| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
-Stimulated Gene-6) in Murine Models of Experimental Arthritis
From the Departments of Orthopedic Surgery*
and
Biochemistry,
Section of Biochemistry and
Molecular Biology, Rush University, Rush-Presbyterian-St. Luke’s
Medical Center, Chicago, Illinois
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TNF-)-stimulated gene-6 (TSG-6) is
up-regulated by various cytokines and growth factors. TSG-6 binds to
hyaluronan in inflamed synovial tissue and forms a complex with a
serine protease inter--trypsin inhibitor (II), increasing
the protease inhibitory effect of II >100-fold. The TSG-6/II
complex then blocks serine proteases, including the
plasminogen-plasmin activation, probably the most important
component in the activation processes of matrix metalloproteinases. To
gain insight into the mechanisms of TSG-6 action in arthritis,
we have used an autoimmune murine model (proteoglycan-induced
arthritis) for systemic, and a monoarticular form of
arthritis (antigen-induced arthritis) for local treatment of arthritis
with recombinant mouse TSG-6 (rmTSG-6). Intravenous injection of
rmTSG-6 induced a dramatic reduction of edema in acutely inflamed
joints by immobilizing CD44-bound hyaluronan and, in long-term
treatment, protected cartilage from degradation and blocked
subchondral and periosteal bone erosion in inflamed joints. The
intra-articular injection of a single dose (100 µg) of rmTSG-6
exhibited a strong chondroprotective effect for up to 5 to 7
days, preventing cartilage proteoglycan from
metalloproteinase-induced degradation. In contrast, rmTSG-6 did
not postpone the onset, nor reduce the incidence of arthritis.
We were unable to detect any significant differences between control
and rmTSG-6-treated animals when various serum markers (including pro-
and anti-inflammatory cytokines, auto- and heteroantibody
productions) or antigen-specific T-cell responses were
compared, nor when the expressions of numerous cell surface
receptors or adhesion molecules were measured. TSG-6 seems to play a
critical negative regulatory feed-back function in
inflammation, especially in arthritic
processes.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TNF-)-stimulated gene 6 (TSG-6) was first identified by
differential screening of a cDNA library of TNF--stimulated human
fibroblasts.1
Although a constitutive expression is low or
not detectable, virtually all cell types are able to produce TSG-6 on
stimulation with various components either in vitro or
in vivo.2-5 TSG-6 is a 35-kd secreted glycoprotein.2 The N-terminal half of the amino acid sequence shows homology to the Link module, which is a conservative sequence in hyaluronan (HA)-binding proteins such as CD44, cartilage link protein, and G1 domains of aggrecan and versican.2,6 The C-terminal half of TSG-6 (CUB domain) shares sequence homologies with the A chain of complement component C1s/C1r, uEGF, a protein involved in development of sea urchin embryos and bone morphogenic protein-1.7,8 Based on these structural homologies, it is strongly believed that TSG-6 may play a crucial role in extracellular matrix formation, inflammatory cell migration, cell proliferation, and developmental processes.4,7
Although TSG-6 cannot be detected in synovial fluids or sera of normal
individuals, synoviocytes from rheumatoid patients constitutively
produce TSG-6 protein and it can be detected in sera of patients with
rheumatoid arthritis, or less frequently in
osteoarthritis.3,9,10
Exogenous TNF- or interleukin
(IL)-1 further increases in TSG-6 production by rheumatoid
synoviocytes.3
A 120-kd complex containing TSG-6 was
regularly detected in TSG-6 (35 kd)-producing cultures in the presence
of serum, but not in serum-free medium.3,9,11
The protein
that bound TSG-6 was purified from human serum and identified as
inter--inhibitor (II).9,11
II is a HA-binding
serum protein with serine protease inhibitory activity.12
The target enzymes include trypsin, chymotrypsin, cathepsin G,
leukocyte elastase, acrosin, and plasmin.8,10,12
Interestingly, when TSG-6 is complexed with II, the complex exhibits
a more than 100-fold higher protease inhibitory effect against plasmin
than the II alone.9,11
This effect of TSG-6 on II
seems to be specific for plasmin, as no such increase in inhibitory
effect was observed on trypsin or neutrophil elastase.11
To further test the inhibitory effects of the TSG-6/II complex on
protease network and inflammation, a murine air-pouch model of acute
inflammation was used.11
Simultaneous injection of human
recombinant TSG-6 with carrageenan into the air pouch exhibited a
potent anti-inflammatory effect.11
Recently, it was shown
that recombinant TSG-6 had a beneficial effect on collagen-induced
arthritis (CIA) reducing both the incidence and severity of
arthritis.13
Taken together, TSG-6 induced by
proinflammatory cytokines at the site of inflammation, binds tightly to
II, and dramatically increasing the II’s serine protease
inhibitory effect. Thus, TSG-6 may be a key component in a negative
feedback loop controlling inflammatory response in local arthritic
processes.
