ARTHRITIS & RHEUMATISM
Vol. 50, No. 8, August 2004, pp 2685–2694
© 2004, American College of Rheumatology
Lack of Requirement of Osteopontin for Inflammation,
Bone Erosion, and Cartilage Damage in the
K/BxN Model of Autoantibody-Mediated Arthritis
Jonathan P. Jacobs,1Allison R. Pettit,2Mari L. Shinohara,3Marianne Jansson,3
Harvey Cantor,3Ellen M. Gravallese,2Diane Mathis,1and Christophe Benoist1
Objective. Osteopontin (OPN) is a secreted glyco-
protein involved in a range of physiologic processes,
including inflammation, immunity mediated by Th1
cells, and bone remodeling. It is expressed in the joints
of rheumatoid arthritis patients and has been the
subject of conflicting reports concerning its role in
arthritis induced by antibodies against type II collagen.
This study assessed the role of OPN in the K/BxN
serum-transfer model of autoantibody-induced arthri-
Methods. Expression of OPN gene transcripts
was assessed by microarray analysis of ankle RNA taken
at 6 time points after transfer of K/BxN serum. OPN-
sufficient or OPN-deficient littermates backcrossed for
10 generations onto the C57BL/6 genetic background
were given K/BxN serum. Arthritis severity was mea-
sured by ankle thickening and a clinical index. Hind
limb sections were stained with hematoxylin and eosin
or toluidine blue and scored for inflammation, cartilage
damage, and bone erosion.
Results. OPN messenger RNA transcripts pro-
gressively increased in ankle joints during the course of
K/BxN serum-transferred arthritis. OPN-deficient mice
receiving K/BxN serum developed arthritis with kinetics
and clinical severity comparable with those of OPN-
sufficient littermates. Histologic assessment of arthritic
joints from OPN-deficient mice revealed synovial hyper-
plasia, pannus formation, mononuclear cell infiltration,
bone erosion, cartilage damage at sites adjacent to and
distal from pannus invasion, and tartrate-resistant acid
phosphatase–positive multinucleated cells at sites of
bone erosion. Histopathologic scoring demonstrated
comparable levels of inflammation, cartilage damage,
and bone erosion in OPN-sufficient and OPN-deficient
Conclusion. OPN does not have a required role in
inflammation, bone erosion, and cartilage damage in
the K/BxN serum-transfer model.
Rheumatoid arthritis (RA), a chronic inflamma-
tory disease involving synovial joints, is characterized by
synovial hyperplasia, leukocyte infiltration into syno-
vium, focal bone erosion, and formation of a region of
invasive synovial tissue, or pannus, that destroys joint
structures. One candidate regulator of inflammation and
bone destruction in RA is osteopontin (OPN), which is
also known as early T lymphocyte activation gene 1
OPN is a phosphorylated glycoprotein that is
present as a cytokine in body fluids and as an immobi-
lized component of the extracellular matrix of mineral-
The Joslin Diabetes and Endocrinology Research Center
cores were supported by the NIH (National Institute of Diabetes and
Digestive and Kidney Diseases). Mr. Jacobs’ work was supported by
the NIH (National Institute of Arthritis and Musculoskeletal and Skin
Diseases [NIAMS] grant AI-46580-04 and National Institute of Allergy
and Infectious Diseases [NIAID] grant AI-54904-01) and by a fellow-
ship from the Lupus Foundation of Massachusetts. Dr. Pettit’s work
was supported by an Arthritis Foundation Fellowship award. Dr.
Shinohara’s work was supported by a National Research Service award
(CA 70083). Drs. Jansson and Cantor’s work (grant AI-12184), Dr.
Gravallese’s work (grant R01-AR-47665), and Drs. Mathis and Be-
noist’s work (NIAMS grant AI-46580-04 and NIAID grant AI-54904-
01) were supported by the NIH.
1Jonathan P. Jacobs, BA, Diane Mathis, PhD, Christophe
Benoist, MD, PhD: Joslin Diabetes Center, Brigham and Women’s
Hospital, and Harvard Medical School, Boston, Massachusetts;2Alli-
son R. Pettit, PhD, Ellen M. Gravallese, MD: Harvard Institutes of
Medicine, Beth Israel Deaconess Medical Center, and Harvard Med-
ical School, Boston, Massachusetts;3Mari L. Shinohara, PhD, Mari-
anne Jansson, PhD, Harvey Cantor, MD: Dana Farber Cancer Insti-
tute, and Harvard Medical School, Boston, Massachusetts.
Address correspondence and reprint requests to Christophe
Benoist, MD, PhD, or Diane Mathis, PhD, Joslin Diabetes Center, 1
Joslin Place, Room 475, Boston, MA 02215. E-mail: cbdm@joslin.
