вход по аккаунту


Human Osteoarthritic Chondrocytes Possess an Increased Number of Insulin-Like Growth Factor 1 Binding Sites but are Unresponsive to its Stimulation.

код для вставкиСкачать
Number 2, February 1994, pp 253-263
0 1994, American College of Rheumatology
Possible Role of IGF-1-Binding Proteins
Objective. To characterize the insulin-like growth
factor 1 (IGF-1) receptor in human osteoarthritic (OA)
and normal adult chondrocytes. The biologic response
of chondrocytes to IGF-1 stimulation was examined, as
was the presence and synthesis of IGF binding proteins
(IGFBP) in these cells.
Methods. Binding studies, Northern blot, immunohistochemical analysis, and affinity cross-linking experiments were performed for characterization of the
IGF receptor, and the latter method was also used for
IGFBP determination. The biologic response was estimated via the incorporation of radiolabeled proline into
a newly synthesized protein.
Results. Binding experiments revealed a single
class of binding sites. The mean +: SEM affinity (Kd) of
normal chondrocytes was 1.4 2 0.4 nM, with 26.8 f 5.5
X lo3 binding siteslcell. OA chondrocytes had a lower
affinity (Kd 15.4 f 4.7 nM) and a higher density (1,178.3
299.5 X lo3 binding sitedcell) compared with normal
cells (P< 0.004 and P < 0.001, respectively). Immuno-
From the Department of Medicine, University of Montreal,
Notre-Dame Hospital Research Center, Montreal, Quebec, Canada.
Sylvain Dor6 was the recipient of a Studentship Grant from the
Arthritis Society.
Sylvain Dor6, PhD: Notre-Dame Hospital Research Center; Jean-Pierre Pelletier, MD: Notre-Dame Hospital Research
Center; John A. DiBattista, PhD: Notre-Dame Hospital Research
Center; Ginette Tardif, PhD: Notre-Dame Hospital Research Center; Paul Brazeau, PhD: Notre-Dame Hospital Research Center;
Johanne Martel-Pelletier, PhD: Notre-Dame Hospital Research
Address reprint requests to Johanne Martel-Pelletier, PhD,
Unit6 Rhumatismale, HBpital Notre-Dame, 1560 est, rue Sherbrooke, Montreal, QuCbec, Canada H2L 4K8.
Submitted for publication February 19, 1993; accepted in
revised form July 6, 1993.
histochemical studies with a monoclonal antibody
(MAb) against the type 1 IGF receptor (aIR3) showed
increased staining in OA cartilage compared with normal tissue. Biologic responses of chondrocytes after
IGF-1 stimulation revealed that OA chondrocytes were
unresponsive, whereas a 2.5-fold increase in new protein
synthesis was observed in normal cells. Competition
studies in normal chondrocytes revealed that both IGF-1
and IGF-2 displaced radiolabeled IGF-1 in a comparable manner; however, insulin at high concentration
weakly competes. Moreover, MAb aIR3 effectively
blocked specific binding in normal chondrocytes (77%),
but not in OA chondrocytes (26%). Northern blot and
covalent cross-linking analyses revealed the specific
band characteristic of type 1 receptor. With the latter
technique, other bands corresponding to the IGFBPs
were also detected. Comparison between normal and
OA chondrocytes showed increased intensity of the
IGFBP bands, particularly those corresponding to the
IGFBP-3 doublet.
Conclusion. It is shown that type 1 IGF receptor
is expressed in human articular cartilage and that the
level of binding sites is significantly increased in OA
chondrocytes. Interestingly, despite the higher level of
binding sites in OA cells, no response to IGF-1 stimulation was found in these cells. Our data suggest that this
increase in specific binding may involve not only the type
1IGF receptor but also IGFBP on the cell surface. The
latter, by binding the IGF-1, will diminish the bioavailability of IGF-1 and thus prevent its anabolic action.
Osteoarthritis (OA) is the most common of the
various arthritic disorders affecting humans. The dis-
ease appears to result from a disruption in the equilibrium between the synthesis and the degradation of
matrix macromolecules in which the degradative process eventually exceeds the reparative one, leading to
a total loss of cartilage and eburnation of bone. Efforts
have been devoted to the search for conditions that
would favor the formation of a durable, functional
articular surface following cartilage damage. This has
led to the study of factors that can stimulate cartilage
repair. Insulin-like growth factor (IGF) appears to be
one of the most important growth factors affecting the
anabolism of the major molecules found in cartilage
(1). Although IGF-1 has been shown to be moderately
mitogenic by itself in human adult articular cartilage, it
strongly stimulated production of extracellular matrix
components by chondrocytes. More particularly, this
growth factor enhances the synthesis of collagen and
proteoglycan in normal cartilage, as shown by both in
vitro and in vivo studies (2-5). Moreover, IGF-1
appears to be a differentiating factor in that it enhances
the expression of type I1 collagen messenger RNA and
prevents the synthesis of type I collagen (6).
This peptide initiates its biologic response by
interacting with specific IGF membrane receptors.
The type 1 IGF receptor is a glycoprotein of -300 kd,
consisting of 2 a-(-130 kd) and 2 psubunits (-90 kd)
joined by disulfide bridges linked in an a2@ form. The
2 a-subunits are entirely extracellular and appear to be
involved in ligand binding. The psubunits traverse the
cell membrane and contain a tyrosine kinase domain in
their cytoplasmic protein. While the type 1 IGF receptor has been identified on growth plate chondrocytes in
humans and animals (6-8), no study has addressed its
presence in adult human articular cartilage. We conducted basic characterization studies to determine if
the receptor was expressed in adult human chondrocytes. In this study, we investigated the specificity of
the IGF receptor, its affinity, as well as the number of
binding sites in normal and OA chondrocytes. The
biologic response to recombinant human IGF-1
(rHuIGF-I) and the presence of IGF binding proteins
(IGFBP) in normal and OA chondrocytes were also
Specimens. Articular cartilage specimens (tibial plateaus and femoral condyles) were obtained from OA patients
(mean L SEM age 66 2 years) undergoing knee arthroplasty. The diagnosis of OA was based on clinical and
radiologic evaluations (9). These specimens represented
moderate to severe OA as defined according to macroscopic
and histologic criteria. For tibial plateau, cartilage was
obtained from the lateral compartment, in which tissue
damage was more limited than on the medial compartment.
