вход по аккаунту


Important role of interleukin-3 in the early phase of collagen-induced arthritis.

код для вставкиСкачать
Vol. 60, No. 5, May 2009, pp 1352–1361
DOI 10.1002/art.24441
© 2009, American College of Rheumatology
Important Role of Interleukin-3 in the Early Phase of
Collagen-Induced Arthritis
Hilke Brühl,1 Josef Cihak,2 Marianne Niedermeier,1 Andrea Denzel,1
Manuel Rodriguez Gomez,1 Yvonne Talke,1 Nicole Goebel,1 Jiří Plachý,3
Manfred Stangassinger,2 and Matthias Mack1
marked exacerbation of the disease, with increased
peripheral blood basophil and plasma IL-6 levels and
increased titers of anticollagen antibody. In studies of
the regulation of IL-3 expression in CD4ⴙ T cells, IL-6
and IL-4 suppressed the release of IL-3 by activated
CD4ⴙ T cells, whereas lipopolysaccharide and CpG
DNA up-regulated IL-3 secretion in activated CD4ⴙ T
cells by acting on costimulatory cells.
Conclusion. Taken together, the present results
demonstrate for the first time that IL-3 has an important role in the early phase of CIA.
Objective. Activation of basophils contributes to
memory immune responses and results in exacerbation
of collagen-induced arthritis (CIA). We undertook the
present study to analyze the production and biologic
effects of interleukin-3 (IL-3), a strong activator of
basophils, in CIA.
Methods. Arthritis was induced by immunization
with type II collagen. Mice were treated with blocking
monoclonal antibodies against IL-3 or with recombinant IL-3. Clinical scoring, histologic analysis,
fluorescence-activated cell sorter analysis, enzymelinked immunosorbent assay, and cell culturing were
performed to assess disease activity and IL-3 production.
Results. IL-3 was produced in large quantities by
collagen-specific CD4ⴙ T cells in the spleen and was
present in the synovial tissue during onset of arthritis,
but was down-regulated in paws with severe inflammation. Blockade of IL-3 during the time of arthritis onset
resulted in profound improvement of the disease, with
reductions in synovial leukocyte and cytokine levels,
peripheral blood basophil levels, and anticollagen antibody titers. Blockade of IL-3 during the late phase of
arthritis had no beneficial effect. Administration of
recombinant IL-3 during onset of arthritis induced a
Interleukin-3 (IL-3), together with IL-5 and
granulocyte–macrophage colony-stimulating factor
(GM-CSF), belongs to a family of hematopoietic cytokines with 4 short ␣-helical bundles. Each of these
cytokines binds to a unique ␣-receptor subunit (e.g.,
IL-3 receptor ␣ [IL-3R␣] for IL-3). Signal transduction
is mediated by a common ␤-receptor subunit, ␤c, that is
unable to bind any of the cytokines (1). In the mouse, a
second ␤-receptor subunit, ␤IL-3, that associates exclusively with the IL-3R␣ subunit, has been identified (2).
IL-3 is produced mainly by CD4⫹ T cells. However, little is known about the regulation of IL-3 secretion from T cells. IL-3 contributes to growth, differentiation, and survival of CD34⫹ hematopoietic progenitor
cells. Although disruption of the IL-3 gene does not
affect basal hematopoiesis, it is necessary for supporting
increased numbers of basophils and tissue mast cells
during parasite infection (3). In vitro, IL-3 promotes the
differentiation of basophils and mast cells from bone
marrow cells (4–7), and it has been reported that it
facilitates and induces histamine and IL-4 release by
basophils (8–12). In monocyte/macrophages, IL-3 upregulates class II major histocompatibility complex expression and enhances lipopolysaccharide (LPS)–
induced IL-1 secretion (13,14). Together with IL-4 or
Supported by DFG grants to Drs. Brühl and Mack.
Hilke Brühl, MD, Marianne Niedermeier, Dipl Biol, Andrea
Denzel, Dipl Biol, Manuel Rodriguez Gomez, Dipl Biol, Yvonne
Talke, Dipl Ing, Nicole Goebel, Matthias Mack, MD: University
Hospital Regensburg, Regensburg, Germany; 2Josef Cihak, PhD,
Manfred Stangassinger, PhD: University of Munich, Munich, Germany; 3Jiřı́ Plachý, PhD: Czech Academy of Sciences, Prague, Czech
Address correspondence and reprint requests to Matthias
Mack, MD, University Hospital Regensburg, Department of Internal
Medicine II, Franz-Josef-Strauss Allee 11, 93053 Regensburg, Germany. E-mail:
Submitted for publication September 8, 2008; accepted in
revised form January 19, 2009.
interferon-␤ (IFN␤), IL-3 supports differentiation of
monocytes into dendritic cells (15,16). Induction of
osteoclast-like cells by IL-3 has also been described
Little is known about the role of IL-3 in arthritis.
In an early study, IL-3 messenger RNA was not detected
in the synovium of patients with RA (19), and no effect
of IL-3 on cultured fibroblast-like synoviocytes was
observed (20). Nevertheless, genetic analysis revealed an
association between a single-nucleotide polymorphism
in the IL-3 promoter and rheumatoid arthritis (21).
Our interest in analyzing the contribution of IL-3
in arthritis was prompted by our recent observation of
marked aggravation of collagen-induced arthritis (CIA)
by activation of basophils with antibodies against IgE
(22). Basophils can be activated not only by crosslinkage
of surface IgE, but by other factors, especially IL-3, as
well. Basophils release not only IL-4, but also the
proarthritogenic cytokine IL-6 (see below). Even at very
low concentrations, IL-3 induces pronounced release of
IL-6 from murine basophils and prolongs the survival of
basophils in culture (see below). Therefore, using the
DBA/1 mouse model of CIA, we analyzed the expression
of IL-3 in the paws at various stages of the disease,
quantified the number of basophils and mast cells in the
synovial tissue, and investigated how disease incidence
and activity were influenced by blockade or administration of IL-3. In addition, we studied the regulation of
IL-3 release from T cells in vitro, to better understand
the stage-specific release of IL-3 in arthritis.