Submitted for publication December 8, 2003; accepted in
revised form April 1, 2004.
ized tissues (1–3). It is secreted by many cell types,
including T cells, macrophages, osteoblasts, and oste-
oclasts, and it can be induced by inflammatory mediators
such as interleukin-1, tumor necrosis factor ?, and
platelet-derived growth factor. OPN interacts with inte-
grin receptors and CD44 to induce chemotaxis, promote
cell adhesion, and modulate cell function. At sites of
inflammation, it promotes macrophage infiltration and
dendritic cell migration to lymph nodes (4,5). It is an
important mediator of Th1 immunity and protective
granuloma responses, and can enhance B cell prolifera-
tion and antibody secretion (6–9). OPN has been di-
rectly linked to autoimmunity in studies finding that
OPN-deficient mice have attenuated experimental auto-
immune encephalomyelitis (EAE) (10,11). However,
these results have recently been challenged by a group of
investigators who observed no attenuation of EAE in
more genetically homogeneous OPN-deficient mice
OPN may also play a role in bone resorption.
OPN-deficient mice have an increased volume of trabec-
ular bone and reduced bone resorption induced by
experimental stresses, including ovariectomy and me-
chanical stress (13–17). Ectopically transplanted bone
from OPN-deficient mice is resorbed less effectively
than is transplanted bone from wild-type mice (18).
Cultured bone from OPN-deficient mice has less resorp-
tion and osteoclast formation than does cultured bone
from wild-type mice in response to parathyroid hormone
and RANKL/macrophage colony-stimulating factor
A few studies have provided experimental evi-
dence linking OPN to inflammatory arthritis (20). Levels
of OPN messenger RNA (mRNA) and protein were
found to be elevated in synovial tissue from RA patients
compared with synovial tissue from osteoarthritis pa-
tients (21,22). In those studies, OPN primarily localized
to fibroblasts in the synovial lining and regions of pannus
invasion into cartilage. OPN was also detected at sites of
osteoclast-mediated bone erosion in mice with collagen-
induced arthritis (CIA) and in the cartilage of mice with
anti–type II collagen (anti-CII) antibody–induced arthri-
tis (23,24). In a study using the latter system, OPN-
deficient mice had reduced levels of inflammation, car-
tilage damage, and inflammatory cell infiltration
compared with OPN-sufficient controls (24). Using the
same model, a second group of investigators found that
treatment with antibodies against the SLAYGLR
epitope of thrombin-cleaved OPN delayed the onset of
arthritis and reduced its severity (25). The role of OPN
in arthritis remains controversial, however, since an-
other group of investigators found that OPN deficiency
did not protect against CIA or anti-CII antibody–
induced arthritis (12).
We studied the role of OPN in the K/BxN
serum-transfer model of arthritis. In this system, arthritis
bearing marked clinical and histologic similarity to RA
was induced by intraperitoneal injection of serum from
arthritic K/BxN mice, the progeny of KRN T cell
receptor (TCR)–transgenic mice (K/B) and nonobese
diabetic mice (N) (26,27). Disease is mediated by patho-
genic autoantibodies produced against glucose-6-
phosphate isomerase, a ubiquitously expressed glycolytic
enzyme that deposits on joint surfaces (28,29). We found
that although OPN was up-regulated during K/BxN
serum-transferred arthritis, it had no required role in
MATERIALS AND METHODS
Mice. OPN-knockout mice were described by Rittling
et al (13) and have been backcrossed onto the C57BL/6 (B6)
genetic background for 10 generations. Matched littermates
from heterozygote crosses or heterozygote/homozygote
crosses were used as controls. Genotypes were assessed by
genomic polymerase chain reaction and confirmed by OPN
enzyme-linked immunosorbent assay (ELISA) of sera from
killed animals. K/BxN mice were generated by crossing KRN
TCR-transgenic B6 mice with NOD mice (26). These experi-
ments were reviewed by the Harvard Medical School Institu-
tional Animal Care and Use Committee (protocol no. 03024).
Generation of serum-transferred arthritis and clinical
scoring. K/BxN serum was collected from 8-week-old arthritic
K/BxN mice and pooled for each experiment. Arthritis was
induced by intraperitoneal injection of 150 ?l of K/BxN serum
on days 0 and 2 or on days 0, 2, 7, and 14. In a dose-response
experiment, mice were injected with 18.75, 37.5, 75, or 150 ?l
of K/BxN serum on days 0 and 2. For doses lower than 150 ?l,
serum was diluted to 150 ?l with phosphate buffered saline
(PBS). Animals were killed for histologic assessment on day 15
or on days 7, 14, and 21; all remaining animals were killed 21
days after serum transfer.
Two clinical parameters of arthritis, ankle thickness
and a clinical index, were measured from the day of injection
until the mice were killed. Ankle thickness was measured with
calipers (J15; Blet, Lyon, France) and reported as ankle
thickening, representing the difference between the ankle
thickness on day 0 and at specific time points thereafter. Each
limb was scored on a scale of 0 (no observable swelling) to 3
(severe inflammation). The scores for the 4 limbs were added
to give the clinical index.