In this article, we investigated whether TSG-6 plays a role in prevention of cartilage breakdown in two murine models of experimental arthritis. We studied the in vivo role of TSG-6 in a systemic autoimmune disease of proteoglycan-induced arthritis (PGIA), and then in a monoarticular form of inflammatory arthritis by injection of recombinant TSG-6 into the knee joints of animals with antigen-induced arthritis (AIA).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Purified plasmid DNA (Lonza pEE14.1 vector; Lonza Biologics plc., Slough, Berkshire, UK) containing a full-length (1654 bp) mouse TSG-6 cDNA clone4 was used for transfection of Chinese hamster ovary (CHO-K1; American Type Culture Collection, Rockville, MD) cells by the standard method of CaCl2 precipitation. Transfected cells were cultured in glutamine-free condition in the presence of 25 µmol/L methionine sulfoximine (Sigma Chemical Co., St. Louis, MO) for the selection of stable transfected cell lines. rmTSG-6 was purified from culture media of stable transfected cloned cell line using HA-coupled EAH-Sepharose (Pharmacia Biotech, Piscataway, NJ). HA-bound rmTSG-6 was eluted with 4 mol/L guanidine hydrochloride in 0.1 mol/L Na-acetate buffer (pH 5.8), dialyzed against distilled water, and then lyophilized. Purified rmTSG-6 was tested by Western blotting using a rabbit polyclonal antibody (TSG-6-CR21) raised against a synthetic peptide (135CGGVFTDPKRIFKSPG), located at the end-terminal end of the CUB domain.4 For quantitative measurement of serum TSG-6 (either measuring serum level in arthritic animals or during the treatments) we used a sandwich enzyme-linked immunosorbent assay (ELISA), confirmed by Western blots. In ELISA, the IgG fraction of rabbit serum to mouse TSG-6 synthetic peptide (36GVYHREARAGRYKL) was used as capture antibody and affinity-purified and biotinylated TSG-6-CR21 antibody for detection of TSG-6 following a protocol described for other serum markers.14 For organ/tissue distribution and kinetic studies, rmTSG-6 was labeled with 125I using a chloramine-T method.
Immunization Protocol for PGIA, Scoring System, and Treatment with rmTSG-6
High-density cartilage proteoglycan aggrecan (PG) was purified by CsCl gradient centrifugation from human cartilage and depleted of glycosaminoglycan side chains as described.15 All animal experiments were approved by the Institutional Animal Care and Use Committee (Rush University, Chicago, IL). Female BALB/c mice (National Cancer Institute, Frederick, MD) were immunized intraperitoneally with cartilage PG.15,16 First antigen injection (100 µg protein) was given in complete Freund’s adjuvant (Difco, MI), and the same doses of antigen were injected as second and third boosters in incomplete Freund’s adjuvant. Typically, BALB/c mice developed swelling and redness in one or more limbs, 7 to 14 days after the third injection of PG.15,16 This stage of immunized animals (from 7 days after the third injection) is regarded as the prearthritic phase of PGIA. A standard scoring system, based on swelling and redness of paws, was used for the assessment of the severity of arthritis.17,18 The time of appearance of swelling and redness was recorded as the time of onset of arthritis. Joint swelling was scored (ranging from 0 to 4 of each paw) and expressed as cumulative acute arthritis score resulting in a maximum score of 16. During the treatment period, paw (joint) thicknesses at frontal and sagittal planes of all ankles and wrists were measured daily with a microcaliper.17 Arthritic animals were injected intraperitoneally or intravenously via the retro-orbicular venous plexus with 100 µl phosphate-buffered saline (PBS), rmTSG-6 (100 µg in PBS), or rmTSG-6 and Pep-1 simultaneously. Pep-1 is a synthetic peptide (NH2-GAHWQFNALTVRGGGS) that shows HA-binding properties19 and can also bind to the Link module of either CD44 or TSG-6 (our unpublished data). The optimal therapeutic dose, serum/tissue distribution, half-life, and kinetics of injected rmTSG-6 was determined in preliminary experiments. Sera were collected before the injection, or every second hour in kinetic studies, and at the end of the experiment. Spleen cells for in vitro tests were isolated at the end of the experiment and paws from all mice were collected and fixed in 10% formalin containing 5% cetylpirimidium chloride (Sigma). Legs were decalcified in 5% formic acid, paraffin embedded, and sections processed for routine histology or immunohistochemistry.15,16,20,21
Immunization Protocol for AIA and Experimental Groups
C57BL/6j mice (National Cancer Institute) were immunized with 100 µg of methylated bovine serum albumin (mBSA; Sigma) in 100 µl of PBS emulsified in 100 µl of complete Freund’s adjuvant. Injections were given subcutaneously in the flanks, and into the proximal tail. After 1 week, mice were subcutaneously boosted with 100 µg of mBSA emulsified in 100 µl of incomplete Freund’s adjuvant, and the same boost was repeated 3 weeks later.21 Two to 3 weeks after the final boost, knees of immunized mice under ketamine/xylazine anesthesia were injected intra-articularly in the following combination: animals of group 1 received a single dose of 60 µg of mBSA in 6 µl of PBS into the right knee, and 6 µl of PBS into the left knee joint. Animals of group 2 received a single dose of 60 µg of mBSA and 100 µg of rmTSG-6 in 6 µl of PBS intra-articularly into the right knee, and 60 µg of mBSA in 6 µl of PBS into the left knee joint. A therapeutic dose of rmTSG-6 injected intra-articularly with mBSA was found to be 25 to 50 µg in preliminary experiments. To avoid repeated intra-articular injections, we used a single double dose (100 µg) of rmTSG-6 throughout this study. Mice (n = 6 to 10 at each time point) were sacrificed on days 1, 3, 5, 7, 12, and 24 after the intra-articular injections. Left and right knees from each mouse were harvested, fixed in 10% formalin containing 5% cetylpirimidium chloride, decalcified, and paraffin embedded. Five- to 7-µm sections were stained with hematoxylin and eosin (H&E), or safranin-O counterstained with fast green. For immunohistochemical analysis, sections were deparaffinized, rehydrated in PBS, and pretreated with chondroitinase ABC (Seikagaku America of Cape Cod, Falmouth, MA) as described.20 Sections were stained with polyclonal antibodies to neoepitopes -VDIPEN341 [a sequence that is generated by stromelysin-1 (MMP-3) cleavage in the interglobular domain of cartilage aggrecan] and -NITEGE373 (aggrecanase-generated neoepitope) (antibodies were generous gifts from Dr. J. Mort, Shriners Hospital, Montreal, Quebec, Canada and Dr. J. Sandy, Shriners Hospital, Tampa, FL).21-23 Affinity-purified peroxidase-labeled anti-rabbit immunoglobulin (Accurate Chemical and Scientific Co., Westbury, NY) was used as a second-step antibody and the peroxidase reaction developed by Fast 3,3'-diaminobenzidine (DAB; Sigma) and H2O2 for 5 minutes as described.20 The DAB reaction was enhanced using DAB Substrate Enhancer (Zymed Laboratories Inc, San Francisco, CA) for 3 minutes. The sections were dehydrated and mounted with Permount.