Induction of CIA and treatment of mice. Arthritis was
induced in male DBA/1 mice by initial subcutaneous injection
of 100–200 ␮g bovine type II collagen (CII) (C1188; SigmaAldrich, Munich, Germany) in complete Freund’s adjuvant at
the tail base on day 0, and restimulation by intraperitoneal (IP)
injection of 100–200 ␮g CII without adjuvant on day 21.
Clinical arthritis was scored on a 0–4 scale under blinded
conditions, as follows: 0 ⫽ normal, 1 ⫽ swelling in 1 joint, 2 ⫽
swelling in ⬎1 joint, 3 ⫽ swelling of the entire paw, and 4 ⫽
deformity and/or ankylosis. In some experiments, animals
received daily IP injections of 35 ␮g of a blocking anti–IL-3
antibody (clone MP2-8F8; Biozol, Munich, Germany) or purified rat IgG (Sigma-Aldrich) from day 21 through day 36. Mice
were killed on day 37. In other experiments, daily IP injections
of 50 ␮g anti–IL-3 antibody or purified rat IgG were started
when the arthritis score of an individual mouse was at least 2;
treatment was continued for 7 days. In additional experiments,
mice were treated from day 20 through day 30 with twice-daily
IP injections of 100 ng IL-3 (PeproTech, Rocky Hill, NJ) or
phosphate buffered saline (PBS). Animal experiments were
performed in accordance with the legal requirements of the
government of Bavaria (Az. 55.2-1-54-2531-109-05).
Preparation of synovial tissue and quantification of
cytokines and infiltrating cells. Fore and hind paws were
removed at the ankle joint, the skin was removed from the
inflamed paws, and the remaining tissue was carefully recovered with a scalpel in a volume of 500 ␮l/1,000 ␮l PBS. Samples
were immediately centrifuged for 10 minutes at 400g. The
supernatant was immediately frozen and used for enzymelinked immunosorbent assay (ELISA) of cytokines. The synovial tissue was digested with type I collagenase (SigmaAldrich) for 20 minutes at 37°C to obtain a single-cell
suspension and used for fluorescence-activated cell sorter
(FACS) analysis.
Histologic analysis. Hind paws were fixed in 3.7%
formalin for 24 hours, decalcified with RDO rapid decalcifier
(Medite, Burgdorf, Germany), and embedded in paraffin. At
least 10 sections of the tarsometatarsal joints (5 ␮m thick) were
stained with hematoxylin and eosin and scored in a blinded
manner on a scale of 0 (normal)–2, for the following categories: synovial inflammation (1 ⫽ focal inflammatory infiltrates;
2 ⫽ inflammatory infiltrate dominating the cellular histology),
synovial hyperplasia (1 ⫽ continuous, at least 3-layer–thick
synovial lining in 1 joint; 2 ⫽ same findings in several joints),
pannus formation and cartilage loss (1 ⫽ cartilage partially
covered by pannus, no cartilage loss; 2 ⫽ same findings but
with cartilage loss), and bone destruction (1 ⫽ small areas of
bone destruction; 2 ⫽ widespread bone destruction).
Flow cytometry and ELISA for cytokines. The following antibodies were used for flow cytometry or magnetic cell
separation: fluorescein isothiocyanate (FITC)–conjugated antiCD45 (30-F11), allophycocyanin (APC)–conjugated antiCD45, FITC-conjugated anti-CD11b (M1/70), phycoerythrin
(PE)–conjugated anti-CD11b, Fc block (2.4G2), PEconjugated anti-CD19 (1D3), APC-conjugated anti–GR-1
(RB6-8C5), APC-conjugated anti-CD4 (RM4-5), PEconjugated anti–c-Kit (2B8), FITC-conjugated anti-IgE (R3572), PE-conjugated anti–IL-3 (MP2-8F8), and FITC- and
PE-conjugated isotype controls (all from BD Biosciences, San
Jose, CA), and APC-conjugated anti-CD49b (DX5; Miltenyi
Biotec, Bergisch Gladbach, Germany). Unfixed cells were
preincubated for 15 minutes on ice with Fc block (5 ␮g/ml) and
then for 45 minutes on ice with combinations of directly
labeled antibodies. After 3 washing steps, red blood cells were
lysed with FACS lysing solution (BD Biosciences) and samples
were analyzed on a FACSCalibur (BD Biosciences). Monocytes and neutrophils were identified by light scatter properties
as well as expression levels of CD11b and GR-1. In synovial
tissue, mast cells and basophils were identified by expression of
IgE and the presence or absence of c-Kit, respectively. In
peripheral blood, basophils could be identified by expression
of IgE only, because no mast cells are present in the peripheral
blood. For quantification of intracellular IL-3, cells were first
stained with APC-conjugated anti-CD4 and then treated with
Fix-Perm and Perm-Wash solutions according to the instructions of the manufacturer (BD Biosciences) and stained with
PE-conjugated antibody against IL-3.
Levels of IL-3, IL-4, and IL-6 were measured with
ELISA kits from BD Biosciences. IL-1␤, IFN␥, tumor necrosis
factor ␣ (TNF␣), GM-CSF, and IL-17 were measured with
Quantikine ELISA kits (R&D Systems, Wiesbaden, Ger-
many). Antibodies against collagen were quantified by ELISA.
Collagen (20 ␮g/ml) was coated overnight on ELISA plates.
Plasma samples were diluted in PBS–3% bovine serum albumin. Immunoglobulins bound to collagen were detected with a
horseradish peroxidase (HRP)–labeled polyclonal rabbit antimouse antibody (P260; Dako, Glostrup, Denmark) or HRPlabeled monoclonal antibodies specific for murine IgG1 (clone
LO-MG1-2; Serotec, Wiesbaden, Germany) or for murine
IgG2a (clone R19-15; BD PharMingen, Heidelberg, Germany). The murine cytokines IL-3, IL-4, and IL-6 were obtained from PeproTech.
