вход по аккаунту


Reversal of the immunosuppressive properties of mesenchymal stem cells by tumor necrosis factor ╨Ю┬▒ in collagen-induced arthritis.

код для вставкиСкачать
Vol. 52, No. 5, May 2005, pp 1595–1603
DOI 10.1002/art.21012
© 2005, American College of Rheumatology
Reversal of the Immunosuppressive Properties of
Mesenchymal Stem Cells by Tumor Necrosis Factor ␣ in
Collagen-Induced Arthritis
Farida Djouad,1 Vanessa Fritz,1 Florence Apparailly,1 Pascale Louis-Plence,1 Claire Bony,1
Jacques Sany,2 Christian Jorgensen,3 and Danièle Noël1
Objective. Adult mesenchymal stem cells (MSCs)
represent promising tools for therapeutic applications
such as tissue engineering and cellular therapy. Recent
data suggest that, due to their immunosuppressive
nature, MSCs may be of interest to enhance allogeneic
hematopoietic engraftment and prevent graft-versushost disease. Using a murine model of rheumatoid
arthritis (RA), this study investigated whether the immunosuppressive properties of MSCs could be of therapeutic value to inhibit reactive T cells in autoimmune
diseases such as RA.
Methods. In mice with collagen-induced arthritis
(CIA), we injected various doses of C3 MSCs at the time
of immunization or booster injection, and subsequently
evaluated the clinical and immunologic parameters. The
immunosuppressive properties of MSCs were determined in vitro in mixed lymphocyte reactions with or
without the addition of tumor necrosis factor ␣ (TNF␣).
Results. In the CIA model of arthritis, MSCs did
not confer any benefit. Both the clinical and the immunologic findings suggested that MSCs were associated
with accentuation of the Th1 response. Using luciferaseexpressing MSCs, we were unable to detect labeled cells
in the articular environment of the knee, suggesting that
worsening of the symptoms was unlikely due to the
homing of MSCs in the joints. Experiments in vitro
showed that the addition of TNF␣ was sufficient to
reverse the immunosuppressive effect of MSCs on T cell
proliferation, and this observation was associated with
an increase in interleukin-6 secretion.
Conclusion. Our data suggest that environmental
parameters, in particular those related to inflammation,
may influence the immunosuppressive properties of
Mesenchymal stem cells (MSCs) are adult progenitor cells present in the bone marrow that are able to
differentiate into several lineages, such as chondrocytes,
osteoblasts, tendinocytes, myocytes, and adipocytes (1).
Due to their differentiation potential and the ease of
expansion, they are largely studied for their use in tissue
engineering for bone and cartilage repair. Numerous
reports have focused on their phenotypic and functional
characterization (for review, see refs. 2 and 3) and, in
particular, it was shown that MSCs display immunosuppressive capacities. This was first shown in vitro, in
experiments in which MSCs were able not only to escape
recognition by alloreactive T cells, but also, when added
in mixed lymphocyte reactions (MLRs), to suppress the
proliferation of T cells (4–6). Our group recently demonstrated that this immunosuppressive effect acts
through the generation of CD8⫹ regulatory T cells (6).
In vivo, intravenous injection of MSCs was also shown to
prolong graft survival in major histocompatibility
complex–mismatched recipient baboons (7).
The immunosuppressive features of MSCs are of
clinical relevance in allogeneic transplantation because
it is expected that the incidence and/or severity of
graft-versus-host disease (GVHD) will be reduced. Consistent with this notion, a recent clinical study showed
that infusion of haploidentical MSCs in a patient with
Supported by the European Community (5th PCRDT program) in the context of the “Stemgenos” consortium (QLRT-200102039).
Farida Djouad, MS, Vanessa Fritz, MS, Florence Apparailly,
PhD, Pascale Louis-Plence, PhD, Claire Bony, MS, Danièle Noël,
PhD: INSERM U475, Montpellier, France; 2Jacques Sany, MD, PhD:
Hôpital Lapeyronie, Montpellier, France; 3Christian Jorgensen, MD,
PhD: INSERM U475 and Hôpital Lapeyronie, Montpellier, France.
Drs. Jorgensen and Noël contributed equally to this work.
