вход по аккаунту


Stromal cells and osteoclasts are responsible for exacerbated collagen-induced arthritis in interferon- deficient mice.

код для вставкиСкачать
Vol. 52, No. 12, December 2005, pp 3739–3748
DOI 10.1002/art.21496
© 2005, American College of Rheumatology
Stromal Cells and Osteoclasts Are Responsible for Exacerbated
Collagen-Induced Arthritis in Interferon-␤–Deficient Mice
Alexandra P. Treschow, Ingrid Teige, Kutty S. Nandakumar, Rikard Holmdahl,
and Shohreh Issazadeh-Navikas
Objective. Clinical trials using interferon-␤
(IFN␤) in the treatment of rheumatoid arthritis have
shown conflicting results. We undertook this study to
understand the mechanisms of IFN␤ in arthritis at a
physiologic level.
Methods. Collagen-induced arthritis (CIA) was
induced in IFN␤-deficient and control mice. The role of
IFN␤ was investigated in both the priming and effector
phases of the disease. The effect of IFN␤ deficiency on
synovial cells, macrophages, and fibroblasts from preimmunized mice was analyzed by flow cytometry, immunohistochemistry, and enzyme-linked immunosorbent
assay. Differences in osteoclast maturation were determined in situ by histology of arthritic and naive paws
and by in vitro maturation studies of naive bone marrow
cells. The importance of IFN␤-producing fibroblasts
was determined by transfering fibroblasts into mice at
the time of CIA immunization.
Results. Mice lacking IFN␤ had a prolonged
disease with a higher incidence compared with control
mice. IFN␤ deficiency was found to influence the effector phase, but not the priming phase, of arthritis.
Compared with control mice, IFN␤-deficient mice had
greater infiltration of CD11bⴙ cells and greater production of tumor necrosis factor ␣ in vivo, and their
macrophages and fibroblasts were both more activated
in vitro. Moreover, IFN␤-deficient mice generated a
greater number of osteoclasts in vitro, and mice immunized to induce arthritis, but not naive mice, had a
greater number of osteoclasts in vivo compared with
control mice. Importantly, IFN␤-competent fibroblasts
were able to ameliorate arthritis in IFN␤-deficient
Conclusion. Our data indicate that IFN␤ is involved in regulating the activation state of osteoclasts
and stromal cells, including macrophages and fibroblasts, but that it has little effect on T cells.
Interferons (IFNs) are potent cytokines that are
classified as either type I or type II. IFN␣ and IFN␤ are
type I IFNs and are considered to be antiinflammatory.
Type I IFNs bind to the same receptor complex, which
consists of 2 transmembrane proteins (1,2). The binding
of IFN␣ and IFN␤ to the receptor has been shown to
have distinct biologic functions (3–6).
Apart from its antiviral properties, IFN␤ has
been shown to have a wide variety of developmental and
immunomodulating effects. IFN␤ has been shown to be
involved in the development of B cells, neutrophils, and
osteoclasts as well as in inhibition of apoptosis of
leukocytes (7–10). However, the immunomodulating
effects of IFN␤ have been of greatest interest in terms of
its therapeutic use. IFN␤ has been shown to downregulate proinflammatory cytokines, such as tumor necrosis factor ␣ (TNF␣) and interleukin-1␤ (IL-1␤), and
to increase the secretion of antiinflammatory mediators,
such as IL-10 and IL-1 receptor antagonist (11). IFN␤
has also been implicated in reducing T cell proliferation
as well as in down-regulation of class II major histocompatibility complex (MHC) on antigen-presenting cells
(APCs) (12,13). Due to its antiinflammatory properties,
this cytokine has been studied and used in various
human immune disorders such as multiple sclerosis
(MS) and cancer (14,15). There has been a positive
Supported by the Swedish Research Council–Natural Science,
the Swedish Research Council–Medicine, The Swedish Rheumatism
Association, the Alfred Österlund Foundation, the Tore Nilson Foundation, King Gustaf V’s 80-Year Foundation, the Royal Physiographic
Society in Lund, the M. Bergvalls Foundation, the Åke Wiberg Foundation, the Börje Dahlin Foundation, and the Crafoord Foundation.
Alexandra P. Treschow, MSc, Ingrid Teige, PhD, Kutty S.
Nandakumar, PhD, Rikard Holmdahl, MD, PhD, Shohreh IssazadehNavikas, PhD: University of Lund, Lund, Sweden.
Address correspondence and reprint requests to Alexandra P.
Treschow, MSc, Section for Medical Inflammation Research, Institute
for Cell and Molecular Biology, University of Lund, BMC I11, S-22184
Lund, Sweden. E-mail:
Submitted for publication April 25, 2005; accepted in revised
form September 13, 2005.
response with IFN␤ treatment of MS patients, and it is
currently the most effective treatment used for this
autoimmune disease.
In recent years, there has been an interest in
determining the possible beneficial effects of IFN␤ in
rheumatoid arthritis (RA). IFN␤ has been investigated
in mice and monkeys with promising results (15,16), and
IFN␤ has been shown to regulate osteoclastogenesis in
mice (9,13,17). Moreover, in a small clinical trial, administration of IFN␤ led to a significant reduction in the
expression of IL-1␤, matrix metalloproteinase 1 (MMP1), and tissue inhibitor of metalloproteinases 1 in the
synovial lining and also to a reduction in CD3⫹ T cell
infiltration (18). That study also showed that in vitro RA
fibroblast-like synoviocytes also had decreased expression of MMP-1 when treated with IFN␤. Together, these
findings indicate that IFN␤ may have a protective effect
against joint destruction. In line with this, a recent report
showed that IFN␤ is highly expressed in the synovium of
RA patients compared with that of patients with osteoarthritis or reactive arthritis (19).
Despite the early success of IFN␤ treatment in
animal models as well as in small clinical trials, recent
clinical studies have shown limited effect (13,18,20,21).
These inconsistent results may, however, have several
explanations. First, animal models of RA differ from
human RA in several aspects, such as in the level of T
cell infiltration into arthritic joints. Second, treatment
protocols used in animal models of RA, including administration of very high amounts of IFN␤ and administration of retrovirus or transformed fibroblasts that
produce IFN␤, are difficult to extrapolate to humans.
Consequently, there is a need to gain an enhanced
understanding of the mechanism of action of IFN␤ at a
physiologic level. The present study utilized IFN␤deficient (IFN␤⫺/⫺) mice in comparison with control
heterozygous (IFN␤⫺/⫹) mice in order to determine the
effect of IFN␤ in collagen-induced arthritis (CIA), the
murine model for RA.