Isolation and culture of cells. Splenocytes from mice
with CIA were depleted of B cells and CD4⫹ T cells with
magnetic beads directed against CD19 and CD4 (Miltenyi
Biotec). Basophils, monocytes, or neutrophils were depleted by
incubation of splenocytes with fluorochrome-labeled antibodies against IgE, CD11b, or GR-1 and subsequent incubation
with magnetic beads directed against the fluorochrome. Total
splenocytes or splenocytes depleted of a specific leukocyte
subset were cultured in 96-well flat-bottomed plates (2 ⫻ 106
cells/200 ␮l medium) for 3 days with or without bovine CII (40
␮g/ml). Cell culture supernatant was used for ELISA, and
collagen-specific release of IL-3 was calculated as IL-3 release
with collagen minus IL-3 release without collagen.
CD4⫹ T cells, B cells, and monocytes were isolated
from the splenocytes of C57BL/6 or Toll-like receptor 4
(TLR-4)–deficient C3H mice (Charles River, Sulzfeld, Germany) using magnetic microbeads against CD4, CD19, and
CD11b (Miltenyi Biotec), respectively. The purity of the
isolated cells was routinely ⬎95%. Cells were cultured for 3
days in 96-well round-bottomed plates in a total volume of 200
␮l medium (RPMI 1640 with 10% fetal calf serum [FCS] and
1% penicillin/streptomycin) per well. The number of cells per
well was 50,000 for each cell type. Beads coated with antibodies against CD3 and CD28 (T cell expander beads; Dynal/
Invitrogen, Hamburg, Germany) were used at a concentration
of 50,000 beads/well. The following reagents were added as
described below: anti-CD3 monoclonal antibody (0.5 ␮g/ml;
clone 2C11), LPS (10 ␮g/ml; Sigma-Aldrich), CpG DNA (1
␮M; TCCATGACGTTCCTGATGCT) (CPG1668; TIB Molbiol, Berlin, Germany), IL-4 (10 ng/ml), and IL-6 (10 ng/ml).
The concentration of IL-3 in the culture supernatant after 3
days was determined by ELISA. For intracellular staining of
IL-3, CD4⫹ T cells were activated for 3 days with anti-CD3
and LPS in the presence of B cells or monocytes. Phorbol
myristate acetate (10 ng/ml) and ionomycin (1 ␮g/ml) were
added during the last 4 hours of culture, and brefeldin A (5
␮g/ml) was added during the last 2.5 hours of culture.
Basophils were enriched from the bone marrow of
C57BL/6 mice with magnetic microbeads against DX-5 and LS
columns (Miltenyi Biotec). Basophils were identified by low
expression of CD45 and high expression of DX-5 and constituted ⬃10% of the enriched cells. Basophils (1,100 cells/well)
were cultured for various periods of time in 96-well roundbottomed plates in a total volume of 200 ␮l medium (RPMI
1640 with 10% FCS and 1% penicillin/streptomycin) per well
with various concentrations of recombinant IL-3. The concentration of IL-4 and IL-6 in the cell culture supernatant was
measured by ELISA. The number of live basophils per well at
each time point was quantified by flow cytometry after staining
of total cells with antibodies against CD45 and DX-5 in
combination with propidium iodide (10 ␮g/ml) and counting
beads (Coulter, Krefeld, Germany).
Statistical analysis. Mean ⫾ SEM values were calculated. The significance of group differences was determined by
Student’s 1-sided t-test. P values less than 0.05 were considered
IL-3 and basophils in CIA. We first investigated
whether IL-3 is produced in the spleen and synovial
tissue of mice with arthritis. On day 31 after the first
immunization with collagen, total splenocytes or splenocytes depleted of specific leukocyte subsets were restimulated with collagen, and collagen-specific release of
IL-3 was determined by subtracting the amount of IL-3
release in the absence of collagen (Figure 1A). Total
splenocytes and splenocytes depleted of CD19⫹ cells (B
cells), IgE⫹ cells (basophils), or GR-1⫹ cells (mainly
neutrophils) produced large amounts of IL-3 after stimulation with collagen. In contrast, depletion of CD4⫹ T
cells or CD11b⫹ cells (mainly monocytes) completely
abrogated the collagen-specific release of IL-3 (Figure
1A), indicating that IL-3 production requires the presence of both CD4⫹ T cells and CD11b⫹ costimulatory
cells. B cells are neither necessary nor sufficient to
support IL-3 production by CD4⫹ T cells, and the
increased release of IL-3 in the absence of B cells
resulted from a higher number of T cells and monocytes
in the assay, since the total number of leukocytes per
well was kept constant and B cells constitute ⬎50% of
the leukocytes in the spleen.
Cytokine production in hind paw synovial tissue
was measured on day 36 after the first immunization
with collagen (Figure 1B). For that purpose the paws
were dissected at the ankle joint, the skin was removed,
and the soft tissue completely recovered in 1 ml PBS.
After centrifugation for 10 minutes at 400g, cytokines in
the supernatant were measured by ELISA. Paws were
classified into 2 groups according to the degree of
clinically apparent arthritis, with 14 paws having a score
of 0–2 and 12 paws having a score of 3 or 4. As expected,
paws with pronounced inflammation had high levels of
IL-6 and IL-1␤ (mean 683 pg/ml and 619 pg/ml, respectively),while IL-6 and IL-1␤ levels in paws with no or
low-grade inflammation were severalfold lower (144
pg/ml and 72 pg/ml, respectively). TNF␣ was detectable
only at very low levels, but was increased in paws with a
clinical score of 3 or 4. In contrast, IL-3 was readily
detectable in paws with a score of 0–2 (mean 66 pg/ml),
but was highly significantly reduced in paws with severe
inflammation (14 pg/ml). Synovial tissue IL-3 levels were
Figure 1. Levels of interleukin-3 (IL-3) and frequency of basophils in
collagen-induced arthritis. A, IL-3 production by splenocytes after
restimulation with collagen. On day 31 after mice were first immunized
with collagen, total splenocytes or splenocytes depleted of specific
leukocyte subsets as indicated were incubated for 3 days with type II
collagen. ⴱⴱ ⫽ P ⬍ 0.01 versus total cells. B, Measurement of synovial
tissue cytokine levels in hind paws with a low amount of inflammation
(score 0–2; n ⫽ 14) and in those with a high amount of inflammation
(score 3 or 4; n ⫽ 12) on day 36 after induction of arthritis. ⴱ ⫽ P ⬍
0.05; ⴱⴱ ⫽ P ⬍ 0.01. GM-CSF ⫽ granulocyte–macrophage colonystimulating factor; TNF␣ ⫽ tumor necrosis factor ␣; IFN␥ ⫽ interferon-␥; NS ⫽ not significant. C, Influence of IL-3 on activation and
survival of basophils in 4-day culture with IL-3 at various concentrations. In the absence of IL-3, there was no detectable release of IL-6
(or IL-4 [data not shown]), and rapid cell death of basophils was
observed. Cytokine release and survival of basophils were markedly
increased with addition of low amounts of IL-3. Values in A–C are the
mean and SEM. D, Flow cytometric detection of basophils and mast
cells in single-cell suspensions prepared from synovial tissue from
inflamed paws. The frequency of basophils (IgE⫹, c-Kit⫺) and mast
cells (IgE⫹, c-Kit⫹) is shown as a percentage of total CD45⫹
infiltrating leukocytes.