Address correspondence and reprint requests to Danièle
Noël, MD, INSERM U475, 99 rue Puech Villa, 34197 Montpellier
Cedex 5, France. E-mail:
Submitted for publication October 20, 2004; accepted in
revised form January 25, 2005.
severe treatment-resistant, grade IV, acute GVHD resulted in a striking immunosuppressive effect as long as
1 year after treatment (8). Other applications of MSCs,
such as in autoimmune diseases, may be of interest
because they may be used in total bone marrow transplantation (BMT), which consists of both hematopoietic
stem cells and stromal cells (9). In the MRL/lpr mouse
model of lupus, it was reported that total BMT resulted
in complete absence of the disease at 40 weeks after
treatment and increased survival at 1 year (10). When
the adherent cells were removed before transplantation,
75% of the mice died within 90 days. These data suggest
that use of MSCs may be of potential interest for the
treatment of diseases in which suppression of T cell
activation is of major importance.
Rheumatoid arthritis (RA) is an autoimmune
disease that is characterized by chronic inflammation of
the joints. Cumulative evidence suggests that CD4⫹ T
cell–mediated autoimmune responses play a critical role
in the pathogenesis of RA, in conjunction with the
activity of B cells and macrophages that infiltrate the
synovium (11). Tumor necrosis factor ␣ (TNF␣) is
secreted by monocyte/macrophages and fibroblasts and
plays a central role in RA by inducing a cascade of
cytokines, including interleukin-1 (IL-1), IL-6, IL-15,
IL-18, and granulocyte–macrophage colony-stimulating
factor. Recent therapeutic approaches have involved the
use of inhibitors capable of blocking either the binding
of TNF␣ or IL-1 to its cell-surface receptors or the
response of CD4⫹ T cells. Three anti-TNF drugs (a
chimeric mouse/human and a fully human antibody to
TNF, and a TNF receptor–immunoglobulin fusion molecule) have been proven to be effective and safe in
appropriate and well-conducted clinical trials and have
shown effectiveness in slowing and even arresting the
progression of radiographic damage (12).
Therapeutic agents with activity against T cells,
including leflunomide, CTLA-4Ig, anti-CD4 antibodies,
cyclosporine, tacrolimus, and T cell receptor V␤-chain
vaccination strategies, have also been studied in RA.
Combination therapies that include any of these T
cell–activation inhibitors in conjunction with non–T cell–
specific agents, such as methotrexate, antimalarial
agents, or anti-TNF biologic agents, may prove to be the
most effective strategies in controlling this complex
disease (13). The collagen-induced arthritis (CIA)
model, induced with bovine type II collagen (CII) in
DBA/1 mice, is the prototype model of RA (14). This
model thus provides a suitable tool to investigate strategies aimed at inhibiting the proliferation of the T cell
compartment and may provide a preclinical model for
treatment evaluation.
The aim of our study was to determine whether
the immunosuppressive properties of MSCs could be of
therapeutic value in autoimmune diseases, by inhibiting
autoreactive CD4⫹ T cell proliferation. We used the
CIA model to explore the effect of MSCs on the disease
course. The results show that the injection of MSCs did
not produce any beneficial effect on arthritic symptoms.
These observations were shown, at least in vitro, to be
related to the presence of TNF␣, which resulted in
reversion of the immunosuppression mediated by MSCs.
Cell culture and transfection experiments. The murine
MSCs C3H10T1/2 (C3) were grown in complete medium,
consisting of Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, and
100 ␮g/ml streptomycin. Transfection of C3 cells was performed according to the standard calcium phosphate coprecipitation procedure. The hygromycin-resistant pREP10 vector
expressing the viral IL-10 (vIL-10) gene under control of the
Rous sarcoma virus promoter, named pREP10-vIL-10, was
kindly provided by Dr. C. Verwaerde (Institut Pasteur, Lille,
France). The neomycin-resistant pCMV-Luc⫹-Neo vector encoding the luciferase gene under control of the cytomegalovirus promoter was a kind gift from Dr. P. Balaguer (15). Stable
clones were obtained after selection using either 400 ␮g/ml of
hygromycin or 1 mg/ml of G418 (Invitrogen, Cergy, France).
Quantification of luciferase activity. The activity of
firefly luciferase was quantified in cell extracts obtained from
cells or tissues. These cell extracts were prepared using the
Luciferase Reporter Gene Assay, High Sensitivity (Roche
Diagnostics, Neuilly-sur-Seine, France), carried out according
to the supplier’s protocol. Luciferase activity was normalized
according to the cell number or to the weight of recovered
Quantification of cytokines by enzyme-linked immunosorbent assay (ELISA). Secretion of murine IL-1␤, IL-2,
IL-4, IL-6, IL-10, interferon-␥ (IFN␥), and TNF␣ in culture
supernatants was determined by ELISAs (BD Biosciences, Le
Pont de Claix, France). Viral IL-10 production was quantified
using the ready-SET-Go kit (Clinisciences, Montrouge,
Mixed lymphocyte reactions. MLRs were performed as
previously described (6). When necessary, MSCs were added
to the MLR to obtain a 300-␮l final volume. Whenever tested,
recombinant murine TNF␣ (R&D Systems, Lille, France) was
added at concentrations of 50–500 ng/ml. Each experiment was
performed in triplicate and repeated at least 3 times.