OPN expression analysis. A microarray analysis was
performed on ankle RNA extracts taken on days 0, 1, 3, 7, 12,
and 18 after serum transfer into B6 mice (Jacobs JP: unpub-
lished observations). Briefly, ankle joints were dissected open,
and the exposed synovial tissue was digested in 6M urea/2%
sodium dodecyl sulfate. RNA samples prepared from these
extracts were amplified using a MessageAmp kit (Ambion,
2686JACOBS ET AL
Austin, TX) and labeled with biotinylated nucleotides using
the BioArray kit (Enzo Diagnostics, Farmingdale, NY). For
each sample, 2–3 mice were pooled, and for each time point,
3–6 samples were collected from separate experiments and
prepared. The resulting labeled RNA was hybridized to
U74Av2 microarrays (Affymetrix, Santa Clara, CA), which
were read with an argon laser detector (GeneArray Scanner;
Affymetrix). Raw hybridization intensities were processed into
expression values using the robust multichip analysis method.
Sera were collected from 3 mice on days 0, 1, 3, 7, 12,
and 18 during the above experiments. Synovial fluid samples
were taken on day 7 by aspirating fluid from dissected ankles.
Synovial fluid protein was obtained by centrifuging the samples
for 5 minutes at 2,400g and collecting the supernatant. Sera
and synovial fluid protein were incubated in triplicate for 2
hours at room temperature in enzyme immunoassay/
radioimmunoassay plate wells (Corning, Corning, NY) that
had been coated with 0.4 ?g/ml of purified AF808 anti-
osteopontin antibody (R&D Systems, Minneapolis, MN). Bi-
otinylated BAF808 anti-osteopontin antibody (0.25 ?g/ml;
R&D Systems) was added for 1 hour at room temperature,
then avidin–horseradish peroxidase secondary antibody
(PharMingen, San Diego, CA) was added for an additional 1
hour at room temperature. Each of the antibody binding steps
was followed by 5 washes with 0.05% Tween 20–PBS.
Substrate solution (TMB Substrate Reagent set;
PharMingen) was added for 15–20 minutes at room tempera-
ture. The plates were then read on an ELISA plate reader
(Molecular Devices, Sunnyvale, CA) at 450 nm. Recombinant
mouse osteopontin (441-OP; R&D Systems) was used as a
Figure 1. Up-regulation of osteopontin (OPN) mRNA in the ankles
of K/BxN serum-transferred mice. A microarray analysis was per-
formed on ankle RNA samples collected on days 0, 1, 3, 7, 12, and 18
after initial injection with K/BxN serum. Each sample consists of
pooled ankle RNA extracts from 2–3 mice; 3–6 replicates were
performed for each time point. A, Time course of arthritis, as
represented by the mean ? SEM ankle thickening each day in the 9
mice that were killed on day 18. B, Fold change in OPN mRNA
expression from day 0 to day 18, as measured by Affymetrix U74Av2
microarrays. Values are the mean ? SEM.
Figure 2. Comparable severity of arthritis in osteopontin (OPN)–
deficient and OPN-sufficient mice upon injection with K/BxN serum.
OPN-sufficient (heterozygous [HZ]) (n ? 6) and OPN-deficient (KO)
mice (n ? 6) were injected with 150 ?l of K/BxN serum on days 0 and
2 in 2 separate experiments. The ankle thickening (A) and clinical
index (B) were measured from day 0 to day 15 (n ? 2) or day 21 (n ?
4). Values are the mean ? SEM. P ? 0.05 at every time point.
ROLE OF OPN IN AUTOANTIBODY-INDUCED ARTHRITIS IN K/BxN MICE2687
Histologic assessment and histopathologic scoring.
Hind limbs were prepared for histology by dissecting off the
skin and outer muscle and then separating the knee and ankle
joints at the mid-tibia. Specimens were fixed in 4% parafor-
maldehyde for a minimum of 12 hours and demineralized for
?2 weeks in 14% EDTA. Specimens were subsequently em-
bedded in paraffin (Citadel 1000; Shandon, Pittsburgh, PA).
For each specimen, at least 15 serial 5-?m sagittal sections
were cut, and every fifth section was stained with hematoxylin
and eosin (H&E; Sigma, St. Louis, MO) for evaluation of
inflammation, bone erosion, and cartilage destruction. An
adjacent section was stained with toluidine blue (Sigma) to
assess cartilage integrity.
Staining for tartrate-resistant acid phosphatase
(TRAP) was performed by a modification of a previously
described method (30). Briefly, sections were incubated for 15
minutes at 37°C in freshly prepared 0.1 moles/liter Tris buffer,
pH 5.0, 1.35 mmoles/liter naphthol-AS-MX phosphate
(Sigma), 0.362 moles/liter N,N-dimethylformamide, 3.88
mmoles/liter violet LB salt (Sigma), and 25 mmoles/liter
sodium tartrate. Slides were rinsed for 10 minutes and coun-
terstained with hematoxylin.