negatively correlated with the clinical arthritis score (r ⫽
⫺0.75). Levels of IL-17, GM-CSF, and IFN␥ in the
synovium did not correlate with the degree of paw
inflammation (Figure 1B).
IL-3 is known to induce and facilitate release of
histamine and IL-4 from basophils. We found that IL-3
by itself also induces substantial release of IL-6 from
murine basophils and markedly prolongs the survival of
isolated basophils in culture (Figure 1C). Release of
IL-6 was observed with administration of IL-3 at very
low concentrations (half-maximal release at ⬃0.3 ng/ml
IL-3). IL-3 also induced release of IL-4 from basophils,
but the release of IL-4 was ⬃3-fold lower than the
release of IL-6 (data not shown). In the absence of IL-3,
only 6% of the basophils survived 4-day culture in
medium, while addition of IL-3 increased the survival of
basophils to ⬃60% (Figure 1C).
Flow cytometry was performed to analyze
whether basophils and mast cells are present in the
inflamed paws of mice with CIA. Synovial tissue from
inflamed paws was digested with collagenase to obtain a
single-cell suspension, and cells were stained with antibodies against IgE, c-Kit, and CD45 to identify basophils
(IgE⫹, c-Kit⫺, and CD45low) and mast cells (IgE⫹,
c-Kit⫹, and CD45⫹). While basophils were unambiguously detectable in all inflamed paws at a frequency of
⬃0.4% of total infiltrating CD45⫹ leukocytes, only very
few mast cells were found (and only in some of the
inflamed paws), with a 20-fold lower frequency than that
of basophils (Figure 1D). The majority of infiltrating
cells were monocytes and neutrophils, which are also
known to be responsive to IL-3.
Functional analysis of IL-3 in CIA. The presence
of IL-3 in early forms of CIA, as well as the presence of
cells (e.g., basophils and monocytes) that are able to
respond to IL-3 by releasing proarthritogenic cytokines
such as IL-6 or IL-1, suggest that IL-3 might be involved
in the pathogenesis of arthritis. We therefore investigated whether blockade of IL-3 with a monoclonal
antibody improves the incidence and severity of arthritis
in mice injected with 200 ␮g CII on day 0 and day 21.
One group of mice (n ⫽ 15) received daily IP injections
of anti–IL-3 antibody (35 ␮g/day) from day 21 through
day 36, while the control group (n ⫽ 15) was injected
with rat IgG at the same dose and time intervals.
Blockade of IL-3 during the time of disease onset highly
significantly reduced the clinical severity of arthritis. On
day 37, the mean arthritis score in the control group was
5.3, whereas in the anti–IL-3 treatment group it was 1.9
(Figure 2A). The maximum score of 4 was reached in 3
of 60 paws from mice in the anti–IL-3 group and 14 of 60
from mice in the control group. The incidence of
recovered synovial tissue (Figure 2B). Histologic analysis of the hind paws showed that the degree of synovial
proliferation and bone destruction was highly significantly reduced in anti–IL-3–treated mice. The amount
of infiltrating cells was significantly decreased, and there
was a trend toward reduced cartilage destruction (P ⫽
0.06) in mice treated with anti–IL-3 (Figure 3). Plasma
titers of anticollagen antibodies were reduced in anti–
IL-3–treated mice on day 37 (Figure 2C); the reduction
Figure 2. Clinical and cellular effects of IL-3 blockade during onset of
collagen-induced arthritis. Mice (n ⫽ 15 per group) were treated with
daily injections of anti–IL-3 or rat IgG from day 21 through day 36
after the first immunization with collagen. A, Arthritis score and
incidence. B, Analysis of synovial tissue from the fore paws on day 37
after the first immunization with collagen. Absolute numbers of
monocytes, neutrophils, and basophils (as identified by flow cytometry) and concentrations of IL-6 and TNF␣ are shown. C, Analysis of
plasma titers of collagen-specific total Ig (plasma dilution 1:100,000)
and IgG1 (plasma dilution 1:5,000) and of peripheral blood leukocyte
subsets on day 37 after the first immunization with collagen. Values are
the mean and SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. OD ⫽ optical density
(see Figure 1 for other definitions).
arthritis was also significantly reduced (by ⬃50%) on
day 37 (Figure 2A).
On day 37, we used the fore paws for analysis of
cells infiltrating the synovial tissue and for measurement
of IL-6 and TNF␣ in the supernatant of the recovered
synovial tissue (500 ␮l/paw). The hind paws were used
for histologic evaluation. The number of monocytes, the
number of basophils, and the total number of CD11b⫹
cells (including monocytes and neutrophils) recovered
per fore paw were significantly reduced in mice treated
with anti–IL-3, as was the level of IL-6 measured in the
Figure 3. Histologic effects of IL-3 blockade during onset of collageninduced arthritis. Mice (n ⫽ 15 per group) were treated as described in
Figure 2, and histologic changes in the lower tarsometatarsal joints
were determined on day 37 after the first immunization with collagen.
A, Sections from a mouse treated with anti–IL-3 and a mouse treated
with rat IgG (control). Low-grade synovial hyperplasia without cartilage or bone destruction is seen in the specimens from the anti–IL-3–
treated mouse, whereas the control specimens exhibit marked bone
destruction and mild cartilage damage, with pronounced synovial
hyperplasia (hematoxylin and eosin stained; magnification ⫻ 40). B,
Summary of histologic scores. Synovial hyperplasia (proliferation),
leukocyte infiltration, cartilage erosion (cartilage), and bone destruction (bone) were scored on a 0–2 scale. Values are the mean and SEM.
ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. See Figure 1 for definitions.
Figure 4. Clinical effects of IL-3 blockade after the onset of collageninduced arthritis. After induction of arthritis, mice were evaluated
daily for clinical disease severity. When the arthritis score was at least
2, mice were randomly assigned to receive daily intraperitoneal
treatment with anti–IL-3 or rat IgG (n ⫽ 10 per group) for 7 days.
Blockade of IL-3 did not reduce the progression of established
arthritis. Values are the mean ⫾ SEM. See Figure 1 for definitions.
was significant for the IgG1 subclass (P ⫽ 0.038), but not
for total Ig (P ⫽ 0.054) or IgG2a (P ⫽ 0.24). FACS
analysis of the peripheral blood on day 37 revealed a
mild but significant reduction in the frequency of basophils, without significant alterations in the frequencies
of neutrophils and monocytes (Figure 2C).
These findings show that IL-3 plays an important
role in the early phase of CIA. In mice treated with
antibodies against IL-3 for only a short period of time
(until day 30), arthritis severity was significantly reduced
at the end of the treatment period but increased thereafter (data not shown), indicating that the effect of IL-3
antibodies is reversible and that treatment should be
continued until the initial inflammation induced by
immunization with collagen has declined.
We next analyzed whether blockade of IL-3 is
able to reduce the progression of already established
arthritis. Mice were immunized twice with 200 ␮g CII
and assessed daily for development of arthritis. When
the arthritis score in an individual mouse was at least 2,
daily IP treatment with either 50 ␮g anti–IL-3 antibody
(mean arthritis score 2.6 before treatment; n ⫽ 10) or
control IgG (mean arthritis score 2.7 before treatment;
n ⫽ 10) was started. As shown in Figure 4, blockade of
IL-3 did not reduce the progression of already established arthritis. These data indicate that IL-3 is not
involved in late progression of arthritis and are consistent with the reduced expression of IL-3 in paws with
severe inflammation (Figure 1).
We also investigated whether administration of
IL-3 during the period of disease onset could increase
the incidence and severity of arthritis. Mice were immu-
nized with 100 ␮g CII on day 0 and day 21. One group
of mice (n ⫽ 21) was treated from day 20 through day 30
with twice-daily IP injection of 100 ng IL-3, while the
control group (n ⫽ 18) was injected with PBS in the
same volume. Injection of IL-3 during the period of
disease onset significantly increased the incidence and
severity of CIA (Figure 5). On day 31 (1 day after the
last injection of IL-3), mice treated with IL-3 showed
significantly increased plasma titers of anticollagen antibodies, a 2-fold increased frequency of peripheral blood
basophils, and almost 5-fold increased plasma levels of
IL-6. There was also a slight, but statistically significant,
increase in the frequency of basophils in PBS-treated
mice from day 19 (1.49% of total leukocytes) to day 24
(1.78% of total leukocytes). Basophilia and increased
plasma IL-6 levels were transient: no significant differences between IL-3– and PBS-treated mice were detectable on day 38 (8 days after the last injection of IL-3).
These data suggest that the restricted availability of IL-3
limits disease onset and progression in DBA/1 mice
immunized with CII and that IL-3 is a diseaseaccelerating factor in the early phase of arthritis.
Figure 5. Exacerbation of established arthritis by interleukin-3 (IL-3).
Mice were treated with twice-daily intraperitoneal injections of 100 ng
IL-3 (n ⫽ 21) or phosphate buffered saline (PBS; n ⫽ 18) from day 20
through day 30 after the first immunization with collagen. A, Arthritis
score and incidence. B, Frequency of basophils in the peripheral blood
(as identified by flow cytometry), plasma titers of collagen-specific
IgG1 (plasma dilution 1:1,000) and IgG2a (plasma dilution 1:2,000),
and plasma levels of IL-6 on day 31 after the first immunization with
collagen. Values are the mean and SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01.
OD ⫽ optical density.
Regulation of IL-3 production by CD4ⴙ T cells.
Activated CD4⫹ T cells are considered to be the main
cellular source of IL-3. However, there is little information on how IL-3 secretion by T cells is regulated. We
therefore investigated which factors up- or downregulate IL-3 production in vitro. Polyclonal activation
of purified CD4⫹ T cells with a soluble antibody against
CD3 and B cells as accessory cells resulted in little
production of IL-3. In contrast, when CD11b⫹ monocytes were used as accessory cells, IL-3 production by
CD4⫹ T cells activated with soluble anti-CD3 was
up-regulated ⬎3-fold (Figure 6A). Addition of the TLR
ligands LPS and CpG DNA markedly enhanced IL-3
secretion by polyclonally activated CD4⫹ T cells in the
presence of accessory B cells or monocytes (Figure 6A).
Stimulation of CD4⫹ T cells and accessory cells with
LPS or CpG DNA in the absence of anti-CD3 did not
result in detectable release of IL-3 (data not shown).
Activation of CD4⫹ T cells with a combination of
antibodies against CD3 and CD28 immobilized on beads
resulted in very high release of IL-3, independent of the
presence of accessory cells or stimulation with LPS or
CpG DNA (Figure 6A).
To confirm that IL-3 is produced by CD4⫹ T
cells and not B cells or monocytes, we measured intracellular levels of IL-3 by flow cytometry (Figure 6B).
CD4⫹ T cells were cultured for 3 days with anti-CD3 in
the presence of LPS-stimulated B cells or LPSstimulated monocytes. Intracellular staining for IL-3 was
detectable only in CD4⫹ T cells, and not in CD4⫺ B
cells or monocytes. Using TLR-4–deficient C3H mice,
we analyzed in greater detail how LPS enhances the
release of IL-3 (Figure 6C). In experiments using accessory B cells that were unable to respond to LPS, release
of IL-3 by CD4⫹ T cells was not enhanced, and when
CD4⫹ T cells from TLR-4–deficient mice were used,
IL-3 production was increased rather than reduced.