Measurements of Antigen-Specific Antibodies, T-Cell Responses, and Cytokine Production
MaxiSorp immunoplates (Nunc International, Hanover Park, IL) were coated with human or mouse cartilage PGs (0.1 µg protein/100 µl/well) for ELISAs as described.24,25 Sera were applied at increasing dilutions and the isotypes of PG-specific antibodies were determined by peroxidase-conjugated rat anti-mouse IgG1 or IgG2a secondary antibodies (Zymed) as described.24,26 Serum antibody levels were normalized to mouse IgG1 and IgG2a.24,25
Antigen-specific T-cell responses (ie, IL-2 production and
proliferation), were measured in quadruplicate samples of spleen cells
(3 x 105
cells/well) cultured in the
presence of 50 µg PG protein/ml. IL-2 was measured in supernatants
harvested on day 2 by the proliferation of the IL-2-dependent CTLL-2
cell line.24,25
Antigen-specific T-cell proliferation was
assessed on day 5 by the incorporation of
3[H]-thymidine. Antigen (PG)-specific
interferon-
(IFN-) and IL-4 production by T cells were determined
in identical culture conditions as described for T-cell proliferation
in 4-day-old conditioned media (2.5 x 106
mononuclear cells/ml) using capture ELISAs from R&D Systems
(Minneapolis, MN). Serum levels of TNF-, IL-6, and IL-10 were
measured by capture ELISAs (R&D Systems or BD Phar-Mingen, San
Diego, CA) and IL-1 by a bioassay using D10S cell line as
described.24,25
Flow Cytometry
Immunostaining of cell-surface molecules and flow cytometry
were performed as described previously.14
Spleen and lymph
node cells were stained with monoclonal antibodies (mAbs) to various
cell surface molecules. These mAbs, purchased from BD PharMingen, were
raised against CD3 (T cell), CD4 (T-helper cell), CD8 (T-suppressor
cell), CD11a/b (leukocyte ß2 integrin ß subunits; LFA-1/Mac-1),
CD18 (ß2 integrin ß subunit; LFA-1), CD19 (B cell), CD29 (ß1
integrin subunit), CD25 (IL-2 receptor), CD44 (IM7), CD45/B220 (B
cell), CD49d (ß1 integrin 4 subunit; VLA-4), CD54 (ICAM-1), CD62L
(L-selectin), and CD62P (P-selectin). Isotype-matched irrelevant mAbs
were used as controls. Cell surface-bound antibodies were detected by
biotinylated goat-anti-rat IgG (Kirkegaard & Perry, Gaithersburg, MD)
and streptavidin-phycoerythrin or streptavidin-fluorescein
isothiocyanate (Life Technologies; Gaithersburg, MD).
Mouse synovial fibroblast and small venule endothelial cells (SVEC4-10) were digested with streptomyces hyaluronidase (Sigma) for 30 minutes at 37°C to remove endogenous HA.17 Mouse synovial fibroblast,17 SVEC4-10,19,27 and 5/4 T cells,18 each at 1 x 106 in 100 µl, were incubated with fluorescein-labeled HA (5 µg/assay)14 and rmTSG-6 (0.04 to 5 mg/ml), or unlabeled HA (5 mg/ml), as competitors for 2 hours at 37°C. Cells were washed in PBS, fixed in 2% paraformaldehyde (Sigma) and analyzed using a FACScan instrument and CellQuest software (Becton Dickinson, San Jose, CA).
