Inducible costimulator ligand regulates bleomycin-induced lung and skin fibrosis in a mouse model independently of the inducible costimulatorinducible costimulator ligand pathway.

ARTHRITIS & RHEUMATISM
Vol. 62, No. 6, June 2010, pp 1723–1732
DOI 10.1002/art.27428
© 2010, American College of Rheumatology
Inducible Costimulator Ligand Regulates Bleomycin-Induced
Lung and Skin Fibrosis in a Mouse Model Independently of the
Inducible Costimulator/Inducible Costimulator Ligand Pathway
Chihiro Tanaka,1 Manabu Fujimoto,1 Yasuhito Hamaguchi,1 Shinichi Sato,2
Kazuhiko Takehara,1 and Minoru Hasegawa1
Objective. Systemic sclerosis is a connective tissue
disease characterized by fibrosis of the skin and internal
organs, including the lungs. Inducible costimulator
(ICOS), which is expressed on activated T cells, and its
ligand ICOSL, which is expressed on antigen-presenting
cells, have been considered a single receptor–ligand
pair. Although the ICOS/ICOSL pathway is known to
play various roles in adaptive immunity, its roles in
innate immunity and tissue fibrosis remain unknown.
Methods. We assessed the roles of ICOS and
ICOSL in tissue fibrosis by administering bleomycin
intratracheally or intradermally into mice deficient in
ICOS and/or ICOSL. Tissue fibrosis was evaluated by
histologic or biochemical examination.
Results. ICOS deficiency attenuated the lung and
skin fibrosis, whereas ICOSL deficiency aggravated it.
Mice deficient in both ICOS and ICOSL exhibited
accelerated fibrosis, reflecting a dominant role of
ICOSL over ICOS in this model. Interestingly, ICOSL
expression on macrophages and B cells derived from
bronchoalveolar lavage fluid was significantly elevated
in ICOS-deficient mice as compared with wild-type mice
during this process. Thus, the levels of ICOSL expres-
sion on B cells and macrophages were inversely associated with the severity of tissue fibrosis.
Conclusion. Our results indicate that ICOSL
expression on antigen-presenting cells plays a previously unknown regulatory role during the development
of bleomycin-induced tissue fibrosis that is independent
of the ICOS/ICOSL pathway. Further studies will be
needed to clarify the roles of ICOS and ICOSL in the
development of systemic sclerosis.
Systemic sclerosis (SSc) is an autoimmune connective tissue disease characterized by excessive extracellular matrix deposition in the skin, lungs, and other
internal organs (1,2). Interstitial lung disease is a major
cause of mortality and morbidity in patients with SSc. A
growing body of evidence suggests that in patients with
SSc, overproduction of extracellular matrix components
by activated fibroblasts results from complex interactions between endothelial cells, lymphocytes, macrophages, and fibroblasts, and via a number of mediators,
including cytokines, chemokines, and growth factors
(1,2).
Bleomycin (BLM)–induced lung fibrosis is widely
used as an established animal model of pulmonary
fibrosis (3,4). Intratracheal administration of BLM induces acute alveolitis and interstitial inflammation.
Yamamoto and colleagues (5) have established a new
mouse model of skin fibrosis induced by daily intradermal injections of BLM. In this model, lung fibrosis also
develops when sufficient amounts of BLM have been
administered (6), thereby making this one of most useful
animal models of SSc.
Inducible costimulator (ICOS) is the third member of the CD28 family of costimulatory molecules and
is induced on the cell surface following T cell activation
(7–9). The ligand of ICOS, or ICOSL (also called B7h,
Supported by grants-in-aid from the Ministry of Education,
Science, and Culture of Japan and by funds for research of intractable
diseases from the Ministry of Health, Labor, and Welfare of Japan.
1
Chihiro Tanaka, MD, Manabu Fujimoto, MD, Yasuhito
Hamaguchi, MD, PhD, Kazuhiko Takehara, MD, PhD, Minoru Hasegawa, MD, PhD: Kanazawa University Graduate School of Medical
Science, Kanazawa, Japan; 2Shinichi Sato, MD, PhD: Tokyo University, Tokyo, Japan.
Address correspondence and reprint requests to Minoru
Hasegawa, MD, PhD, Department of Dermatology, Kanazawa University Graduate School of Medical Science, 13-1 Takaramachi, Kanazawa, Ishikawa 920-8641, Japan. E-mail: minoruha@derma.m.
kanazawa-u.ac.jp.
Submitted for publication August 19, 2009; accepted in
revised form February 18, 2010.
1723
1724
TANAKA ET AL
B7RP-1, LICOS, and GL50), is weakly expressed on
antigen-presenting cells in the steady state and is upregulated after activation of these cells (8,10). The
ICOS/ICOSL pathway promotes T cell activation, differentiation, and effector responses, and T cell–
dependent B cell responses. ICOS-mediated costimulation of T cells leads predominantly to the production of
effector cytokines, such as interleukin-4 (IL-4) and
IL-10, and to a lesser extent, IL-2, interferon-␥ (IFN␥),
and tumor necrosis factor ␣ (TNF␣) (11), thereby
playing a more important role in Th2 responses than in
Th1 responses (7,12–14). Furthermore, recent studies
demonstrated that ICOS influences the expansion of
follicular T cells, Th17 cells, and regulatory T cells
(15–17). Since ICOSL–/– and ICOS–/– mice display similar defects in humoral immunity, ICOS and ICOSL
have been considered a single receptor–ligand pair
(18–20).