Animals. Adult (8–10-week-old) male DBA/1 mice
were grown in our animal facilities, and experiments were
conducted in accordance with the recommendations of the
European Convention for the Protection of Vertebrate Animals Used for Experimentation. Between 5 and 11 mice per
group were included depending on the experiment, and experiments were done at least twice.
Induction of arthritis. Bovine CII was emulsified with
an equal volume of Freund’s complete adjuvant. Mice were
injected at the base of the tail with 100 ␮l of emulsion
containing 100 ␮g of CII. On day 21, animals received a
booster of CII emulsion in Freund’s incomplete adjuvant. C3
cells (106 or 4 ⫻ 106 cells/100 ␮l phosphate buffered saline)
were injected in the tail vein either on the day of primary
immunization or at the time of the booster injection. For the
homing experiments, cells were inoculated after arthritis onset
(day 32).
Development of CIA was assessed every 2–3 days. Paw
swelling was assessed by measuring the thickness of the hind
paws using a caliper. The maximal thickness observed in either
of the 2 hind paws of each mouse was also determined during
the time course of the disease. The clinical score was assessed
using the following system: grade 0 ⫽ no swelling, grade 1 ⫽
ⱖ0.1 mm increase in paw swelling, grade 2 ⫽ ⱖ0.2 mm
increase in paw swelling, grade 3 ⫽ extensive swelling (ⱖ0.3
mm) with severe joint deformity, and grade 4 ⫽ pronounced
swelling (ⱖ0.45 mm) with pronounced joint deformity. After
the animals were killed (between day 42 and day 48 after
immunization), the hind limbs were collected for radiography
and fixed for histology. Radiologic and histologic scoring were
performed as described previously (16).
Assessment of in vitro T cell function. Splenocytes
were collected and cultured at a density of 5 ⫻ 105 cells/well in
200 ␮l of RPMI medium, supplemented with 1% autologous
serum, 2 mM glutamine, 0.1 mM nonessential amino acids, 20
mM HEPES, 1 mM sodium pyruvate, and 5 ⫻ 10⫺5M
2-mercaptoethanol in 96-well round-bottom plates, in the
presence or absence of 5 ␮g/ml concanavalin A. For proliferation assays, cells were cultured for 3 days, and 1 ␮Ci/well of
H-thymidine was added for an additional 18 hours of culture.
For cytokine production, supernatants from 2 ⫻ 106 splenocytes were collected after 24 hours of culture.
Measurement of serum anti-CII antibody levels. Serum samples were collected on day 45 for the detection of
anti-CII IgG1 and IgG2a antibodies by ELISA, as described
previously (16). Results were expressed as the ratio of IgG2a
antibody levels to IgG1 antibody levels at a 1:200 dilution of
Statistical analysis. Statistical comparisons were done
with the Student’s t-test or an impaired Mann-Whitney test to
compare nonparametric data for statistical significance. Percentage comparisons were done using the chi-square test. All
data were analyzed by the program Instat (Graphpad, San
Diego, CA).
Dose-dependent effects of systemic injection of
MSCs on paw swelling in the CIA model. Since we had
previously shown that injection of MSCs could display
immunosuppressive effects (6), we wondered whether
MSCs could prevent or diminish arthritic symptoms in
the CIA model. We first tested the dose of MSCs to
determine a potential dose-response effect. We observed that 4 ⫻ 106 MSCs, versus 106 MSCs, led to an
Figure 1. Effects of dose and time of injection of C3 mesenchymal
stem cells on paw swelling in collagen-induced arthritis. The results are
representative of 1 of 2 experiments with the number of mice varying
from 5 to 10 in each group. ⴱ ⫽ P ⬍ 0.1 versus controls. d0 ⫽ day 0;
d21 ⫽ day 21.
increase in paw swelling (results not shown). We then
investigated whether the time of injection, at immunization or at booster (day 21), may influence the disease
We observed a worsening of the symptoms when
cells were injected closer to the time of onset of CIA
(data not shown). We thus combined the 2 doses of cells
and the 2 times of injection in the same experiment.
According to the assessment of paw swelling, a statistically significant increase could be observed in the mice
receiving the highest dose of MSCs on day 21 (Figure 1).