These data indicate that LPS increases the release of
IL-3 by CD4⫹ T cells by stimulating accessory cells and
that the level of costimulation provided to CD4⫹ T cells
critically influences expression of IL-3.
Based on our finding that IL-3 levels were downregulated in paws with severe arthritis, we explored
whether cytokines present at high concentrations in the
inflamed joints were able to down-modulate expression
of IL-3 by activated CD4⫹ T cells. CD4⫹ T cells were
activated in the presence of various cytokines (IL-6,
IL-4, IL-1␤, TNF␣, and macrophage inflammatory protein 2 [MIP-2]). Addition of IL-1␤, TNF␣, or MIP-2 had
no effect on the release of IL-3 (data not shown).
However, addition of IL-6 or IL-4 significantly reduced
Figure 6. Regulation of IL-3 secretion from CD4⫹ T cells. A, Quantification of IL-3 in the supernatant of CD4⫹ T cells cultured with
CD11b⫹ cells and anti-CD3, with CD19⫹ B cells and anti-CD3, or
with anti-CD3/anti-CD28–coated beads. Lipopolysaccharaide (LPS),
CpG DNA, or anti-CD3/anti-CD28–coated beads were added as
indicated. ⴱⴱ ⫽ P ⬍ 0.01 versus control. B, Flow cytometric detection
of intracellular IL-3 in CD4⫹ T cells cultured with CD11b⫹ cells,
anti-CD3, and LPS (upper panels) or with B cells, anti-CD3, and LPS
(lower panels). After 3 days, intracellular staining of cells was performd with an antibody against IL-3 (left panels) or with an isotype
control antibody (right panels). Number shown in each compartment is
the percentage of positive cells. C, Analysis of the mechanism of the
effect of LPS on IL-3 expression by CD4⫹ T cells. CD4⫹ T cells from
wild-type (WT) mice or from Toll-like receptor 4–deficient C3H mice
were stimulated with anti-CD3 and LPS in the presence of B cells from
WT mice or from C3H mice as indicated. The results showed that LPS
up-regulates IL-3 expression in CD4⫹ T cells by acting on CD19⫹ B
cells. ⴱⴱ ⫽ P ⬍ 0.01 versus cells from WT mice in the presence of B
cells from WT mice. D–F, Effects of IL-4 and IL-6 on the release of
IL-3 from activated CD4⫹ T cells. CD4⫹ T cells were cultured with
CD11b⫹ cells and anti-CD3 or with CD11b⫹ cells, anti-CD3, and LPS
(D), with CD19⫹ B cells and anti-CD3 or with CD19⫹ B cells,
anti-CD3, and LPS (E), or without stimulation or with anti-CD3/antiCD28–coated beads (F). IL-4, IL-6, or both were added as indicated.
The results showed that IL-4 and IL-6 suppress IL-3 release from
activated CD4⫹ cells. Values in A and C–F are the mean and SEM. In
D, ⴱⴱ ⫽ P ⬍ 0.01 versus the corresponding control group without IL-4
or IL-6 treatment. In E and F, ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, versus
control. See Figure 1 for other definitions.
the release of IL-3 by activated CD4⫹ T cells, independent of the costimulatory factors used for T cell activa-
tion (unstimulated monocytes, LPS-stimulated B cells,
or monocytes and anti-CD3/anti-CD28–coated beads)
(Figures 6D–F). IL-6 and IL-4 in combination exerted a
synergistic effect and resulted in very pronounced blockade of IL-3 production (Figures 6D and E). IL-6 and
IL-4 also reduced IL-3 secretion from CD4⫹ T cells
activated only with anti-CD3/anti-CD28–coated beads,
indicating that they act on CD4⫹ T cells and not on
accessory cells.
We have shown in this study that IL-3 is an
important factor in the early development of CIA but
does not contribute to the late phase of the disease.
Blockade of IL-3 with a monoclonal antibody during the
period of onset of arthritis (days 21–36) markedly reduced clinical and histologic signs of arthritis and numbers of infiltrating cells, while IL-3 blockade in mice with
established arthritis had no effect. One possible explanation for the lack of effect of IL-3 antibodies in
advanced arthritis is the low level of production of IL-3
in highly inflamed joints. Joints with a clinical score of 3
or 4 exhibited very low levels of IL-3 but high levels of
IL-6, which proved to be a potent suppressor of IL-3
production by CD4⫹ T cells. Alternatively, the low level
of IL-3 in highly inflamed joints might be due to
increased IL-3 consumption by infiltrating cells.
In the early phase of arthritis, IL-3 is an important disease-accelerating factor. Its restricted availability
seemed to limit arthritis development, and administration of IL-3 markedly increased arthritis severity. IL-3 is
produced systemically by CD4⫹ T cells in the spleen as
well as locally by cells in the synovial tissue, and it may
act systemically as well as locally to aggravate early
Blockade of IL-3 resulted in a significantly reduced frequency of basophils in the peripheral blood
and reduced plasma titers of antibodies against collagen
measured at the end of anti–IL-3 treatment. We have
recently shown that activated basophils contribute significantly to the development of a humoral memory
immune response. Activated basophils release soluble
factors (mainly IL-6) and provide cell contact–
dependent factors that promote proliferation of B cells
and their differentiation into plasma cells in vitro and in
vivo (23). It is tempting to speculate that IL-3 aggravates
early arthritis by increasing the number of basophils in
the peripheral blood, by activation of basophils, and by
increasing plasma titers of antibodies against collagen.
The present findings demonstrate that IL-3 is a very
potent activator of basophils and markedly prolongs
their survival in vitro, and that administration of IL-3 in
vivo results in increased plasma levels of anticollagen
antibodies, 2-fold increased numbers of basophils in the
peripheral blood, and 5-fold increased plasma IL-6
levels. However, it must be kept in mind that plasma
levels of anticollagen antibodies do not correlate very
well with severity of arthritis and that IL-3 also has
several other target cells (e.g., monocytes and dendritic
cells, as noted above) that may contribute to the development of arthritis.