Statistical Analysis
Statistical analysis was performed using SPSS v7.5 (SPSS, Chicago, IL). The Mann-Whitney and Wilcoxon tests were used for intergroup comparisons. For comparison of results of two groups, Student’s t-test and, for the determination of correlation coefficients, Spearman’s gamma test were used. Significance was set at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PGIA was generated in BALB/c mice by systemic immunization with
human cartilage aggrecan. Acutely arthritic animals were injected
intraperitoneally with 100 µl of PBS (control) or 50 to 250 µg
rmTSG-6 (in 100 µl of PBS) for 3 consecutive days. No effect was
observed on joint swelling in either the control or the rmTSG-6-treated
mice after three injections (data not shown). To further test the
feasibility of TSG-6 treatment, acutely arthritic mice with PGIA were
injected intravenously with 100 µg of rmTSG-6. A dramatic reduction
in joint edema of each swollen limb was observed within a few hours
(Figure 1A)
. Subsequent intravenous
injections further reduced joint swelling, whereas the effect was less
and less spectacular until the edematous swelling virtually disappeared
from acutely inflamed paws (Figure 1A
and Figure 2B
). Control mice injected with PBS
showed no change, or only a slight increase, in joint swelling within
this short-term experiment (Figure 1)
.
|
|
, however, was
indistinguishable from those treated with PBS. Because of the lack of
any effect of intraperitoneally injected rmTSG-6 versus the
dramatic reduction of edema after the intravenous injection, we
hypothesized that a free HA-binding capacity of TSG-6 should be
critical in function without affecting the inflammatory cell
infiltration. Indeed, when rmTSG-6 was added together with either a
large dose of HA (data not shown) or with HA-binding peptide Pep-1, the
beneficial effect of TSG-6 disappeared proportionally as the dose of
Pep-1 was increased (Figure 1B)
. In Vitro and In Vivo Tests of rmTSG-6 Ability to Compete for HA-Binding Site
The difference between treated animals and untreated controls was
the regression of joint swelling because of the loss of edema in the
rmTSG-6-treated group. To test if the loss of edema was because of the
ability of rmTSG-6 to compete with cell surface receptor CD44 for HA
binding, in vitro competition assays were performed using
synovial fibroblasts, small venule endothelial cells (SVEC4-10) (both
produce high levels of HA), and 5/4 T-cell hybridoma (a
non-HA-producing cell line). All these cell types express high amounts
of CD44, and hence can bind HA. Cells were incubated with
fluorescein-labeled HA and increasing concentrations of rmTSG-6, and
the HA binding was measured by fluorescence-activated cell sorting
analysis. As shown in Figure 3
, rmTSG-6
could indeed compete with CD44 and the effect was dose-dependent. If
this was the case in vivo, a HA-binding
peptide19
could also reduce joint inflammation and should
exhibit an additive or synergistic effect on TSG-6 treatment. As shown
in Figure 1B
, Pep-1 alone could not reduce joint swelling, even worse,
the co-injection of Pep-1 with rmTSG-6 significantly reduced the
anti-inflammatory effect of the TSG-6 (Figure 1B)
. Taken together,
TSG-6 had a very strong effect on joint inflammation by reducing tissue
edema. This effect was associated with the HA-binding capacity of the
TSG-6, which could be competitively inhibited with a synthetic peptide
with known HA-binding properties, whereas Pep-1 alone did not exhibit
any effect on joint edema.
|
Next, we have monitored the clinical symptoms and histological
abnormalities of inflamed joints during repeated (weekly) treatments
with rmTSG-6 in PGIA up to 28 days. Arthritic BALB/c mice
(n = 14) were injected intravenously with
rmTSG-6 (100 µg in PBS) every 24 hours for 7 days and joint/paw
swelling measured twice a day (Figure 4)
.
As the cumulative joint score (swelling of all paws) almost returned to
the pretreatment level after a 7-day treatment-free period (Figure 4A)
,
mostly because of the newly inflamed paws, a second treatment cycle was
repeated from day 14. As found during the first treatment period,
although less impressively, the cumulative joint score in
rmTSG-6-treated mice showed a statistically significant regression
(P < 0.05) (Figure 4A)
. In contrast, the
cumulative joint swelling in the PBS-injected (control) mice remained
significantly higher (P < 0.05) up to 3 weeks
(Figure 4A)
which was the consequence of the normal progression of the
disease.15,16
Interestingly, additional (new) joints
(paws) became inflamed (swollen) in rmTSG-6-treated mice only after
weaning of TSG-6 treatment (between days 7 and 14). Therefore, when
only the inflamed joint/paw diameters were calculated and shown on
Figure 4B, a
less impressive effect of the second-cycle treatment was
found. Although the anti-inflammatory (anti-edematous) effect remained
significantly low, relative to the joint diameters measured on the zero
day of the experimental group, the second-cycle treatment affected
essentially only the freshly (acutely) inflamed paws, thus the
anti-inflammatory effect of the second treatment period with rmTSG was
almost restricted to these new, acutely inflamed joints/paws (Figure 4A)
. Although the cellular infiltration and/or synovial cell
proliferation of inflamed joints were very similar in PBS- and
TSG-6-treated joints (Figure 2, C and D)
, massive cartilage and bone
erosions were found in joints only in PBS-treated animals. In turn,
systemic application of rmTSG-6 reduced edema and, more strikingly,
prevented articular cartilage from deterioration even in the presence
of massive inflammation (Figure 2, C
versus D).
|
and compared only those paws
that were inflamed before the treatment. The PBS-treated group showed
no change during the first 2 weeks, and then the acute inflammatory
symptoms moderately regressed and previously inflamed joints became
stiff. In contrast, the joint swelling of rmTSG-6-treated groups showed
a statistically significant regression (P <
0.05) within a few days (Figure 1A
and Figure 4A
) that remained
unchanged during the 4-week experimental period (Figure 4B)
. In other
words, once rmTSG-6 reduced edema and redness, inflammation did not
return, not even in the absence of treatment (between days 7 and 14 or
21 and 28) (Figure 4B)
. These results suggest that TSG-6 not only
protects arthritic joints from flare-up but also seems to prevent
nonarthritic limbs from the clinical symptoms of inflammation
determined as redness and swelling. Effect of rmTSG-6 Treatment on the Development of PGIA
During the intravenous first cycle of rmTSG-6 treatment (Figure 4)
, edema waned in arthritic paws and no additional joints became
inflamed. In contrast, previously nonarthritic paws became swollen in
PBS-injected animals (Figure 4, A and B)
, and this was also a frequent
observation in TSG-6-treated mice after the cessation of the treatment.