CIA is one of the most commonly used animal
models for RA, and it is induced in mice by injecting
heterologous type II collagen (CII) in an adjuvant,
leading to a disease resembling RA (22). T cells have
been shown to play an important role in the pathogenesis of CIA (23), possibly by the production of proinflammatory cytokines and by providing help to B cells.
Although CIA is T cell dependent, T cells are primarily
involved in the priming phase of the disease, whereas the
effector phase is driven by B cells producing anti-CII
antibodies that crossreact with mouse CII (24). Histologic changes associated with CIA involve the infiltration
of neutrophils, macrophages, and T cells into the synovia. There is also pannus formation and activation of
stromal cells, such as fibroblasts and macrophages.
These histologic changes are all believed to contribute to
the pathogenesis of CIA.
In the present study, we used IFN␤⫺/⫺ mice in
order to determine the role of IFN␤ in the CIA model.
We found that while the T cell compartment appeared
unaffected by IFN␤ deficiency, fibroblasts, macrophages, and osteoclasts located in the joints were more
activated compared with those in IFN␤⫺/⫹ mice. Therefore, we postulate that IFN␤ deficiency leads to severe
arthritis in which stromal cells and osteoclasts perpetuate the disease. Pinpointing the mechanism of IFN␤ is of
importance in determining whether IFN␤ can be used as
a treatment for arthritis and, if so, which patient group
would benefit most from this treatment.
Mice. The generation of IFN␤-deficient mice has been
described previously (6). The mice were screened for the
deletion of IFN␤ by polymerase chain reaction from tissue (tail
or toe of mice) as previously described (25). Mice were
backcrossed to the B10.RIII strain for 7 and 12 generations by
crossing IFN␤⫺/⫹ mice with B10.RIII mice. Mice were bred
and kept at the conventional animal facility at the Section for
Medical Inflammation Research, University of Lund, and all
experiments had animal ethics committee approval. Unless
stated otherwise, 8–16-week-old male mice were used.
Antigens. CII was prepared from calf cartilage by
pepsin digestion as described previously (26). Peptides were
synthesized as described previously (27). CII was denatured by
heating at 65°C for 20 minutes before use in in vitro proliferation assays.
Immunization, scoring, and anti-CII antibody enzymelinked immunosorbent assay (ELISA). For arthritis experiments, mice (8–14 per group) were immunized intradermally
in the base of the tail with 100 ␮g CII emulsified 1:1 in
Freund’s incomplete adjuvant (IFA; Difco, Detroit, MI). For
in vitro lymphocyte assays, mice were immunized in the base of
the tail and in each hind footpad with 60 ␮g of CII in IFA at
each location. Arthritis was evaluated by visual scoring using
an extended scoring protocol (28); scores ranged from 1 to 15
for each paw with a maximum score of 60 per mouse. Each
arthritic toe and knuckle was scored as 1, with a maximum of
10 per paw. An arthritic ankle or midpaw was given a score of
5. The anti-CII antibody response was determined by measuring the level of CII-specific antibodies in serum collected 121
days postimmunization. The amounts of total anti-CII IgG as
well as the amounts of the IgG1 and IgG2a isotypes were
determined through quantitative ELISA as previously described (29).
Collagen antibody–induced arthritis (CAIA). Arthritis
in 4-month-old mice (9–12 per group) was induced by injecting
intravenously an anti-CII monoclonal antibody cocktail of
CIIC1 and M2139 (9 mg/mouse) as previously described (30)
without a lipopolysaccharide (LPS) booster. After 48 hours,
clinical signs of arthritis were observed, and the arthritis was
monitored for 72 days using the scoring protocol described
Proliferation and cytokine production assays. Ten
days after immunization, cells from the draining inguinal and
popliteal lymph nodes were prepared and restimulated in vitro
in order to determine antigen-specific cell proliferation and
the IFN␥ response as described previously (31,32). Six to 10
mice per group were used.
For determination of anti-CD3 T cell responses, spleen
cells from naive or immunized (10 days prior) mice were
seeded at a concentration of 1 ⫻ 106/well in plates precoated
with anti-CD3 (clone 145-2C11; from our hybridoma collection) and incubated for 48 hours before pulsing with 1 ␮Ci
H-thymidine (Amersham International, Amsterdam, The
Netherlands) for 15–18 hours. Five mice per group were used.
Macrophage preparation and culture. Spleens were
removed 10 days postimmunization, and macrophages were
enriched and stimulated as previously described (25); 8–10
mice per group were used. Supernatants were collected after
36 hours of incubation and assayed for cytokine content using
ELISA. The production of TNF␣ was determined using the
recommended paired antibodies and the protocol of BD
PharMingen (Franklin Lakes, NJ). The plates were read using
a fluorometer (Wallac, Boston, MA).
Synovia preparation and culture. Mice (24–25 per
group) were immunized to induce arthritis as described above.
Thirty days postimmunization, mice were killed, hind legs were
removed, and the synovia of the knees were dissected out,
pooled, and placed in 1.6 mg/ml type IV collagenase (Worthington, Lakewood, NJ) and 0.1% DNase I (Sigma-Aldrich,
St. Louis, MO) in Dulbecco’s modified Eagle’s medium
(DMEM) and incubated for 1 hour at 37°C. The cells were left
untreated or first primed with 10 units/ml of IFN␥ for 60
minutes, then incubated for 36 hours with 50 ng/ml of LPS.
The expression of surface markers on the synovial cells
was determined by flow cytometry using the following conjugated antibodies: fluorescein isothiocyanate (FITC)–
conjugated anti–intercellular adhesion molecule (anti-ICAM)
(clone 3E2; BD PharMingen), biotinylated anti–vascular cell
adhesion molecule (anti-VCAM) (clone 429; BD PharMingen), phycoerythrin (PE)–conjugated anti-CD11b (clone
M1/70; BD PharMingen), allophycocyanin-conjugated anti–
Ly6-G (clone Rb6-8C5; BD PharMingen), FITC-conjugated
anti–class II MHC (clone Y3P; from our hybridoma collection), and biotinylated anti–macrophage F4/80 antigen (clone
F4/80; from our hybridoma collection). In order to determine
TNF␣ production, monosine (3 ␮M/ml; Sigma-Aldrich) was
added 6 hours prior to staining. Intracellular staining was then
performed using BD Cytofix/Cytoperm solution and protocol
(Becton Dickinson, Franklin Lakes, NJ) using an unconjugated
anti-TNF␣ antibody (clone XT22; BD PharMingen) followed
by a biotinylated secondary goat anti-rat antibody (Jackson
ImmunoResearch, West Grove, PA).