It is generally accepted that ICOS/ICOSL signaling is critical for adaptive immunity, including autoimmunity, allergy, infectious diseases, and transplantation
(20–23). However, ICOSL is constitutively expressed not
only on B cells, but also on macrophages and dendritic
cells, which contribute to innate immunity (8,10). Therefore, we hypothesized that on these cells, ICOSL may be
contributing to innate immunity independently of the
ICOS/ICOSL axis. Since tissue fibrosis, including the
BLM-induced scleroderma model, involves a strong
innate immune component, we applied this model to
mice deficient in ICOS and ICOSL to investigate the
roles of these molecules in modulating innate and
adaptive immunity.
MATERIALS AND METHODS
Mice. ICOS–/– mice and ICOSL–/– mice were purchased from The Jackson Laboratory. All mice were backcrossed 8–10 generations onto mice of the C57BL/6 genetic
background. Mating these ICOS–/– mice with ICOSL–/– mice
generated ICOS⫹/–ICOSL⫹/– mice. Then, we generated
ICOS–/–ICOSL–/– mice by crossing the ICOS⫹/–ICOSL⫹/– parents. To verify the ICOS or ICOSL genotype, we conducted
polymerase chain reaction amplification of each gene using
genomic DNA from each mouse. All mice were housed in a
specific pathogen–free barrier facility and screened regularly
for pathogens. Female mice 8–10 weeks of age were used in the
experiments. The Committee on Animal Experimentation at
Kanazawa University Graduate School of Medical Science
approved all studies and procedures.
Intratracheal treatment with BLM. BLM sulfate (Nippon Kayaku) was administered to mice that had been anesthetized by inhalation of diethyl ether. Using aseptic techniques, a
single incision was made at the neck, and the muscle covering
the trachea was snipped to expose the tracheal rings. A single
intratracheal instillation of BLM sulfate (8 mg/kg [7.68 units/
kg]) in 200 ␮l of sterile saline was performed using a 27-gauge
needle.
Intradermal treatment with BLM. BLM was dissolved
in sterile saline at a concentration of 1 mg/ml. Three hundred
microliters of BLM or saline was injected intradermally into
the shaved backs (the para-midline, lower back region) of the
mice with a 27-gauge needle, as described previously (5).
Injections were given daily for 4 weeks.
Histologic examination of lung and skin fibrosis.
Whole lungs were inflated, fixed with 3.5% paraformaldehyde,
and embedded in paraffin. Sections (6 ␮m in thickness) were
stained with hematoxylin and eosin (H&E) to evaluate alveolitis or with Masson’s trichrome to identify collagen deposition in the lungs. The severity of lung inflammation was
determined by a semiquantitative scoring system as previously
described (24). Briefly, lung fibrosis in randomly chosen fields
of sections from the left middle lobe examined at 100⫻
magnification was graded on a scale of 0 (normal lung) to 8
(total fibrous obliteration of fields). All sections were scored
independently by 2 investigators (CT and MH) in a blinded
manner.
All skin sections were taken from the BLM-injected
region of the lower back and were obtained as full-thickness
sections extending down to the body wall musculature. Skin
samples were fixed in formalin, dehydrated, embedded in
paraffin, and used for immunostaining. Six-micrometer sections were stained with H&E or with Azan-Mallory reagents to
identify collagen deposition in the skin. Dermal thickness,
which was defined as the thickness of skin from the top of the
granular layer to the junction between the dermis and intradermal fat, was evaluated independently by 2 investigators (CT
and MH) in a blinded manner. Using the free-hand tool in the
Photoshop Elements 3.0 software package (Adobe Systems),
specific staining with Masson’s trichrome and ␣-smooth muscle
actin (␣-SMA) was quantified in the whole lung and skin
sections.
Hydroxyproline assay. Total lung or dorsal skin tissue
was freeze-dried overnight, and the hydroxyproline content
was determined as previously described (3). Briefly, samples
were hydrolyzed in 6N HCl for 20 hours at 105°C, desiccated
overnight, and dissolved in 1 ml of citrate/acetate buffer (pH
6.0). Chloramine T solution was added to each optimally
diluted sample and left at room temperature for 20 minutes.
Next, Ehrlich’s solution was added to each sample, incubated
for 15 minutes at 65°C, and the optical density at 550 nm was
read. Hydroxyproline standard solutions of 0–20 ␮g/ml were
used to construct a standard curve.
Determination of collagen contents in lung tissue
sections. Approximately 15 ␮m–thick sections were obtained
from the paraffin-embedded lung tissue samples. Groups of
10–20 sections were deparaffinized after incubation with xylol,
xylol:ethanol (1:1), ethanol, water:ethanol (1:1), and water.
Individual samples were placed in small test tubes and covered
with 0.2 ml of a saturated solution of picric acid in distilled
water that contained 0.1% fast green FCF and 0.1% sirius red
F3BA. The samples were rinsed several times with distilled
water until the fluid was colorless. One milliliter of 0.1N NaOH
in absolute methanol (1:1 [volume/volume]) was added, and
the eluted color was read in a spectrophotometer at 540 and
605 nm. This method is based on the selective binding of Sirius
ICOSL REGULATION OF BLEOMYCIN-INDUCED MURINE LUNG AND SKIN FIBROSIS
1725
Figure 1. Lung fibrosis induced by intratracheal treatment with bleomycin (BLM). ICOS–/–, ICOSL–/–, and wild-type (WT) mice were treated with
a single intratracheal injection of BLM and were evaluated through day 21. A, Survival rates in ICOS–/–, ICOSL–/–, and wild-type mice. Results were
compiled from at least 21 mice per group. B–D, Analysis of the lungs for determination of lung fibrosis scores (B), hydroxyproline content (C), and
collagen content (semiquantitative) (D). Values are the mean and SEM of 5 mice per group. E, Quantitative results of immunohistochemical
stainings shown in F. Values are the mean and SEM area of specific antibody staining with Masson’s trichrome and ␣-smooth muscle actin (␣-SMA)
over the total stained area in 5 nonconsecutive whole lung sections from 5 mice per group. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. F, Representative
photomicrographs of lung tissue sections obtained from each BLM-treated mouse strain and from a saline-treated wild-type mouse (day 21), as well
as sections obtained from dying mice (2 wild-type, 1 ICOS⫺/⫺, and 6 ICOSL–/– mice examined; days 10–15). Sections were stained with hematoxylin
and eosin (H&E) or with Masson’s trichrome (original magnification ⫻ 100).