A significant increase in cumulative paw diameter after day 7 in
previously TSG-6-treated animals (Figure 4A)
clearly was the
consequence of these newly inflamed joints. Although conventional
histopathology did not reveal differences in cellular infiltration
(Figure 2)
, the anti-inflammatory effect of TSG-6 was found in
CIA13
and in an air-pouch model.11
All these
observations indicated a preventive effect of TSG-6 on arthritis
development. To test this preventive potential of TSG-6 in arthritis,
PG-immunized mice (n = 8) in the prearthritic
stage (after the third antigen injection) were treated intravenously
with a daily dose of 100 µg rmTSG-6 (a dose found to be optimal for
the suppression of inflammation) or PBS for 3 weeks. Although the
inflammation (cumulative score, 4.2 ± 2.8) was significantly
lower (P < 0.05) in rmTSG-6-treated mice than
in the PBS group (7.6 ± 1.7), neither the incidence nor the time
of onset differed in the two groups (data not shown), ie, TSG-6 had an
anti-edematous but not an anti-arthritic effect.
Pathophysiological Markers in rmTSG-6-Treated and Untreated Mice with PGIA
Because we found differences in severity between TSG-6-treated and
untreated animals, and then differences in cartilage and bone erosions,
it seemed to be prudent to measure serum levels of pro- and
anti-inflammatory cytokines and auto- and heteroantibodies, as well as
monitor the antigen-specific T-cell responses. Serum samples were
collected before the treatment cycles and at 24 hours after the last
(seventh) intravenous injection of rmTSG-6, and T-cell responses were
measured at the end of experiment. Treatment protocols were the same as
shown in Figure 4
. We could not detect any significant differences
between PBS- and rmTSG-6-injected mice measuring serum levels of
TNF-, IL-1, IL-6, IL-10, and auto- and heteroantibodies (including
the ratios of IgG1 and IgG2a), or measuring antigen (PG)-specific
T-cell proliferation, IL-2 (CTLL-2 assay), IL-4, or IFN- production
by spleen cells. Similarly, neither the T-cell subsets nor other
expression of cell surface receptors (mAbs listed in Materials and
Methods) differed when spleen or lymph node cells of the PBS- and
rmTSG-6-treated groups were compared (data not shown). Therefore, we
concluded that the systemic administration of rmTSG-6 probably has no
systemic effect, rather it might be critical at the site of
inflammation where tissue destruction occurs.
Histopathology of Knee Joints in AIA Under the Protection of TSG-6
To gain insight into the mechanism of the local chondroprotective
effect of TSG-6 (Figure 2D)
in a single joint, a murine model of AIA
was used.21
In this model system, the delayed-type
hypersensitivity reaction is localized at a single joint, other joints
can be used as controls, and cartilage degradation can be monitored
precisely on a daily basis. Intra-articular PBS injection in the left
knee joint of mBSA-immunized animals showed no inflammation (Figure 5A)
up to a 24-day period, and no
cartilage damage occurred (Figure 5
; B to E). The earliest response
(lymphocyte infiltration) to the intra-articular injection of mBSA was
detected on days 2 to 3 (data not shown) and became more uniformly
extensive by day 5 (Figure 5F)
. By this time, the loss of cartilage PG
from the superficial middle layers of the articular cartilage became
evident (Figure 5
; G to J). All mBSA-injected joints exhibited heavy
inflammation, independent of whether the mBSA was injected alone
(Figure 5F)
or with rmTSG-6 (Figure 5K)
. The loss of PG from cartilage,
detected by the loss of safranin-O staining, was moderate on day 3 (not
shown), but became extensive by day 5 in mBSA-injected knee joints
(Figure 5H)
. In accordance with the loss of safranin-O staining, an
intensive accumulation of immunostained -VDIPEN and -NITEGE neoepitopes
(the C-terminal cleavage sites of cartilage aggrecan by
metalloproteinases and aggrecanase) were detected in the superficial
layer of damaged cartilage (Figure 5, I and J)
. By day 7, inflammation
further progressed, pannus-like structures eroded bone and cartilage,
safranin-O staining significantly decreased down to the calcifying
zone, and the expression of -NITEGE neoepitope became slightly reduced
(not shown). Knee joints of mBSA-immunized mice were injected with mBSA
with or without rmTSG-6. As shown in Figure 5, F and K
, the
conventional histopathology of these knee joints (ie, H&E-stained
sections) exhibited no differences at any time point. In contrast,
safranin-O staining remained intense (compare the superficial layers of
cartilage in Figure 5, H
with M) up to 5 to 7 days after the
intra-articular administration of mBSA with rmTSG-6. Thus, a single
dose of 100 µg of rmTSG-6 (given together with the mBSA) blocked the
degradation and loss of aggrecan from cartilage (Figure 5
; M to O).