Immunohistochemistry. IFN␤⫺/⫺ and IFN␤⫺/⫹ mice
(5 per group) were immunized to induce arthritis; on day 40,
mice were killed, and paws were dissected and decalcified with
EDTA (for 2–3 weeks). The paws were then embedded in
OCT compound (Sakura Finetek Europe, Zoeterwoude, The
Netherlands) and snap-frozen in isopentane on dry ice. Staining of slides was performed as previously described (25)
Diaminobenzidine (50 mg/ml; Saveen Biotech, Ideon, Sweden)
was used for detection, and slides were counterstained with
hematoxylin. In all studies, the numbers of positive cells were
determined in a blinded manner by calculating the mean count
of 5 distinct areas per section.
Fibroblast preparation and culture. Fibroblasts were
prepared from IFN␤⫺/⫺ and IFN␤⫺/⫹ mice that had shown
clinical signs of arthritis for at least 1 week. The fibroblasts
were prepared by removing the skin and muscle from the hind
legs (8–10 mice per group) and grinding them in a mortar in a
solution of 0.25% trypsin in phosphate buffered saline (PBS).
Cells were then incubated for 30 minutes at 37°C, washed with
DMEM containing 10% fetal calf serum (FCS), and incubated
for an additional 90 minutes in 0.1% collagenase. The fibroblasts were subjected to a minimum of 6 passages (detachment
by 0.5% trypsin in 5 mM EDTA) to obtain a pure culture. To
analyze the phenotype of the fibroblasts, cells (1 ⫻ 104/well)
were seeded into 48-well plates in DMEM containing 10%
FCS and cultured for 4 days before being detached with cell
dissociation media (Sigma-Aldrich). The IL-6 content in the
supernatant of the cultured fibroblasts was determined using
the recommended paired antibodies and protocol of BD
PharMingen. The expression of cell surface markers on fibroblasts was determined by flow cytometry using the following
antibodies: FITC-conjugated anti-ICAM, biotinylated antiVCAM, FITC-conjugated anti–class II MHC, PE-conjugated
anti-CD40 (clone 3/23; BD PharMingen), biotinylated antiCD44 (clone IM7.8.1; BD PharMingen), PE-conjugated antiCD71 (clone C2; BD PharMingen), and biotinylated anti–
IFN␥ receptor ␣-chain (clone GR20; BD PharMingen).
Transfer of fibroblasts. Fibroblasts were detached
from culture bottles using EDTA/trypsin and washed with
PBS, and a single-cell suspension was obtained by passing the
fibroblasts through a 23G needle. The fibroblasts were then
injected periarticularly into the joints of mice (total of 2 ⫻ 106
fibroblasts/mouse) at 6 injection sites in the metacarpal, metatarsal, and ankle joints. At the same time, the mice (8–10 per
group) were immunized to induce CIA as described earlier.
In vitro generation of osteoclasts. Bone marrow cells
were obtained from the tibias of 4 IFN␤⫺/⫺ and 4 IFN␤⫺/⫹
mice by removing the bone ends and flushing with ␣-minimum
essential medium (Gibco BRL Life Technologies, Gaithersburg, MD). Nonadherent cells were washed, 2.5 ⫻ 105/well
were seeded into a 48-well plate, and 10 ng/ml macrophage
colony-stimulating factor (M-CSF; R&D Systems, Minneapolis, MN) was added. Three days later, media were removed,
fresh media containing M-CSF (10 ng/ml) plus recombinant
murine RANKL (100 ng/ml; PeproTech, London, UK) were
added, and cells were cultured for an additional 3–4 days.
Bone marrow cells incubated with M-CSF alone were used as
negative control. The osteoclasts were visualized using
tartrate-resistant acid phosphatase (TRAP) staining according
to Becton Dickinson Technical Bulletin no. 445 (“Tartrate
Resistant Acid Phosphatase staining of osteoclasts”). Cells
were counterstained with hematoxylin. Osteoclasts were classified as multinucleated and TRAP positive. In all studies, the
numbers of positive cells were determined by calculating the
mean count from 5 fields of view per well.
In situ determination of osteoclasts using TRAP staining. Naive IFN␤⫺/⫺ and IFN␤⫺/⫹ mice ages 8–16 weeks or
⬎1.5 years as well as preimmunized (40 days prior) IFN␤⫺/⫺
and IFN␤⫺/⫹ mice (4–6 per group) were killed and paws were
dissected. The paws were fixed in 4% phosphate buffered
formaldehyde for 24 hours at 4°C, decalcified with EDTA (for
2–3 weeks), embedded in paraffin, and sectioned at a thickness
of 5 ␮m. The sections were rehydrated and stained for TRAP
as described above. All joints in the section were counted, and
joints that contained ⱖ1 osteoclast were counted as affected.
Thereafter, the number of affected joints per total number of
counted joints was determined individually in order to compare the 2 groups of mice.
Statistical analysis. The frequency of arthritis was
analyzed by chi-square test. The Mann-Whitney U test was
used in all other statistical analyses.
Exacerbation of CIA in IFN␤-deficient mice in
the chronic phase of the disease. There has recently
been an interest in addressing whether IFN␤ has an
ameliorating effect on arthritis, but the results of these
investigations have been conflicting. We therefore decided to investigate the effect of IFN␤ deficiency on
arthritis in the CIA mouse model. IFN␤⫺/⫺ mice were
backcrossed to the B10.RIII background for 7 generations and then investigated for arthritis susceptibility.
There was no difference in the incidence, day of onset,
or severity of arthritis between mice heterozygous for
IFN␤ deficiency (IFN␤⫺/⫹ mice) and IFN␤ wild-type
littermates (IFN␤⫹/⫹ mice) (data not shown); therefore,
both groups were pooled for subsequent comparison
with the group of IFN␤⫺/⫺ mice. Although there was no
difference in day of onset, IFN␤⫺/⫺ mice were more
susceptible to CIA and developed an exacerbated disease compared with control mice. Relapses of arthritis
were also observed in IFN␤⫺/⫺ mice, and they had a
tendency toward a higher anti-CII antibody response
(Figure 1A and Table 1). A similar disease profile was
also observed when mice that had been backcrossed for
12 generations were used in the CIA model (Table 1).
T cell response to CII not affected by a lack of
endogenous IFN␤. Since the CIA model is T cell dependent, it was feasible that the IFN␤⫺/⫺ mice had an
exacerbated arthritis due to an increased T cell response
to CII. Furthermore, it has been previously reported that
IFN␤ affects the proliferative response of T cells (8,33).