The maximal paw thickness also tended to be worsened
in this group, although the differences were not statistically significant (Table 1). A significantly higher incidence of arthritis, together with higher radiologic and
histologic scores, was observed in the groups of mice
injected with high doses of MSCs on day 21 (Table 1).
These data suggest that MSCs do not display any
beneficial effect in mice with CIA and tend to worsen
clinical symptoms when injected at a high dose and close
to the onset of arthritis.
In parallel, we investigated the mechanisms underlying the severity of CIA in this experiment. Because
the severity of CIA is reflected by the switch from a Th1
to a Th2 response (17), we measured the levels of IgG2a
and IgG1 anti-CII antibodies in the serum. In 3 of 4
groups injected with C3 cells, the IgG2a:IgG1 ratios
were higher than in the control group of arthritic mice
(Table 1).
We also analyzed the cytokine production by
spleen cells to discriminate between a Th1 and Th2
response of effector cells (17). Proinflammatory cyto-
Table 1. Clinical parameters and cytokine levels in treated mice versus control mice with collagen-induced arthritis
Day 0
Clinical parameter
% arthritis incidence
Maximal paw thickness, mean ⫾ SD mm
Days to arthritis onset, mean ⫾ SD
% with severe radiologic score
% with severe histologic score
IgG2a:IgG1, mean ⫾ SD
Cytokine, pg/ml§
Day 21
106 C3
4 ⫻ 106 C3
106 C3
4 ⫻ 106 C3
2.29 ⫾ 0.41
32.5 ⫾ 1.22
0.26 ⫾ 0.34
2.37 ⫾ 0.36
35 ⫾ 5.51
0.37 ⫾ 0.79†
2.22 ⫾ 0.04
0.25 ⫾ 0.55
2.33 ⫾ 0.35
35.22 ⫾ 5.02
0.77 ⫾ 1.13†
2.57 ⫾ 0.48
32.5 ⫾ 1.22
0.57 ⫾ 0.83†
* P ⬍ 0.001 versus controls.
† P ⬍ 0.01 versus controls.
‡ P ⬍ 0.05 versus controls.
§ Cytokine levels were determined in the supernatants of 2 ⫻ 106 splenocytes stimulated with concanavalin A for 24 hours. Results are the mean
level of cytokines expressed in the supernatant of splenocytes from pooled spleens in each group. IFN␥ ⫽ interferon-␥; IL-2 ⫽ interleukin-2;
TNF␣ ⫽ tumor necrosis factor ␣.
kines (IFN␥, TNF␣, IL-1␤, and IL-2) were either stable
or enhanced in all arthritic groups as compared with
those in the control group, but the highest increases
were observed in the group receiving 4 ⫻ 106 C3 cells on
day 21 (Table 1). In all treated arthritic groups compared with the control group, the levels of antiinflammatory cytokines (IL-4 and IL-10) were also enhanced.
Nevertheless, the ratio of IFN␥:IL-4 was enhanced by 2
fold (mean values between 8.9 and 10.9 among the
treated groups) as compared with the control mice
(mean 5.7), which is indicative of a higher Th1 response.
Taken together, these data demonstrate the accentuation of the Th1 helper response when MSCs are injected
in arthritic mice.
Effects of IL-10–expressing MSCs on CIIinduced arthritic symptoms. Because MSCs do not
display any immunosuppression in the CIA model, we
investigated whether the use of genetically modified
MSCs, expressing an antiinflammatory cytokine, could
improve the course of the disease. Since our group and
other investigators have already shown the efficacy of
the viral form of IL-10 (vIL-10) as a possible candidate
for the treatment of arthritis (16,18–21), we derived
clones of C3 MSCs that expressed vIL-10. In MLRs, we
assessed the immunosuppressive effect of the highest
producer clone (580 pg IL-10/106 cells/24 hours), hereafter referred to as C3 IL-10, by comparing the effect of
decreasing concentrations of naive C3 cells with that of
IL-10–expressing C3 cells. Using this experimental con-
struct, we found that secretion of IL-10 significantly enhanced the antiproliferative activity of engineered MSCs
(Figure 2A).