However, this chondroprotective effect of TSG-6 in acutely inflamed
joints gradually decreased after day 7 and essentially no differences
were found between PBS/mBSA and rmTSG-6/mBSA-injected knee joints after
day 12, unless additional doses of TSG-6 were introduced locally or
systematically (results are not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IL-1),
growth factors, and various compounds (including
lipopolysaccharides),2,3,5,28
components that in turn
activate various transcription factors (nuclear factor-
B, nuclear
factor-IL6, or AP-1) and induce the TSG-6 expression in various cell
types.29
In a complex milieu such as the inflamed
rheumatoid synovium, among many other components, TSG-6, HA, and CD44
are up-regulated simultaneously, and increased serum levels of these
compounds correlate directly with the disease activity of rheumatoid
arthritis.3,8,30-33
To study the
pathophysiological relevance of TSG-6 in arthritic processes, and gain
insight into the function of this cytokine-induced molecule in
inflammatory conditions, we used an autoimmune murine model (PGIA) of
rheumatoid arthritis. The intraperitoneal injection of rmTSG-6 had no effect on joint swelling of mice with acute PGIA, whereas the intravenous injection of rmTSG-6 resulted in a dramatic reduction in joint swelling. The lack of any effect of intraperitoneally injected rmTSG-6 is probably because of the entrapment of rmTSG-6 by the excess amount of HA in the peritoneal cavity occupied by adjuvant-induced granuloma, which is the consequence of intraperitoneal immunization.34 Indeed, we were unable to detect radiolabeled rmTSG-6 in serum after intraperitoneal injection, whereas the intravenously injected rmTSG-6 preferentially accumulated at the site of inflammation. In this kinetic study, the half life of intravenously injected rmTSG-6 was less than 4 hours, but the serum level of TSG-6 measured by ELISA was almost constant 60 to 90 minutes after the intravenous injection (unpublished data). Thus, although the exogenous TSG-6 (eg, radiolabeled) reached the inflamed joint soon after the intravenous injection, and was detectable in the inflamed joint at least 24 hours, a significant amount of TSG-6 remained in, or returned to, the circulation.
The loss of edema in the presence of massive cellular infiltration
indicated that TSG-6 affects cell/tissue-bound HA, rather than the
migration of inflammatory cells or the proliferation of synovial cells.
The overnight gap in treatment resulted in a plateau in reduction of
joint swelling which then further declined after repeated
administration of TSG-6 protein (Figure 1A)
. Therefore, we hypothesized
that a relatively high dose of rmTSG-6 can competitively occupy the
binding sites on HA and displace CD44 in the inflamed tissue. As a
result, the CD44-bound HA (and HA-bound water) became diffusible and
receded from the tissue. Using in vitro assays with the
three different cell lines, SVEC4-10 endothelial cells, mouse synovial
fibroblast, and 5/4 T-cell hybridoma, we demonstrated that TSG-6 can
indeed compete with CD44 in a dose-dependent manner. This was further
confirmed in vivo, when increasing doses of HA-binding
peptide Pep-119
reciprocally reduced the effect of rmTSG-6
in short-term experiments. However, although this HA-binding synthetic
peptide Pep-1 competed effectively with TSG-6 in vivo, the
peptide by itself had no effect on joint swelling. This seemingly
contradictory observation indicates that Pep-1 was able to occupy free
binding sites on HA, but it was unable to displace CD44 bound to HA.
The most unexpected observation after noted excessive reduction in
joint swelling in a short-term treatment was the undistinguishable
histopathological picture in PBS- and TSG-6-treated animals. Although
the reduction of edema was evident within hours (Figure 1A)
, and then
visible in paw sections of TSG-6-treated animals (Figure 2B)
, there
were no evident differences in cellular infiltration, synovial lining
cell proliferation, nor pannus formation for up to 4 weeks of the
experimental period. This comparable picture of synovitis was
especially in contrast to the lack of cartilage and bone erosions in
TSG-6-treated animals, which otherwise is a typical consequence of
joint inflammation in RA and in all corresponding animal
models.15,16,20,21,25,35
First, we performed experiments in PGIA. We preferentially use this mouse model for genetic, inflammation, cell migration, and autoimmune studies, rather than CIA, another widely used systemic autoimmune model for human RA. Essentially both inflammatory models show many similarities to RA as indicated by clinical assessment of arthritis and histopathology of the peripheral joints. However, beyond similarities, there are substantial differences between the two autoimmune inflammatory animal models for RA: PGIA shows higher susceptibility (100% in aging female mice),15-17 the joint inflammation is more progressive leading to complete deterioration of articular cartilage accompanied with massive bone erosions,15,16,25 the onset of the disease is dictated by a shift from Th2 to Th1 response,26,36 and it has a recessive inheritance.24,37 Many of these characteristics are comparable to those in rheumatoid arthritis. In contrast to a recently published study for TSG-6 treatment of CIA,13 we were unable to find any significant differences in pathophysiological markers (including the serum level of IL-6) between PBS- and TSG-6-treated animals, synovial cell proliferation, or in cell-surface expression of various phenotypic markers. Fluctuation and differences in serum levels of various cytokines clearly associated with the disease state in both PGIA and CIA, rather than the TSG-6 treatment. We could detect the same, occasionally more prominent, differences in serum markers in leflunomide-,38 IL-4-,36,39,40 and IL-10-treated39-41 animals, or during anti-CD44 immunotherapy,14,17 as found in TSG-6-treated arthritic animals. Although the TSG-6 treatment reduced the severity (edema and cartilage damage) of arthritis, it did not reduce the incidence or delay the onset of arthritis. Therefore, the anti-inflammatory effect of exogenous (human) recombinant TSG-6 associated with increased serum levels of IL-6 measured at a single and late time point in the above-mentioned study13 may coincide with, but can hardly be the consequence of, the TSG-6 treatment. Rather, it might be the consequence of immunization of DBA/1 mice with a large dose of human TSG-6, or the lipopolysaccharide or IL-1 injection might be used for synchronization of CIA to reach a higher incidence. Although IL-6 locally may stimulate PG synthesis,35 IL-6 is a key proinflammatory cytokine required for the development of various forms of arthritis, including CIA.24,25,42-45 In conclusion, it seems to be very likely that a significantly reduced incidence and delayed onset of CIA,13 without suppression of proinflammatory cytokines, antibodies, or the Th1 response, should not be related to increased levels of IL-6 in TSG-6-treated animals.