Local effects of IL-3 in the joint may include
increasing the release of IL-1 from monocytes and
inducing the development of osteoclasts (14,17). In this
study we assessed the presence of basophils and mast
cells in the inflamed synovial tissue of mice with CIA
and found a rather high frequency of basophils (⬃0.4%
of total infiltrating leukocytes), while mast cells were
almost undetectable. Due to conflicting results in studies
using different models of arthritis and different strains
of mast cell–deficient mice (24–26), the role of mast
cells in arthritis development is currently unclear. Our
data suggest that the proarthritogenic effects of IL-3
may be mediated in part by activation of basophils and
are consistent with previous data demonstrating aggravation of arthritis by application of anti-IgE or antiCCR2 antibodies, both of which are known to activate
basophils (22,27).
To better understand why IL-3 expression in the
synovial tissue correlates negatively with the severity of
arthritis, and how the expression of IL-3 is regulated, we
performed in vitro assays with CD4⫹ T cells, considered
to be the main cellular source of IL-3. Apart from results
of promoter analysis and findings of suppression by
cyclosporin A, available data on the regulation of IL-3
production by T cells are very limited (28). IL-3 expression has been observed in both Th1 and Th2 cells (29).
We showed in this study that production of IL-3 by
CD4⫹ T cells is dependent on the level of costimulation
provided to the CD4⫹ T cells. Activation of CD4⫹ T
cells with anti-CD3 in the presence of freshly isolated B
cells resulted in little expression of IL-3, whereas the
presence of monocytes, or B cells and monocytes, activated with ligands for TLR-4 or TLR-9 markedly upregulated production of IL-3 by CD4⫹ T cells. In
experiments using intracellular cytokine staining and
cells from TLR-4–deficient C3H mice we showed that
LPS up-regulated IL-3 production in CD4⫹ T cells by
acting on B cells, but not directly on T cells. It is known
that LPS aggravates CIA, whereas blockade of TLR-4
improves it (30,31).
We also analyzed how proinflammatory cytokines present in inflamed joints modulate the secretion
of IL-3 by polyclonally activated CD4⫹ T cells and
found that IL-6 and IL-4 down-regulate IL-3 expression
by CD4⫹ T cells, whereas IL-1␤, TNF␣, and MIP-2 do
not. Treatment with a combination of IL-4 and IL-6
almost completely prevented IL-3 production induced
by monocytes or LPS-stimulated B cells and monocytes.
Since basophils produce large amounts of IL-4 and IL-6,
one could postulate that there is negative feedback
between activation of basophils and IL-3 production by
T cells. Restimulation of total splenocytes and splenocytes depleted of specific leukocyte subsets with CII
confirmed that IL-3 is produced almost exclusively by
CD4⫹ T cells and requires the presence of antigenpresenting CD11b⫹ monocytes. B cells do not support
IL-3 production by CD4⫹ T cells in the absence of
In summary, our data demonstrate that IL-3 is
involved in the development of CIA. IL-3 may thus
represent a novel therapeutic target for early forms of
rheumatoid arthritis, or for maintenance therapy for
prevention of flares.
We thank K. Schmidbauer and D. Lochbaum for
excellent technical assistance.
Dr. Mack had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy of the
data analysis.
Study design. Brühl, Mack.
Acquisition of data. Brühl, Cihak, Niedermeier, Denzel, Rodriguez
Gomez, Talke, Goebel, Plachý, Stangassinger, Mack.
Analysis and interpretation of data. Brühl, Cihak, Niedermeier,
Denzel, Rodriguez Gomez, Talke, Goebel, Stangassinger, Mack.
Manuscript preparation. Brühl, Mack.
Statistical analysis. Brühl, Mack.
1. Miyajima A, Kitamura T, Harada N, Yokota T, Arai K. Cytokine
receptors and signal transduction. Annu Rev Immunol 1992;10:
2. Hara T, Miyajima A. Two distinct functional high affinity receptors for mouse interleukin-3 (IL-3). EMBO J 1992;11:1875–84.
3. Lantz CS, Boesiger J, Song CH, Mach N, Kobayashi T, Mulligan
RC, et al. Role for interleukin-3 in mast-cell and basophil development and in immunity to parasites. Nature 1998;392:90–3.
4. Ihle JN, Keller J, Oroszlan S, Henderson LE, Copeland TD, Fitch
F, et al. Biologic properties of homogeneous interleukin 3. I.
Demonstration of WEHI-3 growth factor activity, mast cell growth
factor activity, p cell-stimulating factor activity, colony-stimulating
factor activity, and histamine-producing cell-stimulating factor
activity. J Immunol 1983;131:282–7.
Kirshenbaum AS, Goff JP, Dreskin SC, Irani AM, Schwartz LB,
Metcalfe DD. IL-3-dependent growth of basophil-like cells and
mastlike cells from human bone marrow. J Immunol 1989;142:
Valent P, Schmidt G, Besemer J, Mayer P, Zenke G, Liehl E, et al.
Interleukin-3 is a differentiation factor for human basophils. Blood
Dvorak AM, Seder RA, Paul WE, Morgan ES, Galli SJ. Effects of
interleukin-3 with or without the c-kit ligand, stem cell factor, on
the survival and cytoplasmic granule formation of mouse basophils
and mast cells in vitro. Am J Pathol 1994;144:160–70.
Haak-Frendscho M, Arai N, Arai K, Baeza ML, Finn A, Kaplan
AP. Human recombinant granulocyte-macrophage colony-stimulating factor and interleukin 3 cause basophil histamine release.
J Clin Invest 1988;82:17–20.
MacDonald SM, Schleimer RP, Kagey-Sobotka A, Gillis S, Lichtenstein LM. Recombinant IL-3 induces histamine release from
human basophils. J Immunol 1989;142:3527–32.
Kurimoto Y, de Weck AL, Dahinden CA. Interleukin 3-dependent
mediator release in basophils triggered by C5a. J Exp Med
Le Gros G, Ben-Sasson SZ, Conrad DH, Clark-Lewis I, Finkelman
FD, Plaut M, et al. IL-3 promotes production of IL-4 by splenic
non-B, non-T cells in response to Fc receptor cross-linkage.
J Immunol 1990;145:2500–6.