We then compared the effect of C3 IL-10 cells
with that of C3 cells in the CIA model. To this aim, we
injected 106 cells in the tail vein of mice at the time of
CII immunization. This dose and time of cell injection
were chosen to prevent an aggravating effect due to the
presence of MSCs, while still permitting significant
IL-10 secretion. As shown in Figure 2B, all groups of
mice developed arthritis as recorded by the increase in
paw swelling. Similarly, no statistically significant difference was observed between the C3- and C3 IL-10–
treated mice and the control mice, in neither the clinical
score (mean ⫾ SD 5.3 ⫾ 4.23, 8.0 ⫾ 6.29, and 3.6 ⫾ 3.91,
respectively), maximal paw thickness (2.20 ⫾ 0.2 mm,
2.41 ⫾ 0.39 mm, and 2.28 ⫾ 0.13 mm, respectively), nor
the time to onset of CIA (36.4 ⫾ 5.9 days, 35.6 ⫾ 6.7
days, and 35.7 ⫾ 2.7 days, respectively). The percentages
of mice with severe histologic scores (24%, 56%, and
12.5% among the C3-treated, C3 IL-10–treated, and
control mice, respectively) and severe radiologic scores
(17%, 33%, and 0%, respectively) were even enhanced
in the treated groups. Thus, no beneficial effect on the
disease could be observed when MSCs were engineered
to express IL-10.
Worsening of RA symptoms by intraarticular
MSC injection, without homing of MSCs in the joints.
Previous reports have suggested that infiltration of
Figure 2. Effects of engineered C3 mesenchymal stem cells (MSCs)
expressing interleukin-10 (IL-10) on collagen-induced arthritis (CIA).
A, Immunosuppressive effect of naive compared with IL-10–expressing
C3 MSCs (C3-IL10). The proliferative activity of responder splenocytes in a mixed lymphocyte reaction (MLR) using allogeneic splenocytes (allo) is attributed a 100% value. C3 or C3-IL10 MSCs were
added to the MLR at different cell concentrations. Bars show the
mean and SD. B, Mean increase in paw swelling (in mm) in the groups
of CIA mice receiving naive or IL-10–engineered C3 MSCs on the day
of type II collagen immunization, compared with the group of CIA
control mice. No statistically significant differences between groups
were recorded. The results are representative of 1 of 2 experiments
with the number of mice between 6 and 7 in each group.
MSCs from bone marrow into the synovium precedes
the inflammatory cell accumulation and clinical onset of
arthritis (22,23). We thus wondered whether MSCs
injected either systemically or at ectopic sites could
migrate to the arthritic joints. To this aim, we developed
stable clones of C3 cells expressing the firefly luciferase
gene. One clone, C3 Luc-4, was selected according to its
high luminescence intensity. The sensitivity of detection
of C3 Luc-4 cells in vitro was ⬃103 cells, corresponding
to 9.7 relative luciferase units. We then checked that the
transfection procedure did not interfere with their immunosuppressive properties, and confirmed that C3
Luc-4 cells were able to inhibit T cell proliferation in the
same range as naive C3 cells (results not shown).
We then investigated whether MSCs will migrate
to the arthritic joints when administered at the onset of
CIA (day 32), and whether the route of cell administration can influence the homing of MSCs. We injected the
C3 Luc-4 cells either intravenously, intraperitoneally,
intramuscularly, or intraarticularly in both naive mice
and CIA mice. Arthritis, as measured by the extent of
paw swelling (results not shown) and the percentage of
mice with severe radiologic and histologic scores (Figure
3B), developed in all of these animals regardless of the
route of cell administration. However, the clinical score
was higher in mice receiving the C3 Luc-4 cells by
intraarticular injection (Figure 3A).
Detection of the luminescent cells was performed
in cellular extracts from patellae pouches obtained when
the mice were killed, but no signal could be observed,
suggesting that cells, if not absent, were at least undetectable by this method (⬍0.1% of injected cells). In
contrast, cells extracted from the tissues were detected
among all samples tested but at various levels; they were
barely detectable in the heart, liver, marrow, and kidney
(results not shown) but were frequently recovered from
the muscle, lung, spleen, and brain (Figures 3C and D).
Tissue distribution seemed to be unaffected by the route
of administration but was related to the status of the
mice. Thus, the muscle was the main target in arthritic
mice, but high luciferase activity was detected in the lung
after intravenous injection, which was expected after
systemic infusion (Figure 3C). The lung and the muscle
were the 2 main targets in naive mice (Figure 3D), with
a high number of cells recovered in the muscle after
intramuscular cell injection.
Involvement of TNF␣ in the loss of the immunosuppressive properties of MSCs. To understand why
MSCs do not display any immunosuppressive effects in
CIA, we tested the potential role of the inflammatory
environment on MSC behavior. We therefore investigated the role of various cytokines (TNF␣, IFN␥, IL-1␤)
and lipopolysaccharide (LPS) on the antiproliferative
effect induced by MSCs. No effect was observed when
IFN␥ and IL-1␤ were added to the MLR, and only a
slight effect was observed with LPS (results not shown).