red F3BA and fast green FCF to collagens and noncollagenous
proteins, respectively (25).
Immunohistochemical staining. Tissues were harvested before and on days 7, 14, 21, and 28 during BLM
treatment and were assessed for the numbers of infiltrating T
cells, B cells, macrophages, and neutrophils. Deparaffinized
sections were treated with endogenous peroxidase blocking
reagent and proteinase K (both from DakoCytomation) for 6
minutes at room temperature. Sections were then incubated
with rat monoclonal antibodies specific for anti–␣-SMA
(Sigma-Aldrich), macrophages (clone F4/80; American Type
Culture Collection), myeloperoxidase (MPO; NeoMarkers),
CD3 (Dainippon Pharmaceutical), or mouse B220 (BD Biosciences). Rat IgG (SouthernBiotech) was used as a control for
nonspecific staining. Sections were then incubated sequentially
(for 20 minutes at 37°C) with a biotinylated rabbit anti-rat IgG
secondary antibody (Vectastain ABC method; Vector) or a
biotinylated goat anti-rabbit IgG secondary antibody (BD
Biosciences), followed by horseradish peroxidase–conjugated
avidin–biotin complexes. Sections were washed 3 times with
phosphate buffered saline (PBS) between incubations, developed with 3,3⬘-diaminobenzidine tetrahydrochloride and hydrogen peroxide, and then counterstained with methyl green.
Stained cells were counted in 5 random grids under
high magnification (400⫻) using a light microscope. Each
section was examined and scored independently by 2 investi-
gators (CT and MH) in a blinded manner. The mean score was
used for analysis.
Determination of cytokine concentrations in tissues.
Samples of whole lungs were homogenized in 600 ␮l of lysis
buffer (10 mmoles/liter of PBS, 0.1% sodium dodecyl sulfate,
1% Nonidet P40, 5 mmoles/liter of EDTA containing a
complete protease inhibitor mixture; Roche Diagnostics) to
extract proteins. Homogenates were centrifuged at 15,000 revolutions per minute for 15 minutes at 4°C to remove debris (26).
The concentrations of TNF␣, IFN␥, monocyte chemotactic protein 1 (MCP-1), IL-6, IL-10, and IL-12 in supernatants were determined by using a BD Cytometric Bead Array
mouse inflammation kit (BD Biosciences) according to the
manufacturer’s protocol. Flow cytometry analysis was performed using a FACSCalibur flow cytometer (BD Biosciences).
The levels of macrophage inflammatory protein 1␣
(MIP-1␣) (Quantikine Immunoassay) and IL-4 (DuoSet)
(both from R&D Systems) in supernatants were measured by
enzyme-linked immunosorbent assay according to the manufacturer’s instructions. Total protein in the supernatant was
measured with a commercial kit (BCA Protein Assay kit;
Thermo Fischer Scientific).
Reverse transcription–polymerase chain reaction (RTPCR). Whole lungs were harvested 7 days after intradermal
treatment with BLM. Total RNA was isolated from frozen
1726
TANAKA ET AL
Figure 2. Skin and lung fibrosis induced by intradermal treatment with bleomycin (BLM). ICOS–/–, ICOSL–/–, and wild-type (WT) mice were
treated with daily intradermal injections of BLM for 28 days. A, Lung fibrosis was assessed by determining lung fibrosis scores, hydroxyproline
content, and percentage of staining with ␣-smooth muscle actin (␣-SMA). B, Skin fibrosis was assessed by determining dermal thickness,
hydroxyproline content, and percentage of staining with ␣-SMA. Values in A and B are the mean and SEM of 5 mice per group. ⴱ ⫽ P ⬍ 0.05;
ⴱⴱ ⫽ P ⬍ 0.01. C, Representative photomicrographs of lung tissue sections obtained from each BLM-treated mouse strain and from a saline-treated
wild-type mouse. Sections were stained with hematoxylin and eosin (H&E) or with Masson’s trichrome. D, Representative photomicrographs of skin
tissue sections obtained from each experimental group. Sections were stained with H&E or with Azan-Mallory stain. Arrows indicate the dermal
thickness. (Original magnification ⫻ 100.)
lung specimens using RNeasy spin columns (Qiagen) and
digested with DNase I (Qiagen) to remove chromosomal DNA
in accordance with the manufacturer’s protocols. Total RNA
was reverse-transcribed to a complementary DNA using a
reverse transcription system with random hexamers (Promega). Cytokine messenger RNAs (mRNA) were analyzed
using real-time RT-PCR quantification, according to the manufacturer’s instructions (Applied Biosystems). Sequencespecific primers and probes were designed with predeveloped
TaqMan assay reagents (Applied Biosystems). Real-time RTPCR (40 cycles of denaturation at 92°C for 15 seconds and
annealing at 60°C for 60 seconds) was performed on an ABI
Prism 7000 sequence detector (Applied Biosystems). GAPDH
was used to normalize the mRNA. The relative expression of
real-time RT-PCR products was determined according to the
⌬⌬Ct method (27) to compare target gene and GAPDH
mRNA expression. One of the control samples was chosen as
a calibrator sample.