Concordant with a recent report (8), naive IFN␤⫺/⫺
spleen cells (and lymph node cells [data not shown])
were found to have a significantly greater proliferative
response than those of control mice when stimulated
with anti-CD3 (Figure 1B). To evaluate whether
Figure 1. Interferon-␤ (IFN␤) deficiency leads to augmented
collagen-induced arthritis without altering the antigen-specific T cell
response. A, Arthritis index, calculated as the mean arthritis score in
IFN␤-deficient mice and control (IFN␤⫺/⫹ and IFN␤⫹/⫹) mice. Mice
were scored twice weekly, starting from day 14. Data include a total of
8–14 mice per group. ⴱ ⫽ P ⱕ 0.05; ⴱⴱ ⫽ P ⱕ 0.01 versus control mice.
B, Anti-CD3 (␣CD3) response of splenocytes from naive IFN␤⫺/⫺
mice and control IFN␤⫺/⫹ mice. ⴱ ⫽ P ⱕ 0.05 versus IFN␤⫺/⫺ mice.
C, Anti-CD3 and bovine type II collagen (bCII) response of splenocytes from preimmunized IFN␤⫺/⫺ mice and control IFN␤⫺/⫹ mice.
Cells in B and C were incubated for 48 hours before measurement of
H-thymidine (3H-TdR) incorporation. D, Antigen-specific response
of lymphocytes from IFN␤⫺/⫺ and control IFN␤⫺/⫹ mice immunized
10 days prior to in vitro cultivation. Lymphocytes were restimulated
with 50 ␮g/ml of whole bCII, with 50 ␮g/ml or 10 ␮g/ml of the
immunodominant peptide sequence 607–621 of CII (p607), or with 50
␮g/ml of the immunodominant peptide sequence 442–456 of CII
(p442). Cells were cultivated for 72 hours before measurement of
H-TdR incorporation. Values in B–D are the mean ⫾ SD. Results
shown are from 1 of 3 representative experiments. SI ⫽ stimulation
IFN␤⫺/⫺ mice also had an increased antigen-specific
proliferative response, mice were immunized with CII
and cells were subsequently restimulated in vitro with
bovine CII and the immunodominant peptides (607–621
Table 1. Arthritis parameters and anti-CII antibody response in IFN␤⫺/⫺ and control (IFN␤⫺/⫹ and
IFN␤⫹/⫹) mice*
IgG total, units/ml†
IgG1, units/ml
IgG2, units/ml
Incidence, %‡
Maximum score, 0–60¶
Incidence in 12th generation, %‡
CAIA average score, 0–60#
CAIA incidence, %#
CAIA duration¶#
IFN␤⫺/⫺ mice
Control mice
56 ⫾ 11
33 ⫾ 5
88 ⫾ 54
63 ⫾ 6
23 ⫾ 3
10.2 ⫾ 2.1
25.9 ⫾ 1
37 ⫾ 8
20 ⫾ 3
26 ⫾ 5
27 ⫾ 6
12 ⫾ 2
4.2 ⫾ 1.7
19.8 ⫾ 2.1
* Except where indicated otherwise, values are the mean ⫾ SD. CII ⫽ type II collagen; IFN␤⫺/⫺ ⫽
interferon-␤ deficient; CAIA ⫽ collagen antibody–induced arthritis.
† Mean total CII IgG as well as IgG subclass levels in sera collected on day 121 were calculated as arbitrary
units/ml using a polyclonal serum.
‡ Calculated from day 65 after immunization.
§ The average duration of disease was calculated from the first day of evident clinical signs until the end
of the experiment (unaffected mice were assigned a value of 0).
¶ The maximum score of affected mice was calculated as the average of the highest scores obtained in mice
showing clinical signs of arthritis.
# Calculated on day 21 after transfer of CII-specific antibodies.
and 442–456). However, there was no significant difference between the IFN␤⫺/⫺ and control mice in their
proliferative responses (Figures 1C and D) or in production of IFN␥ (data not shown). Similar results were also
observed in mice immunized 3 weeks prior (data not
shown). Furthermore, the detected difference in antiCD3 stimulation in naive mice (Figure 1B) was not
observed in immunized mice (Figure 1C).
Effect of lack of IFN␤ on the effector phase of the
disease. Data so far indicated that IFN␤ deficiency did
not have an effect in the priming phase of CIA. Instead,
the exacerbated disease of IFN␤⫺/⫺ mice may be explained by events occurring in the effector phase. The
effector phase of CIA is mediated via arthritogenic
anti-CII antibodies, which can be mimicked by using the
acute and T cell–independent CAIA model (34).
IFN␤⫺/⫺ and control mice were subjected to passive
transfer of the disease using 2 collagen-specific monoclonal antibodies. As shown in Table 1, IFN␤⫺/⫺ mice
were indeed found to have developed a more severe and
prolonged disease compared with that in control mice,
indicating that IFN␤ operates in the inflammatory phase
in the joints.
Augmented activation in vitro of peripheral and
synovial APCs in IFN␤-deficient mice. Since the enhanced arthritis in the IFN␤⫺/⫺ mice was not due to an
increase in T cell proliferation and could not be explained by an increase in anti-CII antibody production,
the exacerbation had to be due to other cells. We
therefore investigated cells that could be activated by
class II MHC–restricted T cells.
Initial investigation of the spleen macrophage
population showed a significant increase in TNF␣ production (Figure 2A) following 48 hours of culture in
vitro in the presence of both IFN␥ and LPS, but there
was no alteration in the level of IL-1␤ or IL-10 production (data not shown). However, since splenic macrophages are distant from the site of joint inflammation,
we aimed to investigate whether the macrophages as
well as other cells in the synovia were more activated in
the IFN␤⫺/⫺ mice. Mice were immunized, and 30 days
later the synovia were extracted and stimulated in vitro
with IFN␥ and LPS.