A total of 4–8 H&E-stained sections were scored by a
blinded observer (ARP) at low power for inflammation and at
low and high power for bone erosion. Inflammation was
assessed in synovium and periarticular soft tissues of the ankle
and forefoot. Each section was scored on a scale of 0–5 for
inflammation (0 ? normal, 1 ? mild edema and/or minimal
cellular infiltrate, 2 ? mild cellular infiltrate, 3 ? moderate
cellular infiltrate, 4 ? marked cellular infiltrate, and 5 ?
severe cellular infiltrate). Edema was included only in the
minimal inflammation score since it is an early feature of this
arthritis model (26). Bone erosion was assessed separately in
the tibiotalar and forefoot regions and was scored on a 0–5
scale as previously described (31), with the minor modifica-
tions described previously (32). Cartilage was evaluated for
Figure 3. Comparable bone destruction and cartilage damage in the tibiotalar and forefoot regions of osteopontin (OPN)–sufficient and
OPN-deficient mice. OPN-sufficient (wild-type [WT] or heterozygous [HZ]) and OPN-deficient (KO) mice (n ? 3 per group) were injected with
150 ?l of K/BxN serum on days 0, 2, 7, and 14. One mouse from each group was killed on days 7, 14, and 21 after initial K/BxN serum injection.
The clinical index (A) and ankle thickening (B) were measured throughout the 3-week observation period. Histopathologic scoring of bone erosion
(C) and cartilage damage (D) was performed on hind limbs from OPN-sufficient and OPN-deficient mice killed on days 7, 14, and 21. The limbs
were compared in pairs (p) matched for left/right position. The tibiotalar (TT) and forefoot (FF) regions were scored separately.
2688JACOBS ET AL
surface damage and cartilage destruction secondary to pannus
invasion as previously described (32). Cartilage damage was
scored on serial toluidine blue–stained sections based upon
proteoglycan loss in areas remote from inflamed synovium.
Cartilage destruction in areas adjacent to pannus was scored
separately on H&E-stained sections. The mean score for each
histopathologic feature was calculated.
Local up-regulation of OPN mRNA during
K/BxN serum-transferred arthritis. Given the interest
in the role of OPN in inflammatory arthritis, we used
microarray analysis of ankle synovial tissue taken at 6
time points to assess OPN mRNA expression during the
unfolding of K/BxN serum-transferred arthritis (Jacobs
JP: unpublished observations). OPN mRNA was pro-
gressively up-regulated in ankle joints as arthritis devel-
oped (Figure 1). Levels of OPN mRNA began to rise at
about the time of clinical disease onset (day 3), reached
an intermediate value as disease began to peak (day 7),
and rose more slowly thereafter as the disease settled in
(days 12 and 18). The concentration of OPN protein in
synovial fluid on day 7 was 1.01 ? 0.26 ?g/ml (mean ?
SEM) as determined by ELISA. This level was 3.6-fold
greater than the OPN concentration in serum taken
Figure 4. Comparable pannus formation, cartilage damage, and bone erosion in the arthritic joints of osteopontin (OPN)–deficient and wild-type
mice. A and B, Representative sections of the tibiotalar region of matched OPN–wild-type (WT) and OPN-knockout (KO) mice 21 days after the
initial injection of K/BxN serum. Inflammation, bone erosion (arrows), and formation of new woven bone (arrowhead) were evident on hematoxylin
and eosin (H&E)–stained sections (A) of both the WT and KO mice. Pannus formation and associated cartilage destruction (arrow) were observed
in toluidine blue–stained sections (B) of both the WT and KO mice. Cartilage damage and proteoglycan loss also occurred in areas remote from
pannus. C, KO mice demonstrated bone erosion in H&E-stained sections of the forefoot region. D, Staining for tartrate-resistant acid phosphatase
(TRAP) activity in a section adjacent to that in C revealed numerous TRAP? multinucleated osteoclast-like cells at sites of bone erosion (arrows
in C and D). (Original magnification ? 50.)
ROLE OF OPN IN AUTOANTIBODY-INDUCED ARTHRITIS IN K/BxN MICE2689
from the same mice (0.28 ? 0.02 ?g/ml). The elevated
OPN protein in synovial fluid suggests that OPN was
produced locally in the joint, consistent with the mRNA
increase. In contrast to the local up-regulation of OPN
in ankles, no change in peripheral blood OPN mRNA
expression or serum OPN protein was observed during
K/BxN serum-transferred arthritis (data not shown).
Lack of protection of OPN deficiency against
inflammation in serum-transferred arthritis at a range
of disease severities. To test the physiologic relevance of
OPN to K/BxN serum-transferred arthritis, we injected
K/BxN serum into tenth-generation B6-backcrossed
OPN-deficient mice and their matched OPN-sufficient
littermates. The kinetics and severity of arthritis, as
measured by ankle thickening and clinical index, in the
wild-type and heterozygous OPN-sufficient mice were
comparable with those in the OPN-knockout mice (Fig-
ures 2A and B and 3A and B).
Synovial inflammation was assessed in H&E-
stained sections of the tibiotalar and forefoot regions.
Thickening of the synovial lining, mononuclear cell
infiltration, and pannus formation were observed both in
OPN-sufficient and in OPN-deficient mice (Figure 4A).
H&E-stained sections (n ? 4) taken on day 15 were
scored by a blinded observer on a 0–5 scale, correspond-
ing to the degree of inflammatory cell infiltration into
the ankles. Consistent with the clinical findings, OPN-
sufficient and OPN-deficient mice had comparable his-
tologic inflammation scores (2.8 ? 0.4 [mean ? SEM]
versus 3.1 ? 0.2, respectively).