Yoshimoto T, Tsutsui H, Tominaga K, Hoshino K, Okamura H,
Akira S, et al. IL-18, although antiallergic when administered with
IL-12, stimulates IL-4 and histamine release by basophils. Proc
Natl Acad Sci U S A 1999;96:13962–6.
Frendl G, Beller DI. Regulation of macrophage activation by IL-3.
I. IL-3 functions as a macrophage-activating factor with unique
properties, inducing Ia and lymphocyte function-associated antigen-1 but not cytotoxicity. J Immunol 1990;144:3392–9.
Frendl G, Fenton MJ, Beller DI. Regulation of macrophage
activation by IL-3. II. IL-3 and lipopolysaccharide act synergistically in the regulation of IL-1 expression. Immunol 1990;144:
Buelens C, Bartholome EJ, Amraoui Z, Boutriaux M, Salmon I,
Thielemans K, et al. Interleukin-3 and interferon ␤ cooperate to
induce differentiation of monocytes into dendritic cells with potent
helper T-cell stimulatory properties. Blood 2002;99:993–8.
Ebner S, Hofer S, Nguyen VA, Furhapter C, Herold M, Fritsch P,
et al. A novel role for IL-3: human monocytes cultured in the
presence of IL-3 and IL-4 differentiate into dendritic cells that
produce less IL-12 and shift Th cell responses toward a Th2
cytokine pattern. J Immunol 2002;168:6199–207.
Barton BE, Mayer R. IL-3 induces differentiation of bone marrow
precursor cells to osteoclast-like cells. J Immunol 1989;143:
Toyosaki-Maeda T, Takano H, Tomita T, Tsuruta Y, MaedaTanimura M, Shimaoka Y, et al. Differentiation of monocytes into
multinucleated giant bone-resorbing cells: two-step differentiation
induced by nurse-like cells and cytokines. Arthritis Res 2001;3:
Firestein GS, Xu WD, Townsend K, Broide D, Alvaro-Gracia J,
Glasebrook A, et al. Cytokines in chronic inflammatory arthritis. I.
Failure to detect T cell lymphokines (interleukin 2 and interleukin
3) and presence of macrophage colony-stimulating factor (CSF-1)
and a novel mast cell growth factor in rheumatoid synovitis. J Exp
Med 1988;168:1573–86.
Alvaro-Gracia JM, Zvaifler NJ, Firestein GS. Cytokines in chronic
inflammatory arthritis. V. Mutual antagonism between interferon-␥ and tumor necrosis factor-␣ on HLA-DR expression, proliferation, collagenase production, and granulocyte macrophage
colony-stimulating factor production by rheumatoid arthritis synoviocytes. J Clin Invest 1990;86:1790–8.
Yamada R, Tanaka T, Unoki M, Nagai T, Sawada T, Ohnishi Y,
et al. Association between a single-nucleotide polymorphism in the
promoter of the human interleukin-3 gene and rheumatoid arthritis in Japanese patients, and maximum-likelihood estimation of
combinatorial effect that two genetic loci have on susceptibility to
the disease. Am J Hum Genet 2001;68:674–85.
Bruhl H, Cihak J, Plachy J, Kunz-Schughart L, Niedermeier M,
Denzel A, et al. Targeting of Gr-1⫹,CCR2⫹ monocytes in collagen-induced arthritis. Arthritis Rheum 2007;56:2975–85.
Denzel A, Maus UA, Gomez MR, Moll C, Niedermeier M, Winter
C, et al. Basophils enhance immunological memory responses. Nat
Immunol 2008;9:733–42.
Lee DM, Friend DS, Gurish MF, Benoist C, Mathis D, Brenner
MB. Mast cells: a cellular link between autoantibodies and inflammatory arthritis. Science 2002;297:1689–92.
Nigrovic PA, Binstadt BA, Monach PA, Johnsen A, Gurish M,
Iwakura Y, et al. Mast cells contribute to initiation of autoantibody-mediated arthritis via IL-1. Proc Natl Acad Sci U S A
Zhou JS, Xing W, Friend DS, Austen KF, Katz HR. Mast cell
deficiency in KitW-sh mice does not impair antibody-mediated
arthritis. J Exp Med 2007;204:2797–802.
Bruhl H, Cihak J, Schneider MA, Plachy J, Rupp T, Wenzel I, et
al. Dual role of CCR2 during initiation and progression of
collagen-induced arthritis: evidence for regulatory activity of
CCR2⫹ T cells. J Immunol 2004;172:890–8.
Hawwari A, Burrows J, Vadas MA, Cockerill PN. The human IL-3
locus is regulated cooperatively by two NFAT-dependent enhancers that have distinct tissue-specific activities. J Immunol 2002;169:
Stevens TL, Bossie A, Sanders VM, Fernandez-Botran R, Coffman RL, Mosmann TR, et al. Regulation of antibody isotype
secretion by subsets of antigen-specific helper T cells. Nature
Yoshino S, Sasatomi E, Mori Y, Sagai M. Oral administration of
lipopolysaccharide exacerbates collagen-induced arthritis in mice.
J Immunol 1999;163:3417–22.
Abdollahi-Roodsaz S, Joosten LA, Roelofs MF, Radstake TR,
Matera G, Popa C, et al. Inhibition of Toll-like receptor 4 breaks
the inflammatory loop in autoimmune destructive arthritis. Arthritis Rheum 2007;56:2957–67.
DOI 10.1002/art.24495
Clinical Images: Gout attack and stiff knee in an airline passenger in economy class
The patient, a 63-year-old man with a history of chronic gout, which was in remission at the time, took a 2-hour plane ride in
economy class and developed acute gout in his right knee. One week later, he presented with contracture of the knee and diminished
extension. Arthroscopy (A) revealed an easily detachable chalky deposit on the surface of the hyaline cartilage. Adhesions in the
suprapatellar pouch, a “starry sky” appearance, and hemorrhagic areas (B) were also observed. These findings suggest that
immobility of the knees in a sitting position might lead to clinical flare of gout.
Angel Checa, MD
Carolyn Riester O’Connor, MD
Drexel University College of Medicine
Philadelphia, PA
Без категории
Размер файла
355 Кб
induced, arthritis, role, interleukin, importance, phase, collagen, early
Пожаловаться на содержимое документа