To explore the role of TSG-6 during joint inflammation and understand
how the morphologically intact cartilage can be preserved in the
presence of massive synovial inflammation we extended our studies and
used the AIA model. In this model, cartilage damage can be assessed by
loss of safranin-O-stained proteoglycan and, simultaneously, we can
demonstrate the most characteristic, MMP-generated neoepitopes of
cartilage PG (aggrecan), such as the aggrecanase-cleaved -NITEGE and
various MMP-generated -VDIPEN sequences.20,21,23
A single
dose of rmTSG-6 administered simultaneously with the antigen challenge
into the knee joints showed protection against the loss of cartilage
PG. Therefore, the chondroprotective effect of TSG-6 directly, or
indirectly, associated with the inhibition of matrix
metalloproteinases, enzymes that are also up-regulated by cytokines in
inflammatory conditions. It is known that TSG-6 protein can bind to
II, dramatically increasing the serine protease inhibitory effect of
the II.8,9,11
The protease inhibitory effect of
II/TSG-6 complex was especially evident on plasminogen-plasmin
activation, probably the most potent serine protease needed for
activation of pro-MMPs.46
Thus, TSG-6 seems to be a key
component of a negative feedback loop controlling extensive tissue
damage in inflammatory conditions.
A conclusive evidence that the TSG-6/II complex mediates the potent
antiplasmin activity will require additional studies in all arthritis
models (CIA, PGIA, and AIA), and the isolation of the 120-kd
complex.8,11
Although the serine protease inhibitory
activity of II family members via their common bikunin domain has
been known for a long time,47
little is known about their
functions. However, disease-associated presence in various tissues and
fluctuations seen in the serum levels of II and II-related
protein suggest an involvement in pathological processes. Daveau and
colleagues48
reported a distinct pattern of changes in
serum concentrations of the different members of the II family
during acute inflammation. Proteins identical with, or closely related
to, the bikunin chain of II, have been detected in the stroma and
the surrounding connective tissue of malignant tumors,49
in the brain tissue of patients with Alzheimer’s
disease,50
and in the serum and urine of patients with
inflammatory diseases, cancer, and leukemia.51,52
A link
between arthritis and II was suggested more than 20 years ago when
Becker and Sandson53
found II associated with HA in the
synovial fluid of patients with RA. This finding was confirmed to show
that II associates in vitro with HA isolated from the
synovial fluid of healthy patients.54
In conclusion, TSG-6
stimulated by proinflammatory cytokines may function as an
anti-inflammatory component by combining and enhancing the serine
protease inhibitory activity of II.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grants from the National Institutes of Health (AR40310, AR45652, and AR47135) and the Coleman Foundation (Chicago, IL).