Following activation, flow cytometric analyses of
synoviocytes from IFN␤⫺/⫺ and IFN␤⫺/⫹ mice revealed
that the IFN␤⫺/⫺ synoviocytes included a greater number of macrophages; numbers of these increased slightly
after stimulation (Table 2). To our surprise, we did not
see an increase in TNF␣ intracellular staining when the
synovial cells were stimulated with IFN␥ and LPS. This
result was in contrast to that for the macrophages
derived from the spleen (Figure 2A). This could have
been due to a kinetic problem. However, in the synovial
population, there was a greater intracellular expression
of TNF␣ in IFN␤⫺/⫺ synoviocytes compared with that in
control cells, both before and after stimulation. There
was also an increase in ICAM-1⫹ and in ICAM-
Figure 2. Preimmunized mice deficient in interferon-␤ (IFN␤) have greater production of tumor necrosis factor ␣ (TNF␣) in spleen-derived
macrophages and in joints. A, Change in production of TNF␣ (delta TNF␣; calculated as cytokine production by macrophages in the presence of
lipopolysaccharide [LPS], IFN␥, or LPS and IFN␥ minus cytokine production by macrophages cultured in media alone). Results shown are from 1
of 3 representative experiments. B, Number of cells staining positive for CD11b and TNF␣ in hind or front paws of IFN␤⫺/⫺ (n ⫽ 5) and control
IFN␤⫺/⫹ (n ⫽ 5) mice, all having similar symptoms of arthritis. For each mouse, 5 distinct fields on a single sample slide were counted, and the
average number of cells was determined. Histologic findings were obtained from 1 collagen-induced arthritis experiment. Values in A and B are the
mean ⫾ SEM. ⴱ ⫽ P ⱕ 0.05 versus IFN␤⫺/⫺ mice. C and D, Immunohistochemistry 40 days after immunization of 1 representative IFN␤⫺/⫺ mouse
and 1 control IFN␤⫺/⫹ mouse, respectively, with comparable clinical symptoms of arthritis at the time of analyses. Sections were stained for TNF␣,
with positive cells staining brown (original magnification ⫻ 200).
1⫹,VCAM-1⫹ cells after stimulation, compared with
stimulated control cells (Table 2).
Interestingly, after 36 hours of cultivation, there
was a greater number of neutrophils (Ly6-G⫹ and
CD11b⫹ cells) in the IFN␤⫺/⫺ synovial population, which
increased after stimulation (Table 2). However, fresh synoviocytes originating from IFN␤⫺/⫹ mice had a greater
Table 2.
number of neutrophils compared with synoviocytes from
IFN␤⫺/⫺ mice (34.7% versus 25.8%). This result was
expected, since IFN␤⫺/⫺ mice have previously been shown
to have a reduced number of neutrophils (8).
Increased infiltration and fibroblast activation in
synovia of IFN␤ⴚ/ⴚ mice. To investigate the phenotype
of joint inflammation during the effector phase of CIA,
Expression levels of cell surface markers on synovial cells from IFN␤⫺/⫺ and control (IFN␤⫺/⫹)
IFN␤⫺/⫺ mice
Ly6-G ⫹ CD11b
Control mice
* Values are arbitrary units. Cells from 24–25 mice per group were pooled and stimulated. The experiment
was performed twice with similar results, and data shown are from 1 representative experiment.
IFN␤⫺/⫺ ⫽ interferon-␤ deficient; LPS ⫽ lipopolysaccharide; TNF␣ ⫽ tumor necrosis factor ␣;
ICAM-1 ⫽ intercellular adhesion molecule 1; VCAM-1 ⫽ vascular cell adhesion molecule 1.
† Change in expression was calculated by subtracting the expression level of cells cultured in media from
that of cells cultured with LPS ⫹ IFN␥.
hind and fore paws were harvested 40 days postimmunization and were evaluated by immunohistochemistry.
Concordant with the above data (Table 2 and Figure
2A), infiltrated areas of IFN␤⫺/⫺ mice contained more
CD11b⫹ cells and a greater amount of TNF␣, compared
with control mice (Figures 2B–D).
To determine whether fibroblasts were also affected by IFN␤ deficiency, fibroblasts were prepared from
mice with clinical signs of arthritis lasting for at least 1 week
and were then subjected to 6 passages of trypsinization in
vitro before analyses. Fibroblasts from the IFN␤⫺/⫺ mice
produced 10-fold higher amounts of IL-6 compared with
fibroblasts from control mice when both were cultured for
4 additional days (Figure 3A). Flow cytometric analyses of
the fibroblasts also showed that the IFN␤⫺/⫺ fibroblasts
were more activated than the control fibroblasts in terms of
expression of CD44 and ICAM and had a slightly increased
expression of CD40 (Figure 3A).
Ability of IFN␤-competent fibroblasts to protect
against CIA induction in IFN␤-deficient mice. To investigate whether the increased activation status in vitro of
IFN␤⫺/⫺ fibroblasts would also have an impact in vivo
during an inflammatory attack directed to the joints, we
next conducted fibroblast transfer experiments. Neither
IFN␤⫺/⫺ fibroblasts nor control fibroblasts could induce
arthritis when injected into the knee (1 ⫻ 105/knee) of
irradiated or nonirradiated B10.RIII mice (data not
shown). However, transfer of control fibroblasts (6 joint
injection sites, with a total of 2 ⫻ 106/mouse) into IFN␤⫺/⫺
mice resulted in significant protection from subsequent
induction of CIA, compared with IFN␤⫺/⫺ mice injected
with IFN␤⫺/⫺ fibroblasts (Figures 3B and C). The
IFN␤⫺/⫺ mice that had received control fibroblasts had an
arthritis profile similar to that of control IFN␤⫺/⫹ mice
that had received control fibroblasts, demonstrating the
importance of fibroblasts in this model, possibly via production of IFN␤.
Difference in osteoclast generation in vitro and
in vivo in IFN␤-deficient mice. Another important synovial cell that contributes to the arthritis process is the
osteoclast, which degrades cartilage and bone. It has
previously been shown that IFN␤-deficient and IFN␤
receptor–knockout mice have a greater capacity to
generate osteoclasts in vitro and that these mice have an
intrinsic bone erosion disorder in vivo (9). We therefore
analyzed osteoclastogenesis in IFN␤⫺/⫺ mice both in
vitro and in vivo. Bone marrow from IFN␤⫺/⫺ mice
generated more osteoclasts than did bone marrow from
control mice in vitro (Figures 4A–C). However, TRAP
staining analyses did not suggest an increase in osteoclastogenesis in vivo either in 4-month-old or in 16-month-old
naive IFN␤⫺/⫺ mice compared with age-matched naive
Figure 3. Amelioration of arthritis by reconstitution of interferon-␤
(IFN␤)–deficient mice with IFN␤-competent fibroblasts. A, Percent of
fibroblasts staining positive for different cell surface markers (y-axis at
left) and interleukin-6 (IL-6) production following in vitro culture of
fibroblasts from IFN␤⫺/⫺ and control mice (y-axis at right). Error bars
represent the SD. B and C, Arthritis index (mean score) and incidence
of arthritis, respectively, in IFN␤⫺/⫺ mice that had received 2 ⫻ 106
fibroblasts (FB) either from IFN␤⫺/⫺ mice (IFN␤⫺/⫺ ⫹ IFN␤⫺/⫺ FB)
or from control mice (IFN␤⫺/⫺ ⫹ IFN␤⫺/⫹ FB), and in control
IFN␤⫺/⫹ mice injected with control IFN␤⫺/⫹ fibroblasts (IFN␤⫺/⫹ ⫹
IFN␤⫺/⫹ FB). All mice were injected with fibroblasts and immunized
with type II collagen on day 0. Mice were scored twice weekly, starting
from day 14. Data are from a total of 8–10 mice per group pooled from
2 separate experiments, each with balanced groups. ⴱ ⫽ P ⱕ 0.05 for
IFN␤⫺/⫺ fibroblasts versus control IFN␤⫺/⫹ fibroblasts. ICAM-1 ⫽
intercellular adhesion molecule 1; VCAM-1 ⫽ vascular cell adhesion
molecule 1; IFN␥R␣ ⫽ IFN␥ receptor ␣-chain.