Typical OA cartilage surface fibrillation and pitting were
prominent, along with eburnation of variable size. The
specimens were full-thickness strips of tissue cut across the
surface, excluding mesenchymal repair tissue, subchondral
bone, and osteophyte. Control cartilage samples were obtained, within 12 hours of death, from adult subjects (mean
SEM age 53 7 years) with no history ofjoint disease and
whose knee cartilage was macroscopically (10) and microscopically (1 1) normal.
Isolation and culture of chondrocytes. The specimens
were obtained under aseptic conditions and cartilage was
dissected on ice. Chondrocytes were released from articular
cartilage by sequential enzymatic digestion at 37°C as previously described (12): 2.0 mg/ml pronase followed by 1 mglml
collagenase (Sigma, St. Louis, MO). The isolated chondrocytes were seeded at high density and cultured in Dulbecco's
modified Eagle's medium supplemented with 10% heatinactivated fetal calf serum (FCS) and antibiotic mixture
(Gibco, Grand Island, NY) at 37°C in a humidified atmosphere of 5% CO,, 95% air. Unless otherwise stated, only
primary cells were used, in order to ensure the chondrocyte
phenotype. When first-passage cells were utilized, primary
cells were cultured for 3 days, detached using EDTA (0.53
mM), and seeded at high density. Experiments were performed at a final concentration of approximately 5 x lo5
cells/well. After plating, the cells reached confluence in 5-6
days. The medium was changed every second day, and 18
hours before the experiment, the cells were incubated in
fresh medium without FCS.
Binding experiments. Before the experiments, the
confluent chondrocytes were washed twice with the binding
buffer (phosphate buffered saline, 1% bovine serum albumin
[PBS-BSA], pH 8.0) (13) and preincubated for 1 hour with
the same buffer. Cells were then incubated with '"IrHuIGF- 1, and nonspecific binding was determined by adding a 100-fold molar excess of unlabeled rHuIGF-1 (generously provided by Genentech Canada, Burlington, Ontario,
Canada). At the end of the incubation period, the cells were
washed in cold PBS-BSA, solubilized, and the radioactivity
In order to determine the optimal conditions required
to reach steady-state binding, preliminary experiments were
carried out using various temperatures and periods of time.
Specific binding was calculated by subtracting counts bound
in the presence of excess unlabeled ligand from the total
counts bound in the absence of excess unlabeled ligand. The
number of cells was calculated using parallel monolayer
cultures in which chondrocytes were detached. Recombinant human IGF-1 was monoiodinated ('251, 1 mCi/ml;
Amersham, Oakville, Ontario, Canada) using chloramine T
(Sigma) (14). The specific radioactivity of '251-rHuIGF-l
preparations was -600 Ci/mmole.
For competitive binding experiments, the chondrocytes were incubated with '251-rHuIGF-l (0.1 nM) and
increasing concentrations of rHuIGF-1 (0-100 ng/ml),
rHuIGF-2 (&lo0 ng/ml; Bachem, Torrence, CA), and recombinant human insulin (0-10,OOO ng/ml; Peninsula Laboratories, Belmont, CA) for 4 hours at 22°C.
In order to verify IGF receptor type, we used a
specific monoclonal antibody (MAb) against the type 1 IGF
receptor, aIR3 (ascites fluid, generously provided by Dr. S.
Jacobs, Burroughs Wellcome Diagnostic, Research Triangle
Park, NC) (15). Chondrocytes were incubated for 1 hour at
22°C with different dilutions of the antibody (1:8,000 to
1:100) prior to the binding assay. Control experiments were
carried out using nonimmune mouse IgGl (PharMingen, San
Diego, CA) in same concentration, instead of aIR3.
For internalization experiments, cells were washed
as above and incubated in the binding buffer at 4°C for 4
hours with a saturating amount of '251-rHuIGF-1 (1 nM).
They were further incubated at 37°C for various periods of
time (0, 15, 30, 45, 60, 120, or 240 minutes), then quickly
washed once with a stripping buffer consisting of 200 mM
acetic acid and 500 mM NaCl (pH 2.5) followed by PBS,
solubilized, and counted.
AEnity cross-linking and Western ligand blots. M n itycross-linking was done according to the procedure of
Rosenfeld et a1 (14). Briefly, chondrocytes were incubated at
22°C for 4 hours with 0.1 nM '251-rHuIGF-1 in the binding
buffer in the presence or absence of a 100-fold molar excess
of unlabeled rHuIGF-1. Cells were washed once in PBS and
incubated with 50 mM disuccinimidyl suberate (DSS; Pierce
Chemical, Rockford, IL) for 30 minutes at 4°C. After 2
washes, the cells were solubilized directly into boiling sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) buffer (10 mM Tris, 1 mM EDTA, 12.5%
glycerol, 2.5% SDS, 2.5% bromphenol blue, pH 8.0) in the
presence or absence of the reducing agent, dithiothreitol
(Sigma), at 100 mM.
