppetti JM, McCarthy JB, Letourneau PC. Isolation and
partial characterization of a cell-surface heparan sulfate proteoglycan from embryonic rat spinal cord. J Neurosci Res 1994;37:
584–95.
Murakami S, Shimabukuro Y, Miki Y, Saho T, Hino E, Kasai D,
et al. Inducible binding of human lymphocytes to hyaluronate via
CD44 does not require cytoskeleton association but does require
new protein synthesis. J Immunol 1994;152:467–77.
Lesley J, Hyman R. CD44 structure and function. Front Biosci
1998;3:D616–30.
Henke CA, Roongta U, Mickelson DJ, Knutson JR, McCarthy JB.
CD44-related chondroitin sulfate proteoglycan, a cell surface
receptor implicated with tumor cell invasion, mediates endothelial
cell migration on fibrinogen and invasion into a fibrin matrix.
J Clin Invest 1996;97:2541–52.
Barkalow FJ, Schwarzbauer JE. Interactions between fibronectin
and chondroitin sulfate are modulated by molecular context. J Biol
Chem 1994;269:3957–62.
Homandberg GA. Potential regulation of cartilage metabolism in
osteoarthritis by fibronectin fragments. Front Biosci 1999;4:
D713–30.
YASUDA ET AL
39. Homandberg GA, Meyers R, Williams JM. Intraarticular injection
of fibronectin fragments causes severe depletion of cartilage
proteoglycans in vivo. J Rheumatol 1993;20:1378–82.
40. Parker AE, Boutell J, Carr A, Maciewicz RA. Novel cartilagespecific splice variants of fibronectin. Osteoarthritis Cartilage
2002;10:528–34.
41. Screaton GR, Bell MV, Jackson DG, Cornelis FB, Gerth U, Bell
JI. Genomic structure of DNA encoding the lymphocyte homing
receptor CD44 reveals at least 12 alternatively spliced exons. Proc
Natl Acad Sci U S A 1992;89:12160–4.
42. Salter DM, Godolphin JL, Gourlay MS, Lawson MF, Hughs DE,
Dunne E. Analysis of human articular chondrocyte CD44 isoform
expression and function in health and disease. J Pathol 1996;179:
396–402.
43. Bennett KL, Jackson DG, Simon JC, Tanczos E, Peach R, Modrell
B, et al. CD44 isoforms containing exon V3 are responsible for the
presentation of heparin-binding growth factor. J Cell Biol 1995;
128:687–98.
44. Henderson KJ, Edwards JC, Worrall JG. Expression of CD44 in
normal and rheumatoid synovium and cultured synovial fibroblasts. Ann Rheum Dis 1994;53:729–34.
45. Mikecz K, Brennan FR, Kim JH, Glant TT. Anti-CD44 treatment
abrogates tissue oedema and leukocyte infiltration in murine
arthritis. Nat Med 1995;1:558–63.
46. Neidhart M, Gay RE, Gay S. Anti–interleukin-1 and antiCD44 interventions producing significant inhibition of cartilage
destruction in an in vitro model of cartilage invasion by rheumatoid arthritis synovial fibroblasts. Arthritis Rheum 2000;43:
1719–28.
47. Ostergaard K, Salter DM, Andersen CB, Petersen J, Bendtzen K.
CD44 expression is up-regulated in the deep zone of osteoarthritic
cartilage from human femoral heads. Histopathology 1997;31:
451–9.
48. Takagi T, Okamoto R, Suzuki K, Hayashi T, Sato M, Sato M, et al.
Up-regulation of CD44 in rheumatoid chondrocytes. Scand
J Rheumatol 2001;30:110–3.
49. Aguiar DJ, Knudson W, Knudson CB. Internalization of the
hyaluronan receptor CD44 by chondrocytes. Exp Cell Res 1999;
252:292–302.
50. Chow G, Knudson CB, Homandberg G, Knudson W. Increased
expression of CD44 in bovine articular chondrocytes by catabolic
cellular mediators. J Biol Chem 1995;270:27734–41.
51. Johansson S, Svineng G, Wennerberg K, Armulik A, Lohikangas
L. Fibronectin-integrin interactions. Front Biosci 1997;2:
D126–46.
52. Dzamba BJ, Bultmann H, Akiyama SK, Peters DM. Substratespecific binding of the amino terminus of fibronectin to an
integrin complex in focal adhesions. J Biol Chem 1994;269:
19646–52.
53. McKeown-Longo PJ, Mosher DF. Interaction of the 70,000mol-wt amino-terminal fragment of fibronectin with the
matrix-assembly receptor of fibroblasts. J Cell Biol 1985;100:
364–74.
54. Quade BJ, McDonald JA. Fibronectin’s amino-terminal matrix
assembly site is located within the 29-kDa amino-terminal
domain containing five type I repeats. J Biol Chem 1988;263:
19602–9.
55. Schwarzbauer JE. Identification of the fibronectin sequences
required for assembly of a fibrillar matrix. J Cell Biol 1991;113:
1463–73.
56. Sottile J, Schwarzbauer J, Selegue J, Mosher DF. Five type I
modules of fibronectin form a functional unit that binds to
fibroblasts and Staphylococcus aureus. J Biol Chem 1991;266:
12840–3.
57. Bultmann H, Santas AJ, Peters DM. Fibronectin fibrillogenesis
involves the heparin II binding domain of fibronectin. J Biol Chem
1998;273:2601–9.
Документ
Категория
Без категории
Просмотров
1
Размер файла
423 Кб
Теги
matrix, induction, involvement, heparin, terminal, fragmenty, human, fibronectin, cooh, cd44, culture, cartilage, binding, articular, metalloproteinase
1/--страниц
Пожаловаться на содержимое документа