control mice (data not shown). Therefore, in contrast to
the previous report (9), the IFN␤⫺/⫺ mice on the B10.RIII
background did not show signs of an intrinsic bone erosion
disorder. Analyses of arthritic joints showed a clear in-
Figure 4. Enhanced generation of osteoclasts in interferon-␤ (IFN␤)–
deficient mice. A and B, In vitro generation of osteoclasts from bone
marrow (BM) cells from IFN␤⫺/⫺ or control IFN␤⫺/⫹ mice, respectively. Cells were cultured in vitro with macrophage colony-stimulating
factor (M-CSF) and recombinant murine RANKL (RL) for 6–7 days
in 48-well plates (original magnification ⫻ 100). C, Average number of
osteoclasts generated from bone marrow from 4 IFN␤⫺/⫺ and 4
control IFN␤⫺/⫹ mice. Cells were seeded in duplicate, and 5 fields of
view of each well were counted under 200⫻ magnification. Bone
marrow cells from IFN␤⫺/⫺ or control IFN␤⫺/⫹ mice cultured in
M-CSF alone did not support the generation of osteoclasts and were
used as negative control for the experiment (data not shown). D, In
vivo staining of osteoclasts (tartrate-resistant acid phosphatase
[TRAP] positive and multinucleated). Paws from preimmunized mice
(30 days prior) were removed and prepared for paraffin sectioning and
stained for TRAP. All joints in the section were counted, and joints
that contained ⱖ1 osteoclast were counted as affected. Consequently,
the graph represents the number of affected joints divided by the total
number of joints counted in both hind and front paws. Values are the
mean ⫾ SD (4–6 mice per group). ⴱ ⫽ P ⱕ 0.05 versus IFN␤⫺/⫺ mice.
crease in the number of osteoclasts in IFN␤⫺/⫺ mice
compared with control mice (Figure 4D), suggesting that
IFN␤ plays an important role in down-regulating
inflammation-mediated osteoclastogenesis.
IFN␤ has been used in the therapy of various
diseases, such as cancer and viral infections, and it is one of
the few available treatments for MS (14,35). There has also
been a great deal of interest in the possible therapeutic
effects of IFN␤ in RA, but there is a need to gain a better
understanding of the mechanisms of IFN␤ in arthritis. We
have previously shown in experimental autoimmune en-
cephalomyelitis, a murine model of MS, that IFN␤ has the
greatest effects in reducing the activation of macrophages
and microglia, with little effect on T cells. Moreover, IFN␤
deficiency did not cause a shift in the T helper phenotype
(25,36). In accordance with this, in the present study we
found that a lack of IFN␤ in arthritis led to a greater
activation of stromal cells such as macrophages and fibroblasts as well as to an enhanced generation of osteoclasts in
the arthritic joints, and we found that IFN␤ had little effect
on antigen-specific T cell responses.
IFN␤ has not only been shown to have antiinflammatory effects, but it has also been suggested to be
involved in development, homeostasis, and apoptosis of
several cell populations, such as osteoclasts, T cells,
neutrophils, and B cells (8). Osteoclasts are cells that
degrade bone, and they are vital in maintaining bone
homeostasis; however, excessive osteoclastogenesis in
the arthritic joints leads to a net loss of cartilage and
bone. Takayanagi et al elegantly showed the importance
of IFN␤ in osteoclastogenesis, with a clear increase in
osteoclastogenesis both in vitro and in vivo in naive mice
lacking the IFN␤ receptor and in IFN␤⫺/⫺ mice (37). In
the current study, we also show in vitro that IFN␤⫺/⫺ mice
have enhanced osteoclastogenesis. However, we did not
see an increase in osteoclastogenesis in vivo in naive mice,
but instead we found that immunized IFN␤⫺/⫺ mice have
a significant increase in the number of osteoclasts, indicating that IFN␤ in an arthritic joint would be beneficial in
reducing the amount of joint destruction.
IFN␤ has been shown to be involved in apoptosis of
T cells and neutrophils (7,10). Together with stromal
cell–derived factor 1 (CXCL12), IFN␤ was shown to
inhibit apoptosis of T cells located in the joint synovia of
human RA patients and was therefore believed to maintain
T cells within the joint (7). However, the role of T cells in
RA synovia is not known, and it is possible that T cells are
involved in the priming phase of arthritis but have little role
in the effector phase of the disease, since T cells have been
shown to proliferate poorly and secrete few cytokines (38).
In addition, anti–T cell agents seem to have little influence
on ongoing arthritis (39,40), whereas anti–B cell and antimonokine reagents such as anti-TNF␣ have had considerable therapeutic effects (41,42). In the present study, the
IFN␤⫺/⫺ mice showed no difference in T cell numbers
compared with control mice, indicating that there was no
increase in apoptosis in the T cell compartment. Furthermore, there was no difference in antigen-specific T cell
proliferation upon restimulation in vitro. In addition, there
was no difference in the degree of T cell infiltration in the
synovia of these mice (data not shown).
Neutrophils have been suggested to be involved
both in RA and in CIA (10,43). In line with this, IFN␤ has
been shown to prevent apoptosis of neutrophils (10), and
IFN␤⫺/⫺ mice have previously been shown to have a
reduced number of neutrophils (8). Indeed, in the current
study, we observed reduced numbers of neutrophils in both
spleen and lymph nodes (data not shown). However, a
prominent role of neutrophils was not found, since
IFN␤⫺/⫺ mice developed a more exacerbated arthritis.
Interestingly, the neutrophil populations in the spleen and
synovia of IFN␤-deficient mice were able to withstand (and
even proliferate in) in vitro culture (36 hours), which was in
contrast to IFN␤⫺/⫹ neutrophils (data not shown and
Table 2). This indicated that the influence of IFN␤ on
neutrophil apoptosis in vitro differs from that in vivo.