For Western ligand blots (16), culture medium was
dialyzed in cold water, concentrated 30-fold by lyophilization, and the samples subjected to SDS-PAGE under nonreducing conditions. The gel was then electroblotted onto a
nitrocellulose membrane, immersed in TBS (10 mM Tris, 150
mM NaC1,0.05% NaN,, pH 7.4) containing 3% Nonidet P40
(Sigma), incubated overnight at 4°C in 5% skim milk in TBS,
washed with TTBS (TBS plus 0.1% Tween 20; Sigma),
reincubated overnight at 4°C with 1251-IGF-l (lo6 counts per
minute) in TTBS containing 1% BSA, and washed twice in
TTBS and once in TBS.
For both affinity cross-linking and Western ligand
blots, the membranes were further subjected to autoradiography. Samples and prestained high and low molecular
weight standards (Bio-Rad, Richmond, CA) were subjected
Immunohistochemistry studies. Human articular cartilage was processed for immunohistochemical analysis as
previously described (17). The endogenous peroxidase was
quenched with 0.3% H,02 in methanol for 20 minutes at
room temperature and then washed in PBS-BSA. Slides
were further overlaid with 2 pg (100 d m l ) of the anti-type 1
IGF receptor MAb aIR3 (#GR11; Oncogene Sciences,
Manhasset, NY) in PBS-BSA. Control slides were made by
substitution of the primary antibody with the nonimmune
mouse IgGl (PharMingen). After overnight incubation, the
antibody-antigen complex was revealed using a Vectastain
ABC kit (Dimension Laboratories, Mississauga, Ontario,
Canada). Tissues were counterstained with Gill-I hematoxylin (Shandon, Pittsburgh, PA).
Northern blotting and hybridization analysis. RNA
was extracted using the method described by Aiba et al(18),
with modifications. Briefly, freshly released chondrocytes
were lysed in sodium acetate buffer (20 mM sodium acetate,
0.5% SDS, 1 mM EDTA, pH 5.0) and RNA was extracted with
a hot (60°C) solution of phenol saturated with the sodium
acetate buffer at pH 5.0. After being reextracted and precipitated with ethanol, the RNA was redissolved in D E E water
and quantified by spectrophotometry. Poly(A)+ RNA was
obtained following oligo(dT)-cellulosechromatography (Pharmacia LKB Biotechnology, Piscataway, NJ) and was fractionated through a 1.2% agaroseformaldehyde gel and then
electroblotted (Transblot; Bio-Rad) onto Zetaprobe nylon
membrane (Bio-Rad) using Tris-acetateEDTA buffer.
The antisense and sense (32P)-radiolabeled RNA
probes (cRNA) (a3'P-CTP; NEN, Madison, WI) used were
synthesized from the plasmid IGF-1-R.8, containing the
700-basepair human type 1 IGF receptor complementary
DNA subcloned into the Eco RI site of the vector pUC13
(Genentech, South San Francisco, CA) (19) and cloned in
the Eco RI site of the PGEM-4Z vector. The labeling
procedure was carried out using the transcription in vitro
system kit from Promega Biotech (Madison, WI).
Chondrocyte stimulation with rHuIGF-1. In order to
evaluate the putative anabolic effects of rHuIGF-1 on chondrocytes, the incorporation of radiolabeled proline into a
newly synthesized protein was determined (20) on normal
and OA chondrocytes. Prior to the experiments, the chondrocytes were incubated in serum-free medium for 18 hours.
Cells were then incubated (2 days, 37°C) in the presence or
absence (control) of rHuIGF-1 (100 nglml), 3.5 pg/ml cycloheximide, or 10% FCS; each assay was performed in triplicate. Twenty-four hours before the end of the incubation
period, 2 pCi of L-(2,3-'H)-proline (28 Ci/mmole, Amersham) was added to the culture medium. At the conclusion
of the incubation, the medium was removed and the cells
were washed twice with 0.1% nonradioactive proline in PBS,
and solubilized in Tris buffer (50 mM Tris, 150mM NaCl, pH
8.0) containing 1% Nonidet P40. Proteins were then precipitated in 10% (final concentration) ice-cold trichloroacetic
acid, centrifuged, and reprecipitated, the pellet was resuspended in 2.5N NaOH, and the radioactivity counted. The
results were expressed as the percentage of proline incorporation over control.
Statistical analysis. Binding data were analyzed using
the nonlinear regression program LIGAND (21) as adapted
by G. A. McPherson (Biosoft, Milltown, NJ). Data are
expressed as the mean f SEM. Statistical significance was
assessed by Student's 2-tailed t-test, when applicable. Significant differences were confirmed only when the P value
was less than 5%.
Binding. Preliminary binding experiments (n =
4) assessing various temperatures and time periods
revealed that the specific binding of '251-rHuIGF-1 to
confluent cultures of normal chondrocytes reached a
plateau after 2 hours at 22°C and remained stable for at
least 4 hours. Data from studies done at 4°C and 13°C
were similar to one another and showed that the
binding had not completely reached a plateau after 4
hours of incubation. At 37"C, the maximum binding
value was reached at 30 minutes but subsequently
declined. Therefore, all subsequent binding experiments were performed at 22°C with a 4-hour incuba-
Bound (nM)
non specific
Figure 1. Representative curves of the saturation binding of 'Z51-labeledinsulin-like growth factor 1
(IGF-I) to normal (A) and osteoarthritic (B)adult human chondrocytes. Cells were incubated with
increasing concentrations of labeled ligand (total) or in the presence of a 100-fold molar excess of
unlabeled ligand (non specific). Specific binding was determined by subtracting counts bound in the
presence of excess unlabeled IGF-I from counts bound in the presence of labeled IGF-I alone.
Scatchard plots (insets) were analyzed by linear regression analysis.
tion period. Generally, nonspecific binding was less
than 5% of the total bound radioactivity. Further
experiments (n = 3) revealed that maximal internaliza-
tion was reached after a I-hour incubation period and
remained stable for at least 4 hours.