Flow cytometric analysis. Splenic macrophages and B
cells were purified from untreated mice of each strain as
described above. These cells were then used for flow cytometric analysis.
Bronchoalveolar lavage (BAL) cells were prepared as
described elsewhere (28). Briefly, mice were killed on days 7,
14, and 21 after BLM instillation. BAL fluid was collected as
follows: 1 ml of saline was instilled 3 times and withdrawn from
the lungs via an intratracheal cannula.
For 2-color or 3-color immunofluorescence analyses,
BAL cells were stained with the following antibodies: CD4
(RM4-5; BD Biosciences), B220 (RA36B2; BD Biosciences),
CD14 (Sa2-8; eBioscience), ICOS (7E17G9; eBioscience), and
ICOSL (HK5.3; eBioscience). BAL cells with the forward and
side light-scatter properties of lymphocytes were analyzed on a
FACScan flow cytometer (BD Biosciences). Positive and negative populations of cells were determined using unreactive
isotype-matched monoclonal antibodies (BD Biosciences) as
controls for background staining.
Statistical analysis. The Mann-Whitney U test was
used to determine the level of significance of differences in the
sample means. The Bonferroni test was used for multiple
comparisons.
RESULTS
Amelioration of intratracheal BLM–induced
lung fibrosis by ICOS deficiency, but aggravation by
ICOSL deficiency. To assess whether the ICOS/ICOSL
signaling pathway contributed to the development of
lung fibrosis, we first assessed intratracheal treatment of
BLM in ICOS–/–, ICOSL–/–, and wild-type mice. Lungs
were removed from the animals on day 21 following
saline or BLM treatment. As demonstrated by the
survival rates, ICOS deficiency resulted in a significantly
lower mortality rate than was observed for wild-type
mice (Figure 1A). Unexpectedly, ICOSL–/– mice demonstrated significantly higher mortality rates as compared with wild-type mice.
ICOSL REGULATION OF BLEOMYCIN-INDUCED MURINE LUNG AND SKIN FIBROSIS
1727
Figure 3. Inflammatory cell infiltration in the skin and lungs induced by daily intradermal injections of bleomycin (BLM). ICOS–/–, ICOSL–/–, and
wild-type (WT) mice were treated with daily intradermal injections of BLM for 28 days. The numbers of F4/80⫹ macrophages, B220⫹ B cells, CD3⫹
T cells, and myeloperoxidase (MPO)–positive neutrophils per high-power field (hpf) in the lungs (A) and skin (B) were counted at the indicated
times. Values are the mean and SEM of 5 mice per group. ⴱ ⫽ P ⬍ 0.05.
Pathologic findings and fibrosis scores in H&Eand Masson’s trichrome–stained section also revealed
that the loss of ICOS significantly resolved the lung
fibrosis, whereas the loss of ICOSL aggravated it (Figures 1B and F). In addition, visual inspection of fibrotic
tissue and myofibroblasts immunostained with Masson’s
trichrome and ␣-SMA, respectively, demonstrated a
similar tendency in each mouse strain (Figure 1E).
Quantification of the hydroxyproline and collagen contents in the lungs confirmed the histologic findings
(Figures 1C and D).
Since there may be other relevant causes of death
in mutant mice, we examined the histologic features of
the lungs in dying mice (2 wild-type, 1 ICOS⫺/⫺, and 6
ICOSL–/– mice). In both the wild-type and ICOSL–/–
mice, the lungs had a characteristically fibrotic appearance, with disrupted alveolar architecture replaced by
continuous cellular interstitial connective tissue (Figure
1F), which indicates that lung injury was the direct cause
of death in these mice. Thus, ICOS deficiency reduced
the lung fibrosis that was induced by intratracheal
treatment with BLM, whereas ICOSL deficiency aggravated it.
Opposing effects of ICOS and ICOSL deficiency
in skin and lung fibrosis induced by intradermal BLM
treatment. As reported previously (5,6), daily intradermal administration of BLM induces skin and lung fibro-
sis by day 28 in wild-type mice. In these experiments, we
again observed that ICOS loss significantly decreased
the intradermal BLM–induced lung fibrosis, but ICOSL
loss exacerbated it (Figures 2A and C). In addition, the
dermal thickness and histopathologic features noted on
H&E-and Azan-Mallory–stained sections, respectively,
revealed significantly ameliorated skin fibrosis in mice
lacking ICOS (Figures 2B and D). In contrast, ICOSL
deficiency significantly exacerbated skin fibrosis. The
results of measurements of hydroxyproline content in
the skin were consistent with the histologic findings
(Figure 2B). Thus, the tissue pathology, fibrosis score,
and hydroxyproline content all showed similar patterns
in this intradermal BLM–induced skin and lung fibrosis
model.
Consistent with the results of the tissue fibrosis
examination, ␣-SMA⫹ cell numbers were significantly
reduced in both the lung and skin of ICOS–/– mice, but
were increased in ICOSL–/– mice, as compared with
wild-type mice (Figures 2A and B). Thus, ICOS loss
ameliorates the lung and skin fibrosis induced by daily
intradermal injections of BLM, whereas ICOSL loss
aggravates it.
Lung and skin inflammation induced by daily
intradermal injections of BLM. The numbers of F4/80⫹
macrophages, CD3⫹ T cells, B220⫹ B cells, and MPO⫹
neutrophils were assessed in lung and skin tissues of
1728
mice subjected to daily intradermal injections of BLM.