In the present study, lack of IFN␤ expression
seemed to have a prominent effect on stromal cells,
including macrophages and fibroblasts. Flow cytometric
analyses of synoviocytes showed that, upon stimulation
with LPS and IFN␥, there was an up-regulation of
CD11b⫹ cells and an increase in macrophages as well as
an up-regulation of the adhesion molecule ICAM-1.
This indicates that the stromal cells in the synovia are
more readily activated in IFN␤-deficient mice, potentially leading to a more chronic arthritis profile.
Fibroblasts have been implicated in both the
priming and effector phases of RA (7,44); therefore,
targeting fibroblasts should be of therapeutic benefit. In
the present study, a lack of IFN␤ was associated with
fibroblasts having a more active phenotype with increased IL-6 production. There was also an upregulation in the cell surface molecules CD44, CD40,
and ICAM-1, all of which are believed to be involved in
the pathogenesis of RA (44). However, since fibroblasts
require IFN␤ to produce IFN␣ (6), it is possible that the
phenotype of the IFN␤-deficient fibroblast is partially
due to IFN␣ deficiency. Nevertheless, the current study
determined that local injection of IFN␤-producing fibroblasts into the joints of IFN␤⫺/⫺ mice completely reverted the augmented arthritis phenotype to that of
control mice. This is an important finding, since it shows
that physiologic levels of IFN␤ produced by nontransfected fibroblasts can have a beneficial effect in vivo.
In summary, the present study shows that mice
deficient in IFN␤ display a chronic arthritis with a high
incidence. These results may have been expected from
previous animal studies using IFN␤ treatments. However,
we were able to demonstrate that the mechanism of action
of IFN␤ is not mediated through T cells but rather results
from an increased activation of resident cells of the joint
(i.e., fibroblasts, macrophages, and osteoclasts). It is likely
that IFN␤ serves to control the activation state of fibro-
blasts, and in the absence (or in the presence of low levels)
of IFN␤, fibroblasts become more prone to produce cytokines, chemokines, and growth factors that in turn enhance
infiltration of inflammatory cells. In fact, it has previously
been shown that treatment of RA-derived fibroblasts in
vitro with IFN␤ leads to decreased production of chemokines, such as MMP-1 and MMP-2, as well as to decreased
production of prostaglandin E2 (45). The mouse model
used in this study thus supported the conclusion that
fibroblasts have an important role as producers of IFN␤.
We were able to show via transfer of IFN␤-competent
fibroblasts that increasing the amount of IFN␤ indeed has
a profound effect on arthritis. This is an important finding,
since it has recently been proposed that naturally produced
IFN␤ plays an antiinflammatory role in RA patients (19).
There is no doubt that IFN␤ has potent antiinflammatory properties. However, it remains unclear
which cell type and which phase of arthritis is most
receptive to IFN␤ treatment, and this may explain the
limited success of clinical trials. Animal models can aid
in pinpointing mechanism(s) of this cytokine in order to
determine the most effective treatment protocol and the
group of patients that would derive the most benefit.
We express our gratitude to Dr. J. Bäcklund and
Caroline Parsons for critical reading of the manuscript.
1. Uze G, Lutfalla G, Gresser I. Genetic transfer of a functional
human interferon ␣ receptor into mouse cells: cloning and expression of its cDNA. Cell 1990;60:225–34.
2. Novick D, Cohen B, Rubinstein M. The human interferon ␣/␤
receptor: characterization and molecular cloning. Cell 1994;77:
3. Van den Broek MF, Muller U, Huang S, Zinkernagel RM, Aguet
M. Immune defence in mice lacking type I and/or type II
interferon receptors. Immunol Rev 1995;148:5–18.
4. Shen H, Zhang M, Minuk GY, Gong Y. Different effects of rat
interferon ␣, ␤ and ␥ on rat hepatic stellate cell proliferation and
activation. BMC Cell Biol 2002;3:9–17.
5. Deonarain R, Alcami A, Alexiou M, Dallman MJ, Gewert DR,
Porter AC. Impaired antiviral response and ␣/␤ interferon induction in mice lacking ␤ interferon. J Virol 2000;74:3404–9.
6. Erlandsson L, Blumenthal R, Eloranta ML, Engel H, Alm G,
Weiss S, et al. Interferon-␤ is required for interferon-␣ production
in mouse fibroblasts. Curr Biol 1998;8:223–6.
7. Buckley CD. Michael Mason prize essay 2003: why do leucocytes
accumulate within chronically inflamed joints? Rheumatology
(Oxford) 2003;42:1433–44.
8. Deonarain R, Verma A, Porter AC, Gewert DR, Platanias LC,
Fish EN. Critical roles for IFN-␤ in lymphoid development,
myelopoiesis, and tumor development: links to tumor necrosis
factor ␣. Proc Natl Acad Sci U S A 2003;100:13453–8.
9. Takayanagi H, Kim S, Taniguchi T. Signaling crosstalk between
RANKL and interferons in osteoclast differentiation. Arthritis
Res 2002;4 Suppl 3:S227–32.
Wang K, Scheel-Toellner D, Wong SH, Craddock R, Caamano J,
Akbar AN, et al. Inhibition of neutrophil apoptosis by type 1 IFN
depends on cross-talk between phosphoinositol 3-kinase, protein
kinase C-delta, and NF-␬B signaling pathways. J Immunol 2003;
Jungo F, Dayer JM, Modoux C, Hyka N, Burger D. IFN-␤ inhibits
the ability of T lymphocytes to induce TNF-␣ and IL-1␤ production
in monocytes upon direct cell-cell contact. Cytokine 2001;14:272–82.
Rep MH, Hintzen RQ, Polman CH, van Lier RA. Recombinant
interferon-␤ blocks proliferation but enhances interleukin-10 secretion by activated human T-cells. J Neuroimmunol 1996;67:111–8.
Van Holten J, Reedquist K, Sattonet-Roche P, Smeets TJ, PlaterZyberk C, Vervoordeldonk MJ, et al. Treatment with recombinant
interferon-␤ reduces inflammation and slows cartilage destruction
in the collagen-induced arthritis model of rheumatoid arthritis.
Arthritis Res Ther 2004;6:R239–49.
Abdul-Ahad AK, Galazka AR, Revel M, Biffoni M, Borden EC.
Incidence of antibodies to interferon-␤ in patients treated with
recombinant human interferon-␤ 1a from mammalian cells. Cytokines Cell Mol Ther 1997;3:27–32.
Tak PP, Hart BA, Kraan MC, Jonker M, Smeets TJ, Breedveld
FC. The effects of interferon ␤ treatment on arthritis. Rheumatology (Oxford) 1999;38:362–9.
Triantaphyllopoulos KA, Williams RO, Tailor H, Chernajovsky Y.