Competitive binding experiments (n = 2) with
unlabeled rHuIGF-1, rHuIGF-2, and recombinant human insulin using normal chondrocytes demonstrated
that rHuIGF-1 had the highest affinity. At a concentration of 100 ng/ml, unlabeled rHuIGF- 1 displaced the
radiolabeled rHuIGF-1 by -88%. At the same concentration, rHuIGF-2 was almost as efficient as rHuIGF-1
and showed a displacement of 82%. Insulin did not
avidly compete with IGF-1, and a displacement of only
19% was recorded at a very high concentration (10,000
ng/ml). Of note, similar experiments with human skin
fibroblasts showed that insulin at this concentration
displaced IGF-1 by 88%.
Additional experiments were conducted to determine the specificity of the binding toward the type 1
IGF receptor, using MAb d R 3 . Chondrocytes were
preincubated for 1 hour with aIR3 in dilutions ranging
from 1:lOO to 1:8,000, and binding assays were performed with '251-rHuIGF-I as described. With normal
chondrocytes 69% of the binding was blocked at a
dilution of 1:8,000, and a plateau of 77% was obtained
at a 1: 1,000 dilution. Interestingly, with OA chondrocytes, 26% inhibition was found at the highest dilution
and 41% at a dilution of 1:100. Control experiments
performed with the isotype-matched nonimrnune antibody (IgG1) showed no effect on binding.
Saturation binding analyses of normal cells
(Figure 1A) showed binding to be dose-dependent and
saturable. Computer-assisted Scatchard analysis
yielded a linear plot indicating a single binding site.
OA chondrocytes (Figure 1B) showed similar saturability of specific binding, and Scatchard analysis also
revealed a single-site model. Analysis of the binding
experiments revealed a mean ? SEM affinity (Kd) of
1.4 f 0.4 nM, with 26.8 2 5.5 x lo3 binding siteskell
for normal cells (n = 9). OA chondrocytes (n = 7)
demonstrated a weaker affinity, with a Kd of 15.4 +- 4.7
nM (P < 0.004 versus normal chondrocytes) and a
major increase in the binding of radiolabeled
rHuIGF-1, with 1,178.3 f 299.5 X lo' binding sites/
cell (P< 0.001 versus normal chondrocytes). Interestingly, while normal cells demonstrated no significant
change in affinity and density after the first-passage
cells, OA cells (n = 5) showed a statistically significant
decrease in the number of binding siteskell (305.7 2
97.0 X lo3; P < 0.001 compared with normal cells and
P < 0.05 compared with OA primary cells) and a
slightly increased affinity (11.2 f 5.5 nM).
In addition, the effect of age on type 1 IGF
receptor affinity and density in chondrocytes from
human cartilage was examined. The results from these
experiments revealed no age-related change in these
elements for macroscopically normal chondrocytes
(age 20-30 years: Kd = 1.9 a,23.0 x lo3 binding
Figure 2. Affinity cross-linking of radiolabeled insulin-like growth
factor 1 on normal human chondrocytes. Cells were incubated with
0.1 nM radioligand (Total) or in the presence of a 100-fold molar
excess of unlabeled ligand (Non-specific), followed by cross-linking
with 50 mM disuccinimidyl suberate. Samples were analyzed by
sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 6%
gels. Arrows indicate bands at 260 kd and 130 kd. Molecular weight
standards are shown on the right.
siteskell [n = 21; age 40-59 years: Kd = 2.1 nM, 29.0
x lo3 binding sites/ceil [n = 21; age 60-79 years: & =
0.8 nM, 19.9 x lo3 binding siteskell [n = 51).
Figure 3. Representative sections of normal (top) and osteoarthritic
(OA) (bottom) adult human cartilage. Immunolocalization of the
insulin-like growth factor 1 (IGF-1) receptor was performed by
direct immunoperoxidase staining using the anti-type IGF-1 receptor monoclonalantibody, aIR3. Sections of normal cartilage showed
specific staining in only a few chondrocytes, particularly in the
superficial layer. Sections of OA cartilage showed abundant specifically stained chondrocytes in the upper two-thuds of cartilage. Not
shown here are controls performed with nonimmune mouse IgG1,
showing only background staining. (Original magnification x 100.)
Cross-linking of IGF-1 receptor. To assess the
molecular weight of IGF receptors, experiments using
the homobifunctional cross-linker DSS were carried
out, followed by SDS-PAGE on 6% polyacrylamide
gels (Figure 2). Covalent cross-linking experiments
using human chondrocytes with radiolabeled
rHuIGF- 1 under reducing conditions showed a band of
137 kd and another of 275 kd, which were abolished in
the presence of excess radioinert rHuIGF-1. After
subtracting the molecular weight of IGF-1, the lower
molecular weight band corresponded to the a-chain of
the IGF-1 receptor (130 kd), and the higher molecular
weight band (260 kd) likely represented a dimerization
of the a-chain, as previously reported (22,23).
Expression of type 1 IGF receptor. Northern blot
analysis using a specific antisense RNA probe for type
1 IGF receptor revealed that the poly(A)+ enriched
fraction of RNA from freshly released chondrocytes
contained transcripts of 11.O kb (data not illustrated),
characteristic of the human type 1 IGF receptor
Immunohistochemical findings. Immunohistochemical studies using MAb aIR3 were performed on
3 normal cartilage samples and 3 OA cartilage samples. As illustrated in Figure 3, in normal cartilage only
a few chondrocytes at the superficial layer showed
positive specific staining for the type 1 IGF receptor.
In contrast, staining of OA cartilage with aIR3 revealed that a large number of the chondrocytes in all
layers of cartilage, but particularly in the superficial
and middle layers, were positive for the type 1 IGF
Biologic response to rHuIGF-1. In order to compare the relative sensitivity of normal and OA chondrocytes to stimulation by rHuIGF- 1, we investigated
the effect of IGF-1 on the rate of protein synthesis, via
the incorporation of labeled proline in these cells. The
effects of FCS and cycloheximide were also tested.