In comparison with wild-type mice, lung tissues from
ICOS–/– mice exhibited reduced numbers of macrophages, T cells, and neutrophils throughout the 28-day
treatment period (Figure 3A). Conversely, ICOSL–/–
mice demonstrated increased numbers of macrophages,
T cells, and neutrophils. Similarly, the numbers of
macrophages and T cells were reduced in the skin of
ICOS–/– mice (Figure 3B), while the number of macrophages, T cells, and neutrophils were increased in the
skin of ICOSL–/– mice. Thus, inflammatory cell infiltration was modest in ICOS–/– mice and prominent in
ICOSL–/– mice.
Cytokine expression in the lungs of mutant mice.
To assess the effects of ICOS and ICOSL loss on the
production of various cytokines and chemokines, their
concentrations in lung lysates were measured by cytometric bead array analysis on day 5 following intratracheal BLM treatment. In addition, levels of expression
of mRNA for transforming growth factor ␤ (TGF␤),
connective tissue growth factor (CTGF), and FoxP3 in
the lungs were assessed by real-time RT-PCR. The
concentrations or levels of expression of TNF␣, IFN␥,
IL-4, IL-6, IL-10, IL-12, MCP-1, MIP-1␣, TGF␤, CTGF,
and FoxP3 in the lungs of BLM-treated wild-type mice
were significantly higher than those in the lungs of
saline-treated wild-type mice (data not shown). Total
protein concentrations in lung lysates were comparable
between the BLM-treated strains (mean ⫾ SEM 38.7 ⫾
3.3 mg/ml in wild-type, 38.0 ⫾ 3.5 mg/ml in ICOS–/–, and
37.1 ⫾ 3.0 mg/ml in ICOSL–/– mice). The concentration
of TNF␣, IFN␥, IL-4, IL-6, IL-10, IL-12, MCP-1, and
MIP-1␣ in the lungs of BLM-treated mice was not
significantly different between groups (Figure 4A). Expression levels of CTGF and FoxP3 were also similar
among the mouse strains (Figure 4B). However, the
expression levels of TGF␤ were significantly lower in
ICOS–/– mice and higher in ICOSL–/– mice relative to
the levels in wild-type mice (Figure 4B). Thus, TGF␤
expression was most associated with the severity of lung
fibrosis.
Phenotype similarity of mice deficient in both
ICOS and ICOSL as compared with mice deficient in
ICOSL following treatment with BLM. We hypothesized
that either ICOS or ICOSL had functions that were
independent of the ICOS/ICOSL axis in the BLMinduced fibrosis model. To explore this possibility, we
generated mice deficient in both ICOS and ICOSL
(ICOS–/–ICOSL–/–) and assessed the severity of BLMinduced lung and skin fibrosis.
TANAKA ET AL
Figure 4. Levels of cytokines, chemokines, growth factors, and FoxP3
in the lungs of mice treated with intratracheal injections of bleomycin
(BLM). ICOS–/–, ICOSL–/–, and wild-type (WT) mice were treated
with a single intratracheal injection of BLM. A, Concentrations of
tumor necrosis factor ␣ (TNF␣), interferon-␥ (IFN␥), interleukin-4
(IL-4), IL-6, IL-10, IL-12, monocyte chemotactic protein 1 (MCP-1),
and macrophage inflammatory protein 1␣ (MIP-1␣) were measured by
cytometric bead array analysis or enzyme-linked immunosorbent assay
in supernatants of lung homogenates. B, Expression of transforming
growth factor ␤ (TGF␤), connective tissue growth factor (CTGF), and
FoxP3 mRNA was measured by real-time quantitative reverse
transcription–polymerase chain reaction analysis. Values are the mean
and SEM of 5 mice per group. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01.
ICOSL REGULATION OF BLEOMYCIN-INDUCED MURINE LUNG AND SKIN FIBROSIS
1729
Figure 5. Skin and lung fibrosis in ICOS–/–ICOSL–/– double-knockout mice as compared with wild-type (WT) mice following treatment with
bleomycin (BLM). A, Time course of survival, representative photomicrographs of the lungs, lung fibrosis scores, and percentage of total tissue area
in mice treated with intratracheal BLM. H&E ⫽ hematoxylin and eosin. B and C, Inflammation and fibrosis of the lungs (B) and skin (C) following
intradermal injection of BLM, assessed as described in Figures 2 and 3. ␣-SMA ⫽ ␣-smooth muscle actin; hpf ⫽ high-power field; MPO ⫽
myeloperoxidase. D, Concentrations of tumor necrosis factor ␣ (TNF␣), interferon-␥ (IFN␥), interleukin-4 (IL-4), IL-6, IL-10, IL-12, monocyte
chemotactic protein 1 (MCP-1), and macrophage inflammatory protein 1␣ (MIP-1␣) in lung lysates, assessed as described in Figure 4. E, Levels of
mRNA for transforming growth factor ␤ (TGF␤), connective tissue growth factor (CTGF), and FoxP3 in the lungs, assessed as described in Figure
4. ⴱ ⫽ P ⬍ 0.05.
First, we assessed the sensitivity of ICOS–/–
ICOSL–/– mice to intratracheal treatment with BLM.
Deficiency of both ICOS and ICOSL resulted in high
mortality rates, severe fibrosis, and elevated fibrosis
scores (Figure 5A). These findings were quite comparable to what we had observed previously in ICOSL–/– mice.