It remained possible that differences between
OPN-sufficient and OPN-deficient mice were obscured
because the system was overloaded. A dose-response
experiment was performed using doses of K/BxN serum
corresponding to a range of disease severities encom-
passing mild to severe arthritis. No differences in the
clinical parameters of arthritis were observed between
OPN-sufficient and OPN-deficient mice at any of the
doses tested (Figure 5).
Comparable histologic features of bone erosion
and cartilage damage in arthritic OPN-deficient and
OPN-sufficient mice. We considered the possibility that
OPN-deficient mice have the same level of inflammation
as OPN-sufficient mice, but with reduced bone erosion
and/or cartilage damage. H&E-stained sections of the
tibiotalar region revealed bone erosions in both OPN-
sufficient and OPN-deficient arthritic mice (Figure 4A).
Bone erosion was also observed in the forefoot region of
OPN-deficient mice (Figure 4C) and OPN-sufficient
mice (results not shown), in some cases with marked
trabecular and subchondral bone loss. Staining of serial
sections of the forefoot region for TRAP activity re-
vealed numerous TRAP? multinucleated osteoclast-
like cells at sites of bone erosion in both groups (Figure
4D and data not shown). As has been reported previ-
ously with this model (33), we observed not only bone
erosion, but also the formation of new woven bone
undergoing remodeling by multinucleated osteoclast-
like cells. This process was seen in the joints of both
OPN-deficient (Figure 4A) and OPN-sufficient (results
not shown) mice.
Damage to the cartilage was assessed on tolu-
idine blue– and H&E-stained sections of the tibiotalar
and forefoot regions. Invasion of the pannus into carti-
lage was associated with loss of cartilage matrix and
proteoglycan content in OPN-sufficient and OPN-
deficient mice (Figure 4B). Proteoglycan and cartilage
matrix loss in areas distal from pannus also occurred in
both groups (Figure 4B), as did full-depth cartilage loss,
generally in association with loss of the underlying
subchondral bone (Figure 4C and data not shown).
We directly compared the extent of bone erosion
and cartilage damage in OPN-sufficient and OPN-
deficient mice by semiquantitative histopathologic scor-
ing. Sections were taken from mice killed 7, 14, or 21
days after the initial K/BxN serum injection. The fore-
foot and tibiotalar regions were evaluated independently
to reflect the variation in disease severity that can occur
in these areas. Bone erosion on H&E-stained sections
was scored on a 0–5 scale. OPN-deficient and OPN-
sufficient mice demonstrated comparable erosion in the
tibiotalar and forefoot regions at the 3 time points
(Figure 3C). The degree of bone erosion was generally
proportionate to the degree of inflammation in all mice
(data not shown), consistent with previously published
observations (33). Surface cartilage damage and carti-
lage destruction secondary to pannus invasion were
scored on toluidine blue– and H&E-stained sections,
respectively. The degree of cartilage damage (Figure
3D) and pannus destruction of cartilage (data not
shown) was similar in arthritic OPN-deficient and OPN-
sufficient mice. Comparable histologic scores for bone
erosion and cartilage damage were also found for ar-
thritic OPN-sufficient and OPN-deficient mice killed on
day 15 (data not shown).
OPN-deficient mice injected with K/BxN serum
developed inflammation, cartilage damage, and bone
erosion comparable with that of OPN-sufficient con-
trols. Previous studies have examined the role of OPN in
2690JACOBS ET AL
anti-CII antibody–induced arthritis (12,24). This model
bears close mechanistic resemblance to K/BxN serum-
transferred arthritis. In both systems, transferred auto-
antibodies deposit in the joints and initiate pathogenic
inflammation via complement activation and cellular
innate immunity. Transfer experiments with various
knockout mice have yielded largely concordant results
between the two models, including demonstration of
crucial roles for interleukin-1 receptor, C5a receptor,
Fc? receptor (Fc?R), Fc?RIIB, and Fc?RIII (33–38).
The similar pathogenic mechanisms suggest that the role
of OPN in each model should also be comparable.
Our findings with the K/BxN serum-transfer
model differed from those of Yumoto et al (24), who
observed that OPN-deficient mice were resistant to
anti-CII antibody–induced arthritis, but were consistent
with those of Blom et al (12), who found no such
protective effect. Both groups used similar experimental
methods—cotransfer of anti-CII antibodies with lipo-
polysaccharide (LPS) into OPN-sufficient and OPN-
deficient mice—except that different sources of antibod-
ies and different lines of OPN-knockout mice were used.
It could have been argued that the system used by Blom
et al was not a robust one because only 1 of 4 injected
Figure 5. Similar response of osteopontin (OPN)–deficient and OPN-sufficient mice to a range of doses
of K/BxN serum. OPN-sufficient (heterozygous [HZ]) and OPN-deficient (KO) mice were injected on
days 0 and 2 with 150, 75, 37.5, or 18.75 ?l of K/BxN serum (n ? 2 mice at each dose). The ankle
thickening and clinical index were measured from day 0 to day 15 (150-?l dose group) or day 21 (all other
dose groups). Data for the 150-?l dose group were incorporated into Figure 2.