Preliminary experiments using normal chondrocytes
and increasing concentrations of rHuIGF-1 (0,30,100,
300, and 1,OOO ng/ml) showed a dose-response curve
which reached a plateau at 100 nglml. For OA chondrocytes, the concentration of rHuIGF- 1 was extended (&l0,000 ng/ml), and no response was found
even at 10,000 ng/ml. A concentration of 100 ng/ml was
then used for additional testing. Although under basal
conditions OA chondrocytes (n = 6) had a higher level
of synthetic activity (1,392 ? 508 X lo3 disintegrations
per minute) than normal chondrocytes (n = 4) (379 5
171 x lo3 dpm), the former were unresponsive to
stimulation by rHuIGF- 1 (Figure 4). However, normal
chondrocytes showed a 2.5-fold enhancement in protein synthesis. Cycloheximide induced a similar reduction of proline synthesis in both cell types. FCS at 10%
had a substantial effect on normal chondrocytes with a
3.5-fold increase in protein synthesis, while OA chondrocytes showed only a 2.2-fold increment.
Identification of IGFBPs. In view of the results
demonstrating a discrepancy between the level of
binding of IGF-1 to OA chondrocytes and the biologic
response of these cells to stimulation by this factor,
one must consider the possibility of interference by
endogenous binding proteins (IGFBP) in the IGF- 1
binding assay, particularly since insulin failed to compete for radiolabeled IGF-1. We thus examined if
chondrocytes were capable of synthesizing IGFBPs.
Normal chondrocytes were incubated for 2 days at
€3 200
Figure 4. Incorporation of radiolabeled proline in adult human normal and osteoarthritic chondrocytes, in
response to treatment with cycloheximide (Cyclo.), insutin-like growth factor 1 (IGF-l), or fetal calf serum
(FCS). Values are the mean and SEM percentage in relation to controls (Cont.).
37"C, and the presence of binding proteins in culture
medium was analyzed by the Western ligand blot
technique. Bands of approximately 43, 37, 33, 25, and
23 kd were found and appeared to correspond to the
molecular weight described for the IGFBP (25,26). On
the other hand, although the first 2 bands correspond
to the IGFBP-3 doublet, the others might represent
degradation products of the latter (26,27). Characterization of these bands is under way. A comparison
between normal (n = 3) and OA (n = 3) chondrocytes
showed that the bands corresponding to IGFBP-3 had
a much higher intensity than those from normal chondrocytes.
We further examined the presence of the
IGFBP on normal and OA chondrocytes using the
affinity cross-linking technique on an SDS-PAGE
buffer with a 13% polyacrylamide gel. Normal and OA
chondrocytes were incubated with "'I-IGF-l
in the
presence or absence of a 1,000-fold molar concentration of IGF-1 or insulin. For comparison purposes,
human adult skin fibroblasts were also included in
these experiments. As illustrated in Figure 5 , normal
and OA chondrocytes showed bands corresponding to
the molecular weight of the IGFBPs. After subtraction
of the molecular weight of the ligands (IGF-1), OA
cells showed bands of approximately 40,38,35,30,24,
and 15 kd, corresponding to IGFBP-3 (doublet of 40 kd
and 38 kd), IGFBP-2 (35 kd), IGFBP-1, -5, or the
glycosylated form of IGFBP-4 (30 kd), and IGFBP-4 or
-6 (24 kd). The smaller band (15 kd) has not yet been
identified but may represent either unknown IGFBPs
or degradation products (26,27). Higher bands were
also noted: a band at 130 kd reflects the type 1 IGF
receptor, and another band, of -74 kd was often
present and has not yet been identified. In normal
cells, only the 74-kd, 40-kd, and 38-kd bands, and the
band corresponding to type 1 IGF receptor were
apparent. The 40/38-kd bands had a lower intensity
than in OA cells. Human skin fibroblasts demonstrated
mostly the 130-kd band. Insulin, as expected, did not
compete with IGFBPs.
In osteoarthritis, the balance between the synthesis and degradation of matrix macromolecules is
disturbed, favoring tissue catabolism. In this context,
it would seem of utmost interest to have a better
understanding of the role of growth factors in this
equilibrium. Among the various growth factors identified to date, IGF-1 appears to be one of the most
important candidates in terms of its influence on the
anabolism of cartilage macromolecules. More precisely, IGF-1 in fetal calf serum (3) and in human
synovial fluid (28) has been shown to influence proteoglycan synthesis on bovine cartilage. However, the
"e 8
Mr ( m a )
v i v
+ IGF-1
$e <*
$e $
v j u v
v ; d d
Figure 5. Affinity cross-linking of radiolabeled insulin-like growth
factor 1 (IGF-I), performed on 13% polyacrylamide gels. Normal
and osteoarthritic human chondrocytes (C.N and C.OA), as well as
normal human adult skin fibroblasts (S. Fibro), were incubated with
0.1 nM radioligand alone (Total Binding) or in the presence of a
1,000-fold molar excess of unlabeled IGF-1 (Binding + IGF-I) or
insulin (Binding + Insulin), followed by cross-linking with 50 nM
disuccinimidyl suberate. Molecular weight standards are shown on
the left.
effect of IGF-1 in OA cartilage remains largely unknown.