Next, we performed daily intradermal injections
of BLM. In this model of BLM-induced lung and skin
fibrosis, the ICOS–/–ICOSL–/– mice exhibited significantly increased fibrosis scores, ␣-SMA⫹ cell infiltration, and hydroxyproline content, reflecting severe lung
fibrosis, as compared with wild-type mice (Figure 5B).
ICOS–/–ICOSL–/– mice exhibited significantly greater
lung infiltration by macrophages, B cells, T cells, and
neutrophils on day 7 as compared with wild-type mice
(Figure 5B). Furthermore, deficiency of both ICOS and
1730
TANAKA ET AL
Figure 6. Expression of inducible costimulator (ICOS) and ICOS ligand (ICOSL) in mutant mice. A, ICOS expression on CD4⫹ T cells and ICOSL
expression on CD14⫹ macrophages and on B220⫹ B cells in the spleens of untreated wild-type (WT), ICOS–/–, and ICOSL–/– mice, as determined
by flow cytometry. Histograms are representative of at least 5 experiments. Values at the left are the mean and SEM of 5 mice per group. ⴱⴱ ⫽ P ⬍
0.01. B, ICOS and ICOSL expression on T cells, macrophages, and B cells in bronchoalveolar lavage fluid prepared in wild-type, ICOS–/–, and
ICOSL–/– mice after intratracheal treatment with bleomycin, as determined by flow cytometry. Histograms are representative of at least 5
experiments conducted on day 7. Values at the left are the mean and SEM of 5 mice per group, as determined on the indicated days. ⴱⴱ ⫽ P ⬍ 0.01.
ICOSL led to significant increases in dermal thickness,
hydroxyproline content, and ␣-SMA⫹ cells in the skin
relative to the findings in wild-type mice (Figure 5C).
Numbers of infiltrating macrophages, B cells, T cells,
and neutrophils were increased in the skin of ICOS–/–
ICOSL–/– mice (Figure 5C).
The total protein concentration in lung lysates
was comparable in ICOS–/–ICOSL–/– mice and in wildtype mice (mean ⫾ SEM 38.7 ⫾ 3.3 mg/ml in wild-type
mice and 38.1 ⫾ 2.9 mg/ml in ICOS–/–ICOSL–/– mice).
The concentrations of cytokines and chemokines in
lysates of ICOS–/–ICOSL–/– mice were not significantly
different from those in lysates of wild-type mice (Figure
5D). Levels of mRNA for TGF␤, but not CTGF or
FoxP3, were significantly higher in the lung tissues of
ICOS–/–ICOSL–/– mice as compared with wild-type mice
(Figure 5E). Thus, the results of BLM-induced fibrosis
in ICOS–/–ICOSL–/– mice resembled those in ICOSL–/–
mice.
Inverse association between ICOSL expression
levels and the severity of lung fibrosis. In light of the
results obtained in ICOS–/–ICOSL–/– mice, we considered that expression of ICOSL, rather than ICOS,
played a dominant role in BLM-induced fibrosis. We
assessed the expression levels of ICOS and ICOSL by
fluorescence-activated cell sorting of splenocytes from
mice that had not been treated with BLM (Figure 6A).
The expression levels of ICOS on CD4⫹ T cells were
significantly elevated by ⬃135% in ICOSL–/– mice than
in wild-type mice. In addition, the expression levels of
ICOSL on macrophages and B cells were increased
⬃302% and ⬃348%, respectively, in ICOS–/– mice as
compared with wild-type mice.
Next, we evaluated the association of ICOS and
ICOSL expression levels in BAL fluid cells on days 7, 14,
and 21 after a single intratracheal treatment with BLM
(Figure 6B). The expression levels of ICOS on CD4⫹ T
cells and the expression of ICOSL on macrophages and
B cells reached a maximum ⬃7 days after the single
intratracheal BLM treatment, and then gradually declined. On day 7, the expression level of ICOS on T cells
was elevated ⬃233% in ICOSL–/– mice as compared
with wild-type mice. In contrast, the expression levels of
ICOSL on macrophages and B cells were increased
⬃161% and ⬃290%, respectively, in ICOS–/– mice relative to wild-type mice. Thus, ICOSL expression levels
were inversely associated with disease severity in this
mouse model.
ICOSL REGULATION OF BLEOMYCIN-INDUCED MURINE LUNG AND SKIN FIBROSIS
DISCUSSION
In the present study, we demonstrated that
ICOSL plays roles that are independent of the ICOS/
ICOSL pathway in a mouse model of BLM-induced lung
and skin fibrosis. In addition, the severity of tissue
fibrosis appeared to correlate with the expression levels
of ICOSL on B cells and macrophages.
We made the unexpected finding that ICOS–/–
mice and ICOSL–/– mice exhibited contrasting phenotypes in the BLM induced lung and skin fibrosis model.
Before BLM treatment, no histologic difference was
found in the lungs and skin of these mutant mouse
strains as compared with wild-type mice (data not
shown). However, ICOS deficiency significantly inhibited, whereas ICOSL deficiency exacerbated, BLMinduced lung and skin fibrosis (Figures 1 and 2).