In contrast, although C3 MSCs totally inhibited the
allogeneic response of T lymphocytes in MLRs, they
were unable to suppress the proliferative response when
Figure 3. Assessment of MSC migration to the joints when administered at the onset of CIA (day 32), and
effects of route of cell administration. A, Clinical score recorded during the onset of CIA in mice receiving 106
C3 Luc-4 MSCs through various routes of injection. The naive group comprised the pool of nonimmunized mice
that received 106 C3 Luc-4 MSCs through various routes of injection on day 32. Arrow indicates the day of MSC
injection at the beginning of CIA (day 32). B, Radiologic score expressed as the percentage of mice injected with
C3 Luc-4 MSCs and displaying severe scores (score ⬎1), and histologic score expressed as the percentage of mice
displaying severe scores (score ⱖ2). Bars show the mean and SD. C and D, Detection of the luciferase activity
in various tissues of naive (C) or CIA (D) mice injected with C3 Luc-4 MSCs through various routes. Data are
represented as the relative luciferase units/gm of tissue. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.001. IA ⫽
intraarticular; IM ⫽ intramuscular; IV ⫽ intravenous; IP ⫽ intraperitoneal (see Figure 2 for other definitions).
cultured in the presence of TNF␣, since no statistically
significant difference was observed compared with the
allogeneic response (Figure 4A). TNF␣ alone had no
effect on the allogeneic reaction (results not shown),
whereas the reversion of immunosuppression was observed when TNF␣ was added at 50 or 500 ng/ml.
Moreover, T cell proliferation was statistically higher
than that observed when MSCs were present alone.
We then determined the expression profile of
cytokines secreted by MSCs in the absence or presence
of TNF␣. No variations were observed for most of the
cytokines tested, since low levels of IL-1␤ were measured (Figure 4B) and IL-2, IL-4, IL-12, and IFN␥ were
absent (results not shown). However, the concentrations
of IL-6 were greatly enhanced in the presence of TNF␣
(up to 16-fold increase with the highest dose of TNF␣)
(Figure 4B). Thus, the reversion of MSC-induced immunosuppression and the high levels of IL-6 in the presence
of TNF␣ might account for the absence of a beneficial
effect of MSCs in CIA.
The immunosuppressive properties of MSCs
have been established (4–6) and, recently, treatment of
GVHD with MSCs has been reported (8), suggesting
that MSCs have therapeutic potential to inhibit the
unwarranted host immune response. In the present
study, we investigated whether MSCs might be of therapeutic value in autoimmune inflammatory disorders.
Figure 4. Reversion of the immunosuppressive effect of C3 MSCs in
the presence of tumor necrosis factor ␣ (TNF␣). A, The proliferative
activity of responder splenocytes in an MLR using allogeneic splenocytes (allo) is attributed a 100% value. C3 MSCs were added to the
MLR with or without TNF␣ at the concentrations of 50 ng/ml (TNF50)
or 500 ng/ml (TNF500). ⴱⴱⴱ ⫽ P ⬍ 0.001. B, Levels of various
cytokines in the supernatants of C3 MSCs incubated for 24 hours in the
presence or absence of 50 or 500 ng/ml of TNF␣ (TNF50 or TNF500,
respectively). Bars show the mean and SD. See Figure 2 for other
We showed that MSCs provide no benefit in the CIA
model of arthritis. Indeed, we observed a switch in the
behavior of MSCs depending on the inflammatory environment, alloreactivity or autoimmunity, and we showed
in vitro that TNF␣ is responsible for the reversion of the
immunosuppressive effect of MSCs, possibly accounting
for the lack of amelioration of RA.
CIA is a well-established model for RA that has
been used by many investigators to test the effects of
various treatments aimed at inhibiting the T cell response (14). We used C3 cells because we have previ-
ously shown that they share the immunosuppressive
characteristics of primary MSCs (6). Unexpectedly, in
the CIA model, no beneficial effect on arthritis was
observed at the doses and times of MSC injection tested.
Together with the clinical findings, the immunologic
parameters tended to reverse, further suggesting that
instead of being inhibited, the T cell response was
activated. This activation was unlikely the result of CII
expression by MSCs that could thus act as a boost,
because we have previously demonstrated that naive C3
cells were unable to differentiate in vivo into CIIexpressing chondrocytes (24). These findings suggest
that environmental parameters might influence the
properties of MSCs, since inflammation reverses immunosuppression.