ROLE OF OPN IN AUTOANTIBODY-INDUCED ARTHRITIS IN K/BxN MICE2691
OPN-heterozygous mice and 3 of 10 injected OPN-
knockout mice developed arthritis. This was not true of
our system, since we observed a 100% incidence of
arthritis after K/BxN serum transfer and similar disease
kinetics and severity within batches of serum. The
difference between our results and those of Yumoto et
al also cannot be attributed to overloading the system,
since OPN-deficient and OPN-sufficient mice had com-
parable responses to a wide range of serum doses.
As also argued by Blom et al (12), the discordant
results reported by Yumoto et al (24) might be attribut-
able to polymorphic genes linked to OPN. Yumoto et al
compared OPN-knockout mice and wild-type litter-
mates of mixed B6/129Sv F2background. Such a com-
parison does not control for the potential confounding
effects of 129Sv genes linked to the OPN mutant allele.
OPN is situated in a region on mouse chromosome 5
that carries many genes that could influence an inflam-
matory disease, including a cluster of CXCL chemo-
kines, and has been found to encode quantitative trait
loci associated with arthritis and EAE (12). Blom et al
used OPN-knockout mice backcrossed onto the
C57BL/10 background for 12 generations in order to
minimize the possibility of any linked genes having an
effect. The OPN knockout mice used in our study were
derived from the same mutant line used by Yumoto et
al, but the mutation has been backcrossed onto the B6
background for 10 generations, thereby minimizing con-
founding influences of 129Sv genome segments.
Recently, Yamamoto et al (25) reported that
prophylactic administration of polyclonal antibodies
against the SLAYGLR epitope on thrombin-cleaved
OPN moderately reduced the clinical severity of anti-CII
antibody–induced arthritis and delayed its onset. The
polyclonal antibodies also slowed the progression of
disease when given therapeutically. These findings raise
the possibility that when OPN is absent from develop-
ment, it may not be required for inflammation, since
other factors could compensate in its absence, but in
settings where OPN is present, blocking OPN may have
a modulatory effect on inflammation. Alternatively, the
apparent role of OPN in these experiments may reflect
the use of LPS to boost arthritis induction (39,40). LPS
induces macrophage expression of OPN in vitro and
could induce local expression of OPN within joints by
synovial macrophages in anti-CII antibody–induced ar-
thritis (41). Since LPS also activates coagulation, much
of this OPN may be cleaved by thrombin (42). This
conversion could facilitate the proinflammatory activity
of OPN, since thrombin-cleaved, but not full-length,
OPN induces monocyte migration in vitro (25). Since the
dose of anti-CII antibodies used in previous studies was
insufficient to induce arthritis without LPS, inhibiting
the effects of LPS would, by itself, suppress arthritis,
even if the arthritogenic effects of the anti-CII antibod-
ies are unaffected. Hence, our results and those of
Yamamoto et al are consistent with OPN having an
important mechanistic role downstream of LPS but not
OPN is potentially relevant to arthritis not only
for its effects on inflammation, but also for its role in
bone resorption. Bone resorption induced in vivo by
experimental stresses, including estrogen withdrawal,
ectopic transplantation, and mechanical stress, is atten-
uated in OPN-deficient bone (15,17,18). In culture,
OPN-deficient bone showed reduced bone resorption
and osteoclast differentiation in response to RANKL, an
osteoclast differentiation factor that plays a central role
in the regulation of bone remodeling (18,43). In K/BxN
serum-transferred arthritis, RANKL-deficient mice had
inflammation comparable with that of control mice, but
had dramatically reduced bone erosion (32). If OPN is a
downstream factor in osteoclast-mediated bone resorp-
tion, OPN-deficient mice injected with K/BxN serum
would be expected to have diminished bone resorption
compared with OPN-sufficient mice, even though they
had comparable inflammation. In a routine histopatho-
logic analysis of H&E-stained joint sections from OPN-
deficient and OPN-sufficient arthritic mice, both groups
demonstrated similar inflammation and bone resorp-
tion. Subtle differences in bone erosion between the two
groups would not be detected in our analysis but might
be observed using more sensitive methods of quantita-
tion. Nevertheless, significant osteoclast-mediated bone
erosion does occur in the absence of OPN in this
arthritis model. We conclude that if OPN does act
downstream of RANKL, it is not a necessary mediator
of bone erosion in antibody-mediated arthritis.
We have shown that OPN does not play a re-
quired role in inflammation, bone erosion, or cartilage
damage in the K/BxN serum-transfer model of
autoantibody-mediated arthritis. Indeed, in the absence
of OPN, these 3 parameters of arthritis are unaltered. It
remains possible that OPN is involved in the recruitment
of inflammatory cells or in bone erosion during arthritis
but the effects of its absence are too small to be detected
in our analysis, or that in its absence, other factors
compensate. In either scenario, OPN by itself is not an
attractive target for reducing inflammation and bone
destruction in RA patients.