R. V. K. and T. B. were equally involved in this study and both regarded as first author.
Accepted for publication July 23, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-inhibitor. Biochemistry 1994, 33:7423-7429[Medline]
-inhibitor and exerts a strong anti-inflammatory effect in vivo. J Immunol 1996, 156:1609-1615[Abstract]
-trypsin inhibitor proteoglycan family—a group of proteins binding and stabilizing the extracellular matrix. Eur J Biochem 1998, 252:339-346[Medline]
, and interleukin-6 in cartilage proteoglycan metabolism and destruction. Effect in situ blocking in murine antigen- and zymosan-induced arthritis. Arthritis Rheum 1995, 38:164-172[Medline]
-trypsin inhibitor, and 1 proteinase inhibitor to synovial fluid hyaluronate and the influence of these proteins on its degradation by oxygen derived free radicals. Ann Rheum Dis 1988, 47:377-385This article has been cited by other articles:
 |
|
 |
|
  V. Maina, A. Cotena, A. Doni, M. Nebuloni, F. Pasqualini, C. M. Milner, A. J. Day, A. Mantovani, and C. Garlanda Coregulation in human leukocytes of the long pentraxin PTX3 and TSG-6 J. Leukoc. Biol., July 1, 2009; 86(1): 123 - 132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Colon, A. Shytuhina, M. K. Cowman, P. A. Band, K. W. Sanggaard, J. J. Enghild, and H.-G. Wisniewski Transfer of Inter-{alpha}-inhibitor Heavy Chains to Hyaluronan by Surface-linked Hyaluronan-TSG-6 Complexes J. Biol. Chem., January 23, 2009; 284(4): 2320 - 2331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. W. Sanggaard, C. S. Sonne-Schmidt, T. P. Krogager, T. Kristensen, H.-G. Wisniewski, I. B. Thogersen, and J. J. Enghild TSG-6 Transfers Proteins between Glycosaminoglycans via a Ser28-mediated Covalent Catalytic Mechanism J. Biol. Chem., December 5, 2008; 283(49): 33919 - 33926. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Mahoney, K. Mikecz, T. Ali, G. Mabilleau, D. Benayahu, A. Plaas, C. M. Milner, A. J. Day, and A. Sabokbar TSG-6 Regulates Bone Remodeling through Inhibition of Osteoblastogenesis and Osteoclast Activation J. Biol. Chem., September 19, 2008; 283(38): 25952 - 25962. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Garcia, H.-G. Wisniewski, M. S. Lucia, N. Arevalo, T. J. Slaga, S. L. Kraft, R. Strange, and A. P. Kumar 2-Methoxyestradiol Inhibits Prostate Tumor Development in Transgenic Adenocarcinoma of Mouse Prostate: Role of Tumor Necrosis Factor-{alpha}-Stimulated Gene 6 Clin. Cancer Res., February 1, 2006; 12(3): 980 - 988. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Kuznetsova, A. J. Day, D. J. Mahoney, M. S. Rugg, D. F. Mosher, and D. D. Roberts The N-terminal Module of Thrombospondin-1 Interacts with the Link Domain of TSG-6 and Enhances Its Covalent Association with the Heavy Chains of Inter-{alpha}-trypsin Inhibitor J. Biol. Chem., September 2, 2005; 280(35): 30899 - 30908. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Mahoney, B. Mulloy, M. J. Forster, C. D. Blundell, E. Fries, C. M Milner, and A. J. Day Characterization of the Interaction between Tumor Necrosis Factor-stimulated Gene-6 and Heparin: IMPLICATIONS FOR THE INHIBITION OF PLASMIN IN EXTRACELLULAR MATRIX MICROENVIRONMENTS J. Biol. Chem., July 22, 2005; 280(29): 27044 - 27055. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Bardos, Z Szabo, M Czipri, C Vermes, M Tunyogi-Csapo, R M Urban, K Mikecz, and T T Glant A longitudinal study on an autoimmune murine model of ankylosing spondylitis Ann Rheum Dis, July 1, 2005; 64(7): 981 - 987. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-G. Wisniewski, E. S. Snitkin, C. Mindrescu, M. H. Sweet, and J. Vilcek TSG-6 Protein Binding to Glycosaminoglycans: FORMATION OF STABLE COMPLEXES WITH HYALURONAN AND BINDING TO CHONDROITIN SULFATES J. Biol. Chem., April 15, 2005; 280(15): 14476 - 14484. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lesley, I. Gal, D. J. Mahoney, M. R. Cordell, M. S. Rugg, R. Hyman, A. J. Day, and K. Mikecz TSG-6 Modulates the Interaction between Hyaluronan and Cell Surface CD44 J. Biol. Chem., June 11, 2004; 279(24): 25745 - 25754. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Szanto, I. Gal, A. Gonda, T. T. Glant, and K. Mikecz Expression of L-Selectin, but Not CD44, Is Required for Early Neutrophil Extravasation in Antigen-Induced Arthritis J. Immunol., June 1, 2004; 172(11): 6723 - 6734. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Milner and A. J. Day TSG-6: a multifunctional protein associated with inflammation J. Cell Sci., May 15, 2003; 116(10): 1863 - 1873. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Fulop, S. Szanto, D. Mukhopadhyay, T. Bardos, R. V. Kamath, M. S. Rugg, A. J. Day, A. Salustri, V. C. Hascall, T. T. Glant, et al. Impaired cumulus mucification and female sterility in tumor necrosis factor-induced protein-6 deficient mice Development, May 15, 2003; 130(10): 2253 - 2261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Ochsner, D. L. Russell, A. J. Day, R. M. Breyer, and J. S. Richards Decreased Expression of Tumor Necrosis Factor-{alpha}-Stimulated Gene 6 in Cumulus Cells of the Cyclooxygenase-2 and EP2 Null Mice Endocrinology, March 1, 2003; 144(3): 1008 - 1019. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L. Espey and J. S. Richards Temporal and Spatial Patterns of Ovarian Gene Transcription Following an Ovulatory Dose of Gonadotropin in the Rat Biol Reprod, December 1, 2002; 67(6): 1662 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lesley, N. M. English, I. Gal, K. Mikecz, A. J. Day, and R. Hyman Hyaluronan Binding Properties of a CD44 Chimera Containing the Link Module of TSG-6 J. Biol. Chem., July 12, 2002; 277(29): 26600 - 26608. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. A. Nentwich, Z. Mustafa, M. S. Rugg, B. D. Marsden, M. R. Cordell, D. J. Mahoney, S. C. Jenkins, B. Dowling, E. Fries, C. M. Milner, et al. A Novel Allelic Variant of the Human TSG-6 Gene Encoding an Amino Acid Difference in the CUB Module. CHROMOSOMAL LOCALIZATION, FREQUENCY ANALYSIS, MODELING, AND EXPRESSION J. Biol. Chem., May 3, 2002; 277(18): 15354 - 15362. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