Among the cytokines produced by macrophages
and fibroblasts, TNF␣ has been found to play a pivotal
role in RA (12). We thus investigated whether TNF␣
and/or the subsequent cytokine cascade could influence
the immunosuppressive properties of MSCs. We found
that in the presence of TNF␣, MSCs were unable, at
least in vitro, to inhibit the proliferation of allogeneic T
cells, and this was associated with an increase in the level
of IL-6. In culture supernatants, IL-6 was increased at
least 12–16-fold in the presence of TNF␣ and, although
detected in patellae pouches, was barely detectable
(between 30 and 150 pg/ml) in the sera of CIA mice.
Low levels of systemic IL-6 have already been reported
in the CIA model (25). In our culture conditions, C3
cells secreted basal levels of IL-1␤, whereas no IL-2,
IL-4, or IFN␥ was detectable, which confirms previous
data (26). These cytokines were not enhanced upon
addition of TNF␣ (results not shown). On the contrary,
IL-12 was undetectable with or without TNF␣ induction,
although it has been shown to be expressed by primary
human MSCs (26). Whether the increase in IL-6 was
sufficient to account for the aggravation of arthritis or
whether this cytokine may influence the properties of
MSCs still needs to be investigated. Detection of high
levels of IL-6 both in RA patients and in a mouse model
of arthritis has been reported (27,28), and IL-6 is one of
the cytokines measured as a reflection of arthritis aggravation. These observations suggest that IL-6 plays a role
in the worsening of CIA in the presence of MSCs.
Human MSCs have been isolated from synovial
membrane (29) and shown to maintain their multilineage differentiation potential in vitro (30) and in vivo in
a model of skeletal muscle repair (31). MSCs have also
been identified in the synovial fluid of patients with
arthritis (32) and in articular cartilage (33). These data
suggest that MSCs are present both in normal joints and
arthritic joints. Other reports even have suggested that
in the presence of TNF, marrow-derived MSCs could
accumulate in the synovial membrane and bone marrow
and initiate the clinical onset of arthritis (23). We
investigated whether exogenously added MSCs could
influence the clinical course of arthritis by migrating to
the joints and participating in pannus formation. Using
luciferase-expressing MSCs, enzymatic activity has been
detected in all organs tested, with variations depending
on tissue and individual variability, except in the knee
joints. It is likely that the sensitivity limit of ⬃1,000 cells
in the tissue extract is too high to permit the detection of
the injected cells. Nevertheless, clinical, radiologic, and
histologic parameters were significantly worsened when
cells were injected intraarticularly. This suggests that
although MSCs were not detected in the patellae
pouches, part of the cells may still have contributed to
the aggravation of CIA symptoms.
Although we could not detect luciferase-positive
MSCs in the knee joints, we wanted to determine
whether the homing capacities of MSCs varied depending on the route of injection and the status of mice
(arthritic versus naive). In CIA mice, principally targeted tissues were the muscle, lung, spleen, and brain,
with the muscle being the first targeted tissue except in
the case of intravenous injection, and, to a lesser extent,
the marrow, heart, kidney, and liver were targeted. No
striking difference was observed in comparison with
naive mice, although the lung and muscle were the 2
tissues mainly targeted in this latter group. Similar
findings were previously obtained using murine bone
marrow adherent cells injected in irradiated mice
(34,35). A recent study in a baboon model demonstrated
a less abundant engraftment of ex vivo–expanded MSCs
when the recipients were not previously conditioned by
lethal total body irradiation (36). Thus, depletion of the
hematopoietic compartment by irradiation may help cell
engraftment but does not seem to modify the biodistribution. Similarly, when comparing our experiments with
those of Pereira et al (34,35), distribution of infused cells
appears to be unmodified in naive and irradiated mice.
However, a more precise quantification of infused cells
in the various tissues needs to be performed. Our data
further suggest that MSCs could distribute in a similar
manner and in many tissues, following local or systemic
infusion, and this could occur independent of the inflammatory context.
IL-10 is an antiinflammatory cytokine whose
beneficial role has been shown in CIA (16,19,20). Although IL-10–expressing MSCs have a greater in vitro
immunosuppressive potential, when injected into mice,
they behave similarly to naive MSCs, since both forms
were ineffective in reducing the clinical parameters of
CIA. Secretion of IL-10 by the genetically modified
MSCs was undetectable in the sera of the injected mice
(results not shown). It was previously described that
using recombinant adenoviral gene transfer, a significant
benefit was associated with a level of 30 ng/ml IL-10 in
the animal sera, whereas levels of ⬃600 pg/ml had only
a minimal effect (16). Although we cannot compare the
delivery systems used in both experiments, one may still
speculate that the dose of IL-10 necessary to display a
beneficial effect has to be in the range of nanograms per
milliliter in the bloodstream. Nevertheless, we cannot
exclude the possibility that the local delivery of IL-10–
engineered MSCs directly into the joints might have a
therapeutic effect. Use of human MSCs expressing the
soluble TNF receptor II has recently been reported in
the SCID model of arthritis induced by anti-CII antibodies, and investigators reported a reduction of arthritic symptoms (Barry F: personal communication).