2692JACOBS ET AL
We would like to thank Quynh-Mai Pham for assisting
with the OPN K/BxN serum-transfer experiments and Robert
Saccone, Jennifer Johnson, and Joyclyn Yee from the core
facilities of the Joslin Diabetes and Endocrinology Research
Center for assisting with the microarray experiments.
1. Denhardt DT, Noda M. Osteopontin expression and function: role
in bone remodeling. J Cell Biochem Suppl 1998;30:92–102.
2. O’Regan A, Berman JS. Osteopontin: a key cytokine in cell-
mediated and granulomatous inflammation. Int J Exp Pathol
3. Denhardt DT, Noda M, O’Regan AW, Pavlin D, Berman JS.
Osteopontin as a means to cope with environmental insults:
regulation of inflammation, tissue remodeling, and cell survival.
J Clin Invest 2001;107:1055–61.
4. Giachelli CM, Lombardi D, Johnson RJ, Murray CE, Almeida M.
Evidence for a role of osteopontin in macrophage infiltration in
response to pathological stimuli in vivo. Am J Pathol 1998;152:
5. Weiss JM, Renkl AC, Maier CS, Immig M, Liaw L, Ahrens T, et
al. Osteopontin is involved in the initiation of cutaneous contact
hypersensitivity by inducing Langerhans and dendritic cell migra-
tion to lymph nodes. J Exp Med 2001;194:1219–29.
6. Ashkar S, Weber GF, Panoutsakopoulou V, Sanchirico ME,
Jansson M, Zawaideh S, et al. Eta-1 (osteopontin): an early
component of type-1 (cell-mediated) immunity. Science 2000;287:
7. O’Regan AW, Chupp GL, Lowry JA, Goetschkes M, Mulligan N,
Berman JS. Osteopontin is associated with T cells in sarcoid
granulomas and has T cell adhesive and cytokine-like properties in
vitro. J Immunol 1999;162:1024–31.
8. Lampe MA, Patarca R, Iregui MV, Cantor H. Polyclonal B cell
activation by the Eta-1 cytokine and the development of systemic
autoimmune disease. J Immunol 1991;147:2902–6.
9. Iizuka J, Katagiri Y, Tada N, Murakami M, Ikeda T, Sato M, et al.
Introduction of an osteopontin gene confers the increase in B1 cell
population and the production of anti-DNA autoantibodies. Lab
10. Chabas D, Baranzini SE, Mitchell D, Bernard CC, Rittling SR,
Denhardt DT, et al. The influence of the proinflammatory cyto-
kine, osteopontin, on autoimmune demyelinating disease. Science
11. Jansson M, Panoutsakopoulou V, Baker J, Klein L, Cantor H.
Cutting edge: attenuated experimental autoimmune encephalomy-
elitis in Eta-1/osteopontin-deficient mice. J Immunol 2002;168:
12. Blom T, Franzen A, Heinegard D, Holmdahl R. Comment on the
influence of the proinflammatory cytokine, osteopontin, on auto-
immune demyelinating disease [letter]. Science 2003;299:1845.
13. Rittling SR, Matsumoto HN, McKee MD, Nanci A, An X, Novick
KE, et al. Mice lacking osteopontin show normal development and
bone structure but display altered osteoclast formation in vitro.
J Bone Miner Res 1998;13:1101–11.
14. Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JM, Hogan BL.
Altered wound healing in mice lacking a functional osteopontin
gene (spp1). J Clin Invest 1998;101:1468–78.
15. Yoshitake H, Rittling SR, Denhardt DT, Noda M. Osteopontin-
deficient mice are resistant to ovariectomy-induced bone resorp-
tion. Proc Natl Acad Sci U S A 1999;96:8156–60.
16. Chellaiah MA, Kizer N, Biswas R, Alvarez U, Strauss-Schoen-
berger J, Rifas L, et al. Osteopontin deficiency produces osteoclast
dysfunction due to reduced CD44 surface expression. Mol Biol
17. Ishijima M, Rittling SR, Denhardt DT, Noda M. Enhancement of
osteoclastic bone resorption and suppression of osteoblastic bone
formation in response to reduced mechanical stress do not occur in
the absence of osteopontin. Proc Natl Acad Sci U S A 1999;96:
18. Asou Y, Rittling SR, Yoshitake H, Tsuji K, Shinomiya K, Nifuji A,
et al. Osteopontin facilitates angiogenesis, accumulation of oste-
oclasts, and resorption in ectopic bone. Endocrinology 2001;142:
19. Ihara H, Denhardt DT, Furuya K, Yamashita T, Muguruma Y,
Tsuji K, et al. Parathyroid hormone-induced bone resorption does
not occur in the absence of osteopontin. J Biol Chem 2001;276:
20. Gravallese EM. Osteopontin: a bridge between bone and the
immune system. J Clin Invest 2003;112:147–9.