It is currently thought that ICOS and ICOSL
represent a single receptor/ligand pair with no other
known binding partners (19). Consistent with this idea,
previous studies have shown that ICOS–/– mice and
ICOSL–/– mice have similar phenotypes (18,19). Perhaps
because of these previous reports, no simultaneous
investigation of the roles of both ICOS and ICOSL in
murine disease models has been performed. Furthermore, most studies have investigated the roles of ICOS
or ICOSL in disease models that were related to
antigen-specific adaptive immunity (20–23). While
ICOS is expressed on effector and memory T cells,
ICOSL is constitutively expressed on macrophages and
dendritic cells, which contribute to innate immunity, in
addition to its expression on B cells. Several studies have
shown that ICOSL is expressed on endothelial cells and
some epidermal cells, although we did not detect expression on these cell types, at least by immunohistochemical
staining (results not shown). Our findings therefore
suggest that ICOSL contributes to innate immunity in
BLM-induced tissue fibrosis independently of the ICOS/
ICOSL pathway. Since ICOS–/–ICOSL–/– mice displayed
similar phenotypes as ICOSL–/– mice after BLM treatment (Figure 5), this suggests that the ICOS/ICOSLindependent functions of ICOSL dominate over its
ICOS-dependent functions in this model system.
Our study also revealed that the expression levels
of ICOS and ICOSL were related to each other. That is,
ICOS expression was markedly up-regulated by ICOSL
loss, and ICOSL expression was dramatically increased
by the lack of ICOS before and during BLM treatment
(Figure 6). These results reflect the previously reported
increase in ICOSL expression on B cells from ICOSdeficient patients (29). Although these findings may
1731
arise from feedback regulation resulting from interactions between ICOS and ICOSL, the underlying mechanism remains unclear. A recent study demonstrated
that ICOSL expression on B cells was reduced in
ICOS-transgenic mice, resulting in a phenotype resembling that of ICOS–/– mice (30). These findings suggested
that ICOSL expression on B cells is tightly regulated by
ICOS expression on T cells, which interacted with these
B cells. In our current study, ICOSL expression on
macrophages and B cells from BAL fluid was markedly
elevated during BLM treatment in ICOS-deficient mice
as compared with wild-type mice (Figure 6). Since
ICOSL loss exacerbated tissue fibrosis in spite of the
presence of ICOS, the levels of ICOSL expression on
macrophages or B cells were found to be inversely
associated with the severity of fibrosis in this mouse
model. Therefore, ICOSL expression on antigenpresenting cells may play regulatory roles that are
independent of the ICOS/ICOSL pathway during development of BLM-induced tissue fibrosis.
As described above, the mechanism by which
ICOSL expression levels regulate tissue inflammation
and fibrosis remains unclear. However, the severity of
lung involvement was found to be correlated with lung
TGF␤ expression levels in each strain we studied (Figures 4B and 5E). That is, TGF␤ was significantly reduced in ICOS–/– mice and significantly increased in
ICOSL–/– or ICOS–/–ICOSL–/– mice. Since TGF␤ has
been considered the central player in tissue fibrosis, we
proposed that ICOSL regulation of TGF␤ might be
affecting the development of lung fibrosis. Our findings
suggest that independently of its interactions with ICOS,
ICOSL plays a dominant role in regulating TGF␤ induction and the development of fibrosis in BLM-induced
murine fibrosis models.
A limitation of this study is that there are alternative explanations for the findings, among which is the
possibility that the procedure we followed to obtain
ICOSL–/– mice could have generated additional unwanted gene-encoded mutations that represent a more
relevant explanation of the findings. In addition, the
findings in the BLM-induced mouse model of fibrosis
are not simply translatable to the pathogenesis and the
therapeutic approach to SSc in humans. A recent article
proposed some criteria to use in selecting the most
promising molecular targets for trials in SSc (31). One of
those criteria is that the antifibrotic effects should be
confirmed in at least 2 complementary animal models of
SSc. Further studies that include assessments in additional animal models will be needed to clarify the roles
of ICOS and ICOSL in the development of SSc.
1732
TANAKA ET AL
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors approved
the final version to be published. Dr. Hasegawa 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 conception and design. Fujimoto, Hasegawa.
Acquisition of data. Tanaka.
Analysis and interpretation of data. Tanaka, Fujimoto, Hamaguchi,
Sato, Takehara, Hasegawa.
REFERENCES
1. Varga J, Abraham D. Systemic sclerosis: a prototypic multisystem
fibrotic disorder. J Clin Invest 2007;117:557–67.
2. Gabrielli A, Avvedimento EV, Krieg T. Scleroderma. N Engl
J Med 2009;360:1989–2003.
3. Schrier DJ, Phan SH, McGarry BM. The effects of the nude
(nu/nu) mutation on bleomycin-induced pulmonary fibrosis: a
biochemical evaluation. Am Rev Respir Dis 1983;127:614–7.
4. Ponticos M, Holmes AM, Shi-wen X, Leoni P, Khan K, Rajkumar
VS, et al. Pivotal role of connective tissue growth factor in lung
fibrosis: MAPK-dependent transcriptional activation of type I
collagen. Arthritis Rheum 2009;60:2142–55.
5. Yamamoto T, Takagawa S, Katayama I, Yamazaki K, Hamazaki
Y, Shinkai H, et al. Animal model of sclerotic skin. I. Local
injections of bleomycin induce sclerotic skin mimicking scleroderma. J Invest Dermatol 1999;112:456–62.
6. Yoshizaki A, Iwata Y, Komura K, Ogawa F, Hara T, Muroi E, et
al. CD19 regulates skin and lung fibrosis via Toll-like receptor
signaling in a model of bleomycin-induced scleroderma. Am J
Pathol 2008;172:1650–63.
7. Hutloff A, Dittrich AM, Beier KC, Eljaschewitsch B, Kraft R,
Anagnostopoulos I, et al. ICOS is an inducible T-cell co-stimulator
structurally and functionally related to CD28. Nature 1999;397:
263–6.
8. Yoshinaga SK, Whoriskey JS, Khare SD, Sarmiento U, Guo J,
Horan T, et al. T-cell co-stimulation through B7RP-1 and ICOS.