Amelioration of collagen-induced arthritis and suppression of
interferon-␥, interleukin-12, and tumor necrosis factor ␣ production by interferon-␤ gene therapy. Arthritis Rheum 1999;42:90–9.
Van Holten J, Plater-Zyberk C, Tak PP. Interferon-␤ for treatment of rheumatoid arthritis? Arthritis Res 2002;4:346–52.
Smeets TJ, Dayer JM, Kraan MC, Versendaal J, Chicheportiche R,
Breedveld FC, et al. The effects of interferon-␤ treatment of synovial
inflammation and expression of metalloproteinases in patients with
rheumatoid arthritis. Arthritis Rheum 2000;43:270–4.
Holten JV, Smeets TJ, Blankert P, Tak PP. Expression of interferon ␤ (IFN-␤) in synovial tissue from rheumatoid arthritis
patients compared to osteoarthritis and reactive arthritis patients.
Ann Rheum Dis 2005. E-pub ahead of print.
Tak PP. IFN-␤ in rheumatoid arthritis. Front Biosci 2004;9:
Van Holten J, Pavelka K, Vencovsky J, Stahl H, Rozman B,
Genovese M, et al. A multicentre, randomised, double-blind,
placebo controlled phase II study of subcutaneously administered
interferon ␤1a in the treatment of patients with active rheumatoid
arthritis. Ann Rheum Dis 2004;64:64–9.
Courtenay JS, Dallman MJ, Dayan AD, Martin A, Mosedale B.
Immunisation against heterologous type II collagen induces arthritis in mice. Nature 1980;283:666–8.
Corthay A, Johansson A, Vestberg M, Holmdahl R. Collageninduced arthritis development requires ␣ ␤ T cells but not ␥ delta
T cells: studies with T cell-deficient (TCR mutant) mice. Int
Immunol 1999;11:1065–73.
Mo JA, Holmdahl R. The B cell response to autologous type II
collagen: biased V gene repertoire with V gene sharing and
epitope shift. J Immunol 1996;157:2440–8.
Teige I, Treschow A, Teige A, Mattsson R, Navikas V, Leanderson
T, et al. IFN-␤ gene deletion leads to augmented and chronic
demyelinating experimental autoimmune encephalomyelitis. J Immunol 2003;170:4776–84.
Miller EJ. Structural studies on cartilage collagen employing
limited cleavage and solubilization with pepsin. Biochemistry
Michaelsson E, Andersson M, Engstrom A, Holmdahl R. Identi-
fication of an immunodominant type-II collagen peptide recognized by T cells in H-2q mice: self tolerance at the level of
determinant selection. Eur J Immunol 1992;22:1819–25.
Holmdahl R, Carlsen S, Mikulowska A, Vestberg M, Brunsberg U,
Hansson A, et al. Genetic analysis of mouse models for rheumatoid arthritis. In: Adolph KW, editor. Human genome methods.
New York: CRC Press; 1998. p. 215–38.
Holmdahl R, Klareskog L, Andersson M, Hansen C. High antibody response to autologous type II collagen is restricted to H-2q.
Immunogenetics 1986;24:84–9.
Nandakumar KS, Svensson L, Holmdahl R. Collagen type IIspecific monoclonal antibody-induced arthritis in mice: description
of the disease and the influence of age, sex, and genes. Am J
Pathol 2003;163:1827–37.
Svensson L, Jirholt J, Holmdahl R, Jansson L. B cell-deficient
mice do not develop type II collagen-induced arthritis (CIA). Clin
Exp Immunol 1998;111:521–6.
Malmstrom V, Michaelsson E, Burkhardt H, Mattsson R, Vuorio
E, Holmdahl R. Systemic versus cartilage-specific expression of a
type II collagen-specific T-cell epitope determines the level of
tolerance and susceptibility to arthritis. Proc Natl Acad Sci U S A
Noronha A, Toscas A, Jensen MA. Interferon ␤ decreases T cell
activation and interferon ␥ production in multiple sclerosis. J Neuroimmunol 1993;46:145–53.
Nandakumar KS, Backlund J, Vestberg M, Holmdahl R. Collagen
type II (CII)-specific antibodies induce arthritis in the absence of
T or B cells but the arthritis progression is enhanced by CIIreactive T cells. Arthritis Res Ther 2004;6:R544–50.
Giovannoni G, Miller DH. Multiple sclerosis and its treatment. J
R Coll Physicians Lond 1999;33:315–22.
Matheu V, Treschow A, Navikas V, Issazadeh-Navikas S. Upregulation of B7 molecules (CD80 and CD86) and exacerbated eosinophilic pulmonary inflammatory response in mice lacking the
IFN-␤ gene. J Allergy Clin Immunol 2003;111:550–7.
Takayanagi H, Ogasawara K, Hida S, Chiba T, Murata S, Sato K,
et al. T-cell-mediated regulation of osteoclastogenesis by signalling
cross-talk between RANKL and IFN-␥. Nature 2000;408:600–5.
Korganow AS, Ji H, Mangialaio S, Duchatelle V, Pelanda R,
Martin T, et al. From systemic T cell self-reactivity to organspecific autoimmune disease via immunoglobulins. Immunity
Firestein GS, Zvaifler NJ. How important are T cells in chronic
rheumatoid synovitis? [editorial]. Arthritis Rheum 1990;33:768–73.
Fox DA. The role of T cells in the immunopathogenesis of
rheumatoid arthritis: new perspectives [review]. Arthritis Rheum
Kazkaz H, Isenberg D. Anti B cell therapy (rituximab) in the
treatment of autoimmune diseases. Curr Opin Pharmacol 2004;4:
Ehrenstein MR, Evans JG, Singh A, Moore S, Warnes G, Isenberg
DA, et al. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNF␣ therapy. J Exp Med
Suzuki M, Uetsuka K, Shinozuka J, Nakayama H, Doi K. Immunohistochemical study on type II collagen-induced arthritis in
DBA/1J mice. Exp Anim 1997;46:259–67.
Ritchlin C. Fibroblast biology: effector signals released by the
synovial fibroblast in arthritis. Arthritis Res 2000;2:356–60.
Taylor PC, Peters AM, Paleolog E, Chapman PT, Elliott MJ,
McCloskey R, et al. Reduction of chemokine levels and leukocyte
traffic to joints by tumor necrosis factor ␣ blockade in patients with
rheumatoid arthritis. Arthritis Rheum 2000;43:38–47.
Без категории
Размер файла
163 Кб
responsible, induced, exacerbated, deficiency, mice, arthritis, collagen, stroma, osteoclast, cells, interferon
Пожаловаться на содержимое документа