Since previous studies have shown that 1)
IGF-1 reduces the degradation of normal cartilage in
vitro by its direct action on decreasing both the basal
and the cytokine-stimulated degradation of proteoglycans (29,30), 2) serum IGF-1 levels appear to be
reduced in OA (31) and in juvenile chronic arthritis
(32), and 3) synovial fluid levels of IGF-1 are decreased in various arthritides (33), one may be tempted
to conclude that a decrease in the availability of this
anabolic factor may be responsible for the persistence
of the osteoarthritic condition. However, recent findings showing the increased synthesis of IGF-1 by OA
chondrocytes (34) combined with the enhancement of
IGF- 1 in the matrix and IGF- 1 mRNA in chondrocytes
of the superficial layer of human OA articular cartilage
(35) render this hypothesis very unlikely. As the
disease progresses, the loss of cartilage occurs in
combination with a decreased matrix synthesis. Thus,
a plausible hypothesis might be that a disturbance in
the action of IGF-1 at the chondrocyte level is an
important element in the disease tissue, causing a
sequence of events that favors the degradation of
The data presented in this report represent the
first characterization of the type 1 IGF receptor in
adult human articular chondrocytes. Moreover, our
data revealed that the higher binding capacity of OA
chondrocytes toward IGF-1 is related not only to an
increase in type 1 IGF receptor, but also to the
presence of IGFBPs. The higher level of these proteins
at the chondrocyte surface may be responsible for the
resistance of OA cells to stimulation by IGF-1.
In this study, we demonstrated that the type 1
IGF receptor is expressed (mRNA) and present in
adult human articular chondrocytes and corresponds
to a single class of high-affinity receptors. The equilibrium dissociation constant as well as the number of
binding sites on normal adult chondrocytes correlates
well with previously published results using bovine,
lapine, or rat chondrocytes (13,36,37). The presence of
such a receptor on the chondrocyte membrane is also
confirmed by the results of internalization experiments
showing a rapid disappearance of radiolabeled IGF- 1
from the cell membranes.
Although insulin showed only a slight displacement of IGF-1 in chondrocytes, additional evidence of
the receptor specificity was provided. Experiments
with aIR3, the anti-type 1 IGF receptor MAb, showed
effective inhibition of the binding of IGF-1 by normal
chondrocytes. Furthermore, the identification, by affinity cross-linking (Figure 2), of bands of 130 kd and
260 kd are also characteristic of the type 1 IGF
receptor. The 260-kd band may very well correspond
to the cross-linked dimer of the type 1 IGF receptor
a-subunits (22,23).
The results of binding experiments using OA
chondrocytes indicate that these cells bind IGF- 1 with
a weaker aftinity, but with a much greater capacity,
than their normal counterparts. This finding may indicate that the level of type 1 IGF receptor is elevated in
these cells. Although the results of immunohistochemistry studies are indicative of an increase in the receptor level in OA chondrocytes, one must not exclude
the possibility that the diseased cartilage also contains
factors that are conducive to increased IGF-1 binding.
The presence of such factors in these diseased cells is
very plausible and is substantiated by other findings in
this study. First, we did not find any age-related
changes in IGF-1 binding in chondrocytes from macroscopically normal cartilage. Second, the fact that the
elevation in IGF- 1 binding was partly reversible after
passaging of the OA cells and that aIR3 only partially
blocked the binding in the diseased cells strongly
suggests that in situ factors may influence the level of
binding. Third, and most important, is the nonresponsiveness of OA chondrocytes to IGF-1 stimulation,
which contrasts with normal cells. This nonresponse
of OA chondrocytes was not caused by a general
inhibition of protein synthesis, since these cells were
shown to respond to FCS.
The reduced effect of FCS on OA chondrocytes
may be explained by the presence in the FCS of
factors other than IGF-1, which would act via another
metabolic pathway. The resistance of chondrocytes to
IGF-1 does not appear to be restricted to human OA
chondrocytes: Joosten et a1 (38) and Schalkwijk et al
(39) reported that chondrocytes from mice with experimental reactive arthritis are also nonresponsive to
IGF-1, with respect to maintenance of proteoglycan
In view of the results from the present study,
one might suspect that IGF binding protein interferes
in OA tissue. The human IGFBP family consists of at
least 6 proteins (termed IGFBP-1 to IGFBP-6), whose
genes have been cloned and proteins sequenced (40).
These proteins are found extracellularly as well as
membrane bound (25; for review, see ref. 41), and
their association constants (Kd 0.2-1 nM) are comparable with the affinity of the IGF receptors. The
secretion of these proteins has been extensively delineated in several cell types, including cartilage chondrocytes from young rabbits (26).
In the present study, we showed that adult
human chondrocytes in monolayer culture produce
IGFBP and that chondrocytes possess membraneassociated IGFBP. In normal chondrocytes, the cellassociated IGFBP appeared to be mostly IGFBP-3
(25,42). In OA, a large variety of bands representing
known IGFBP were found. An important finding was
the major enhancement in the density of bands corresponding to IGFBP-3 in the OA tissue. This is of
special interest since IGFBP-3 was found, in other cell
types, to be more effective than other IGFBP in
blocking the binding of IGF-1 to its receptor (43).
However, in vitro studies have shown discrepancies
with regard to the inhibition or potentiation effects of
IGF-I by IGFBP-3 at the cellular level (4345). Nonetheless, these binding proteins may be of particular
importance in modulating IGF-1 interaction with its
receptor in OA cartilage. This hypothesis appears
even more likely given the finding by Tesch et al (2)
that when IGF-1 is complexed with binding proteins,
the chondrocytes are unable to maintain their proteoglycan synthesis level.
In summary, our findings indicate that articular
chondrocytes possess type 1 IGF receptor and, although OA chondrocytes appear to have more binding
sites than do normal chondrocytes, these diseased
cells are not able to respond to IGF-1 stimulation. The
latter phenomenon appears to be related to an increased level of IGFBP in the cartilage. These results
are likely to have important pathophysiologic relevance in OA, in that the presence of these binding
proteins may have a marked effect on the bioavailability of IGF-I. We may have identified, in this study, a
metabolic disorder in the action of IGF-I , associated
with the development of OA. Further experiments
designed to characterize these IGFBP, their level, and
their role in OA are now under way. These are of
prime importance since the regulation of IGF-1 action
may be an important factor in the development of the
chronic form of this degenerative disease.