21. Petrow PK, Hummel KM, Schedel J, Franz JK, Klein CL, Muller-
Ladner U, et al. Expression of osteopontin messenger RNA and
protein in rheumatoid arthritis. Arthritis Rheum 2000;43:
22. Ohshima S, Yamaguchi N, Nishioka K, Mima T, Ishii T, Umeshita-
Sasai M, et al. Enhanced local production of osteopontin messen-
ger RNA and protein in rheumatoid arthritis. J Rheumatol
23. Ohshima S, Kobayashi H, Yamaguchi N, Nishioka K, Umeshita-
Sasai M, Mima T, et al. Expression of osteopontin at sites of bone
erosion in a murine experimental arthritis model of collagen-
induced arthritis. Arthritis Rheum 2002;46:1094–101.
24. Yumoto K, Ishijima M, Rittling SR, Tsuji K, Tsuchiya Y, Kon S, et
al. Osteopontin deficiency protects joints against destruction in
anti-type II collagen antibody-induced arthritis in mice. Proc Natl
Acad Sci U S A 2002;99:4556–61.
25. Yamamoto N, Sakai F, Kon S, Morimoto J, Kimura C, Yamazaki
H, et al. Essential role of the cryptic epitope SLAYGLR within
osteopontin in a murine model of rheumatoid arthritis. J Clin
26. Kouskoff V, Korganow AS, Duchatelle V, Degott C, Benoist C,
Mathis D. Organ-specific disease provoked by systemic autoreac-
tivity. Cell 1996;87:811–22.
27. Korganow AS, Ji H, Mangialaio S, Duchatelle V, Pelanda R,
Martin T, et al. From systemic T cell self-reactivity to organ-
specific autoimmune disease via immunoglobulins. Immunity
28. Matsumoto I, Staub A, Benoist C, Mathis D. Arthritis provoked by
linked T and B cell recognition of a glycolytic enzyme. Science
29. Matsumoto I, Maccioni M, Lee DM, Maurice M, Simmons B,
Brenner M, et al. How antibodies to a ubiquitous cytoplasmic
enzyme may provoke joint-specific autoimmune disease. Nat
30. Glowacki J, Rey C, Glimcher MJ, Cox KA, Lian J. A role for
osteocalcin in osteoclast differentiation. J Cell Biochem 1991;45:
31. Bendele A, McAbee T, Sennello G, Frazier J, Chlipala E, McCabe
D. Efficacy of sustained blood levels of interleukin-1 receptor
antagonist in animal models of arthritis: comparison of efficacy in
animal models with human clinical data. Arthritis Rheum 1999;
32. Pettit AR, Ji H, von Stechow D, Muller R, Goldring SR, Choi Y,
et al. TRANCE/RANKL knockout mice are protected from bone
erosion in a serum transfer model of arthritis. Am J Pathol
33. Ji H, Pettit A, Ohmura K, Ortiz-Lopez A, Duchatelle V, Degott C,
et al. Critical roles for interleukin-1 and tumor necrosis factor-? in
antibody-induced arthritis. J Exp Med 2002;196:77–85.
34. Ji H, Gauguier D, Ohmura K, Gonzalez A, Duchatelle V, Danoy
ROLE OF OPN IN AUTOANTIBODY-INDUCED ARTHRITIS IN K/BxN MICE2693
P, et al. Genetic influences on the end-stage effector phase of
arthritis. J Exp Med 2001;194:321–30.
35. Ji H, Ohmura K, Mahmood U, Lee DM, Hofhuis FM, Boackle SA,
et al. Arthritis critically dependent on innate immune system
players. Immunity 2002;16:157–68.
36. Kagari T, Doi H, Shimozato T. The importance of IL-1? and TNF-?,
and the noninvolvement of IL-6, in the development of monoclonal
antibody-induced arthritis. J Immunol 2002;169:1459–66.
37. Grant EP, Picarella D, Burwell T, Delaney T, Croci A, Avitahl N,
et al. Essential role for the C5a receptor in regulating the effector
phase of synovial infiltration and joint destruction in experimental
arthritis. J Exp Med 2002;196:1461–71.
38. Kagari T, Tanaka D, Doi H, Shimozato T. Essential role of Fc?
receptors in anti-type II collagen antibody-induced arthritis. J Im-
39. Terato K, Hasty KA, Reife RA, Cremer MA, Kang AH, Stuart
JM. Induction of arthritis with monoclonal antibodies to collagen.
J Immunol 1992;148:2103–8.
40. Terato K, Harper DS, Griffiths MM, Hasty DL, Ye XJ, Cremer
MA, et al. Collagen-induced arthritis in mice: synergistic effect of
E. coli lipopolysaccharide bypasses epitope specificity in the
induction of arthritis with monoclonal antibodies to type II
collagen. Autoimmunity 1995;22:137–47.
41. Guo H, Cai CQ, Schroeder RA, Kuo PC. Osteopontin is a negative
feedback regulator of nitric oxide synthesis in murine macro-
phages. J Immunol 2001;166:1079–86.
42. Okajima K. Regulation of inflammatory responses by natural
anticoagulants. Immunol Rev 2001;184:258–74.
43. Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation and
activation. Nature 2003;423:337–42.
2694JACOBS ET AL