Nature 1999;402:827–32.
9. Greenwald RJ, Freeman GJ, Sharpe AH. The B7 family revisited.
Annu Rev Immunol 2005;23:515–48.
10. Swallow MM, Wallin JJ, Sha WC. B7h, a novel costimulatory
homolog of B7.1 and B7.2, is induced by TNF␣. Immunity
1999;11:423–32.
11. McAdam AJ, Chang TT, Lumelsky AE, Greenfield EA, Boussiotis
VA, Duke-Cohan JS, et al. Mouse inducible costimulatory molecule (ICOS) expression is enhanced by CD28 costimulation and
regulates differentiation of CD4⫹ T cells. J Immunol 2000;165:
5035–40.
12. Nurieva RI, Chung Y, Hwang D, Yang XO, Kang HS, Ma L, et al.
Generation of T follicular helper cells is mediated by interleukin-21 but independent of T helper 1, 2, or 17 cell lineages.
Immunity 2008;29:138–49.
13. Coyle AJ, Gutierrez-Ramos JC. The role of ICOS and other
costimulatory molecules in allergy and asthma. Springer Semin
Immunopathol 2004;25:349–59.
14. Lohning M, Hutloff A, Kallinich T, Mages HW, Bonhagen K,
Radbruch A, et al. Expression of ICOS in vivo defines CD4⫹
effector T cells with high inflammatory potential and a strong bias
for secretion of interleukin 10. J Exp Med 2003;197:181–93.
15. Bauquet AT, Jin H, Paterson AM, Mitsdoerffer M, Ho IC, Sharpe
AH, et al. The costimulatory molecule ICOS regulates the expression of c-Maf and IL-21 in the development of follicular T helper
cells and TH-17 cells. Nat Immunol 2009;10:167–75.
16. Burmeister Y, Lischke T, Dahler AC, Mages HW, Lam KP, Coyle
AJ, et al. ICOS controls the pool size of effector-memory and
regulatory T cells. J Immunol 2008;180:774–82.
17. Akbari O, Freeman GJ, Meyer EH, Greenfield EA, Chang TT,
Sharpe AH, et al. Antigen-specific regulatory T cells develop via
the ICOS–ICOS-ligand pathway and inhibit allergen-induced airway hyperreactivity. Nat Med 2002;8:1024–32.
18. Mak TW, Shahinian A, Yoshinaga SK, Wakeham A, Boucher LM,
Pintilie M, et al. Costimulation through the inducible costimulator
ligand is essential for both T helper and B cell functions in T
cell-dependent B cell responses. Nat Immunol 2003;4:765–72.
19. Nurieva RI, Mai XM, Forbush K, Bevan MJ, Dong C. B7h is
required for T cell activation, differentiation, and effector function. Proc Natl Acad Sci U S A 2003;100:14163–8.
20. Dong C, Juedes AE, Temann UA, Shresta S, Allison JP, Ruddle
NH, et al. ICOS co-stimulatory receptor is essential for T-cell
activation and function. Nature 2001;409:97–101.
21. Rottman JB, Smith T, Tonra JR, Ganley K, Bloom T, Silva R, et
al. The costimulatory molecule ICOS plays an important role in
the immunopathogenesis of EAE. Nat Immunol 2001;2:605–11.
22. Kopf M, Coyle AJ, Schmitz N, Barner M, Oxenius A, Gallimore A,
et al. Inducible costimulator protein (ICOS) controls T helper cell
subset polarization after virus and parasite infection. J Exp Med
2000;192:53–61.
23. Ozkaynak E, Gao W, Shemmeri N, Wang C, Gutierrez-Ramos JC,
Amaral J, et al. Importance of ICOS–B7RP-1 costimulation in
acute and chronic allograft rejection. Nat Immunol 2001;2:591–6.
24. Ashcroft T, Simpson JM, Timbrell V. Simple method of estimating
severity of pulmonary fibrosis on a numerical scale. J Clin Pathol
1988;41:467–70.
25. Lopez-De Leon A, Rojkind M. A simple micromethod for collagen and total protein determination in formalin-fixed paraffinembedded sections. J Histochem Cytochem 1985;33:737–43.
26. Mori R, Kondo T, Nishie T, Ohshima T, Asano M. Impairment of
skin wound healing in ␤-1,4-galactosyltransferase-deficient mice
with reduced leukocyte recruitment. Am J Pathol 2004;164:
1303–14.
27. Meijerink J, Mandigers C, van de Locht L, Tonnissen E, Goodsaid
F, Raemaekers J. A novel method to compensate for different
amplification efficiencies between patient DNA samples in quantitative real-time PCR. J Mol Diagn 2001;3:55–61.
28. Moore BB, Ballinger MN, White ES, Green ME, Herrygers AB,
Wilke CA, et al. Bleomycin-induced E prostanoid receptor
changes alter fibroblast responses to prostaglandin E2. J Immunol
2005;174:5644–9.
29. Grimbacher B, Hutloff A, Schlesier M, Glocker E, Warnatz K,
Drager R, et al. Homozygous loss of ICOS is associated with
adult-onset common variable immunodeficiency. Nat Immunol
2003;4:261–8.
30. Watanabe M, Takagi Y, Kotani M, Hara Y, Inamine A, Hayashi
K, et al. Down-regulation of ICOS ligand by interaction with ICOS
functions as a regulatory mechanism for immune responses.
J Immunol 2008;180:5222–34.
31. Distler JH, Distler O. Criteria to select molecular targets for
anti-fibrotic therapy. Rheumatology (Oxford) 2008;47 Suppl
5:v12–3.