The authors wish to thank M. Zafarullah for helpful
discussions regarding molecular biology and K. Kiansa for
providing technical assistance. We also thank S. Fiori for
secretarial support.
1. Isaksson OG, Ohlsson C, Nilsson A, Isgaard J, Lindahl A:
Regulation of cartilage growth by growth hormone and insulinlike growth factor I. Pathology 5:451-453, 1991
2. Tesch GH, Handley CJ, Cornell HJ, Herington AC: Effects of
free and bound insulin-like growth factors on proteoglycan
metabolism in articular cartilage explants. J Orthop Res 10: 1&
22, 1992
3. McQuillan DJ, Handley CJ, Campbell MA, Bolis S,Milway VE,
Herington AC: Stimulation of proteoglycan biosynthesis by
serum and insulin-like growth factor-1 in cultured bovine articular cartilage. Biochem J 240:423-430, 1986
4. Willis DH Jr, Liberti JP: Post-receptor actions of somatomedin
on chondrocyte collagen biosynthesis. Biochim Biophys Acta
844:7240, 1985
5. Guenther HL, Guenther HE, Froesch EF, Fleisch H: Effect of
insulin-like growth factor on collagen and glycosaminoglycan.
Expenentia 38:97!&981, 1982
6. Demarquay D, Dumontier MF, Tsagns L, Bourguigon J, Nataf
V, Corvol MT: In vitro insulin-like growth factor I interaction
with cartilage cells derived from postnatal animals. Hormone
Res 33:111-114, 1990
7. Neely EK, Beukers MW, Oh Y, Cohen P, Rosenfeld RG:
Insulin-like growth factor receptors. Acta Paediatr Scand
(Suppl) 372:116-123, 1991
8. Trippel SB, Chernausek SD, Van Wyk JJ, Moses AC, Mankin
HJ: Demonstration of type I and type I1 somatomedin receptors
on bovine growth plate chondrocytes. J Orthop Res 65317-826,
9. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K,
Christy W, Cooke TD, Greenwald R, Hochberg M, Howell D,
Kaplan D, Koopman W, Longley S 111, Mankin H, McShane
DJ, Medsger T Jr, Meenan R, Mikkelsen W, Moskowitz R,
Murphy W, Rothschild B, Segal M, Sokoloff L, Wolfe F:
Development of criteria for the classification and reporting of
osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 29:1039-1049, 1986
10. Meachim G, Hardinge K, Williams DR: Methods for correlating
pathological and radiological findings in osteoarthritis of the hip.
Br J Radio1 49670476, 1972
11. Mankin HJ, Dorfman H, Lippiello L, Zarins A: Biochemical
and metabolic abnormalities in articular cartilage from osteoarthritic human hips. 11. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg 53523-537, 1971
12. Martel-Pelletier J, McCollum R, DiBattista J, Faure M-P, Chin
JA, Fournier S, Sarfati M, Pelletier J-P: The interleukin-1
receptor in normal and osteoarthritic human articular chondrocytes: identification as the type I receptor and analysis of
binding kinetics and biologic function. Arthritis Rheum 35530540, 1992
13. Watanabe N, Rosenfeld RG,Hintz RL, Dollar LA, Smith R L
Characterization of a specific insulin-like growth factor4
somatomedin-C receptor on high density, primary monolayer
cultures of bovine articular chondrocytes: regulation of receptor
concentration by somatomedin, insulin and growth hormone. J
Endocrinol 107:275-283, 1985
14. Rosenfeld RG, Conover CA, Hodges D, Lee PDK, Misra P,
Hintz L, Li CH: Heterogeneity of insulin-like growth factor-I
affinity for the insulin-like growth factor-I1 receptor: comparison of natural, synthetic and recombinant DNA-derived insulinlike growth factor-I. Biochem Biophys Res Commun 143:199205, 1987
15. Kull FC, Jacobs S, Su YF, Svoboda ME, Van Wyk JJ,
Cuatrecasas P: Monoclonal antibodies to receptors for insulin
and somatomedin-C. J Biol Chem 256:65614566, 1983
16. Hossenlopp P, Seurin D, Segovia-Quinson B, Hardoin S,
Binoux M: Analysis of serum insulin-like growth factor binding
proteins using Western blotting: use of the method for titration
of the binding proteins and competitive binding sites. Anal
Biochem 154:138-143, 1986
17. Pelletier J-P, Mineau F, Faure M-P, Martel-Pelletier J: Imbalance between the mechanisms of activation and inhibition of
metalloproteases in the early lesions of experimental osteoarthritis. Arthritis Rheum 33:1466-1476, 1990
18. Aiba H, Adhya S, de Crombrugghe B: Evidence for two
functional gal promoters in intact Escherichia coli cells. J Biol
Chem 256:11905-11910, 1981
19. Ullrich A, Gray A, Tam AW, Yang-Feng T, Tsubokawa M,
Collins C, Henzel W, Bon TL, Kathuria S, Chen E, Jakobs S,
Francke U, Ramachandran J, Fujita-Yamaguchi Y: Insulin-like
growth factor I receptor primary structure: comparison with
insulin receptor suggests structural determinants that define
functional specificity. EMBO J 5:2503-2512, 1986
20. Willis DH Jr, Liberti JP: Post-receptor actions of somatomedin
on chondrocyte collagen biosynthesis. Biochim Biophys Acta
8M72-80, 1985
21. Munson PJ, Rodbard D: Ligand: a versatile computerized
approach for characterization of ligand-binding systems. Anal
Biochem 107:220-239, 1980
22. Park JHY, Vanderhoof JA, Blackwood D, MacDonald RG:
Characterization of type I and type I1 insulin-like growth factor
receptors in an intestinal epithelial cell line. Endocrinology
126:2998-3005, 1990
23. Conover CA, Mistra P, Hintz RL, Rosenfeld RG: Differential
binding of '251-IGF-Ipreparations to human fibroblast monolayers. Acta Endocrinol (Copenh) 118513-520, 1988
24. Abbott AM, Bueno R, Pedrini MT, Murray JM, Smith RJ:
Insulin-likegrowth factor I receptor gene structure. J Biol Chem
1910759-10763, 1992
25. Drop SLS, Schuller AGP, Lindenbergh-Kortleve DJ, Groffen
C, Brinkman A, Zwarthoff EC: Structural aspects of the IGFBP
family. Growth Regul2:69-79, 1992
26. Froger-Gaillard B, Hossenlopp P, Adolphe M, Binoux M:
Production of insulin-like growth factors and their binding
proteins by rabbit articular chondrocytes: relationships with cell
multiplication. Endocrinology 124:2365-2372, 1989
27. Cohen P, Graves HCB, Peehl DM, Kamarei M, Giudice LC,
Rosefeld RG: Prostate-specific antigen (PSA) is an insulin-like
growth factor binding protein-3 found in seminal plasma. J Clin
Endocrinol Metab 75: 1046-1053, 1992
28. Schalkwijk J, Joosten LAB, van den Berg WB, van Wyk JJ, van
de Putte LBA: Insulin-like growth factor stimulation of chondrocyte proteoglycan synthesis by human synovial fluid. Arthritis Rheum 32:66-71, 1989
29. Fosang AJ, Tyler JA, Hardingham TE: Effect of Interleukin-1
and insulin-like growth factor-I on the release of proteoglycan
components and hyaluronan from pig articular cartilage in
explant culture. Matrix 11:17-24, 1991
30. Tyler JA: Insulin-like growth factor I can decrease degradation
and promote synthesis of proteoglycan in cartilage exposed to
cytokines. Biochem J 260543448, 1989
31. Denko CW, Boja B, Moskowitz RW: Growth promoting peptides in osteoarthritis: insulin, insulin-like growth factor-I,
growth hormone. J Rheumatol 17:1217-1221, 1990
32. Aitman TJ, Palmer RG, Loftus J, Ansell BM, Royston JP, Teale
JD, Clayton RN: Serum IGF-I levels and growth failure in
juvenile chronic arthritis. Clin Exp Rheumatol 7557-561, 1989
33. Coates CL, Burwell RG, Lloyd-Jones K, Swannell AJ, Walker
G, Selby C: Somatomedin activity in synovial fluid from patients with joint diseases. Ann Rheum Dis 37:303-314, 1978
34. Dor6 S, Rousseau N, Pelletier JP, TardifG, Abribat T, Brazeau
P, Martel-Pelletier J: Expression of growth hormone receptor
and absence of response to growth hormones in human articular
cartilage (abstract). Seventy-fifth Annual Meeting of The Endocrine Society, Las Vegas, Nevada, 1993
35. Middleton JFS, Tyler JA: Upregulation of insulin-like growth
factor I gene expression in the lesions of osteoarthritic human
articular cartilage. Ann Rheum Dis 51:440447, 1992
36. Schalch DS, Sessions CM, Farley AC, Masakawa A, Emler CA,
Dills D: Interaction of insulin-like growth factor Vsomatomedin-C
with cultured rat chondrocytes: receptor binding and internalization. Endocrinology 118:1590-1597, 1986
37. Trippel SB, Corvol MT, Dumontier MF, Rappaport R, Hung
HH, Mankin HJ: Effect of somatomedin-C/insulin-like growth
factor I and growth hormone on cultured growth plate and
articular chondrocytes. Pediatr Res 2576-82, 1989
38. Joosten LA, Schalkwijk J, van den Berg WB, van de Putte
LBA: Chondrocyte unresponsiveness to insulin-like growth
factor-1: a novel pathogenetic mechanisms for cartilage destruction in experimental arthritis. Agents Actions 26: 193-195, 1989
39. Schalkwijk J, Joosten LAB, van den Berg WB, van de Putte
LBA: Chondrocyte nonresponsiveness to insulin-like growth
factor I in experimental arthritis. Arthritis Rheum 32:89&590,
40. Rechler MM, Brown AL: Insulin-like growth factor binding
proteins: gene structure and expression. Growth Regul2:5548,
41. Cohick WS, Clemmons DR: The insulin-like growth factors.
Ann Rev Physiol 55:131-153, 1993
42. Sommer A, Maack CA, Spratt SK, Hunt TK, Spencer EM:
Molecular genetics and actions of recombinant IGFBP-3, Mode m Concepts of IGFs. Edited by EM Spencer. New York,
Elsevier, 1991
43. McCusker RH, Busby WH, Dehoff MH, Camacho-Hubner C,
Clemmons DR: Insulin-like growth factor (IGF) binding to cell
monolayers is directly modulated by the addition of IGF-binding
proteins. Endocrinology 129:939-949, 1991
44. Baxter RC, Martin JL: Binding proteins for the insulin-like
growth factors: structure, regulation and function. Prog Growth
Factor Res 1:49-68, 1989
45. Conover CA, Ronk M, Lombana F, Powell DR: Structural and
biological characterization of bovine insulin-like growth factor
binding protein-3. Endocrinology 127:2795-2893, 1990
Без категории
Размер файла
1 065 Кб
site, growth, unresponsive, osteoarthritis, possess, human, stimulating, like, factors, increase, number, insulin, binding, chondrocyte
Пожаловаться на содержимое документа