Thus, by blocking TNF␣, this therapeutic molecule not
only prevents the cytokine cascade responsible for cell
proliferation and degradation in the joints, but also
possibly preserves the immunosuppressive effect of
Our results thus suggest that although nonengineered MSCs seem to be unsuitable for the treatment of
inflammatory diseases, use of engineered MSCs could
be of interest to effectively deliver a therapeutic molecule. These data need to be confirmed in other autoimmune disease models.
We thank A. Pillon and P. Ballaguer (INSERM U540,
Montpellier, France) for the kind gift of the pCMV-Luc⫹-Neo
vector, and C. Verwaerde (Institut Pasteur, Lille, France) for
the generous gift of the pREP10-vIL-10 vector. We are grateful to Denis Greuet for providing excellent animal care,
Michèle Radal (Centre de Recherches et de Lutte contre le
Cancer Val d’Aurelle, Montpellier, France) for performing the
histologic work, and Dr. Taourel and colleagues at Lapeyronie
Hospital in Montpellier for carrying out the radiographic
1. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R,
Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–7.
2. Barry FP, Murphy JM. Mesenchymal stem cells: clinical applications and biological characterization. Int J Biochem Cell Biol
3. Noel D, Djouad F, Jorgensen C. Regenerative medicine through
mesenchymal stem cells for bone and cartilage repair. Curr Opin
Investig Drugs 2002;3:1000–4.
Bartholomew A, Sturgeon C, Siatskas M, Ferrer K, McIntosh K,
Patil S, et al. Mesenchymal stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp
Hematol 2002;30:42–8.
Di Nicola M, Carlo-Stella C, Magni M, Milanesi M, Longoni PD,
Matteucci P, et al. Human bone marrow stromal cells suppress
T-lymphocyte proliferation induced by cellular or nonspecific
mitogenic stimuli. Blood 2002;99:3838–43.
Djouad F, Plence P, Bony C, Tropel P, Apparailly F, Sany J, et al.
Immunosuppressive effect of mesenchymal stem cells favors tumor
growth in allogeneic animals. Blood 2003;102:3837–44.
Bartholomew A, Patil S, Mackay A, Nelson M, Buyaner D, Hardy
W, et al. Baboon mesenchymal stem cells can be genetically
modified to secrete human erythropoietin in vivo. Hum Gene Ther
Le Blanc K, Rasmusson I, Sundberg B, Gotherstrom C, Hassan M,
Uzunel M, et al. Treatment of severe acute graft-versus-host
disease with third party haploidentical mesenchymal stem cells.
Lancet 2004;363:1439–41.
Jorgensen C, Djouad F, Apparailly F, Noel D. Engineering
mesenchymal stem cells for immunotherapy. Gene Ther 2003;10:
Ishida T, Inaba M, Hisha H, Sugiura K, Adachi Y, Nagata N, et al.
Requirement of donor-derived stromal cells in the bone marrow
for successful allogeneic bone marrow transplantation: complete
prevention of recurrence of autoimmune diseases in MRL/MPIpr/Ipr mice by transplantation of bone marrow plus bones (stromal cells) from the same donor. J Immunol 1994;152:3119–27.
Firestein GS. Evolving concepts of rheumatoid arthritis. Nature
Toussirot E, Wendling D. The use of TNF-␣ blocking agents in
rheumatoid arthritis: an overview. Expert Opin Pharmacother
Lorenz HM. T-cell-activation inhibitors in rheumatoid arthritis.
BioDrugs 2003;17:263–70.
Robbins PD, Evans CH, Chernajovsky Y. Gene therapy for
arthritis. Gene Ther 2003;10:902–11.
Pillon A, Gauthier P, Balaguer P, Pelegrin A, Nicolas JC. Bioluminescent imaging: applications to cancerology and endocrinology. J Soc Biol 2004;198:157–61.
Apparailly F, Verwaerde C, Jacquet C, Auriault C, Sany J,
Jorgensen C. Adenovirus-mediated transfer of viral IL-10 gene
inhibits murine collagen-induced arthritis. J Immunol 1998;160:
Delgado M, Abad C, Martinez C, Leceta J, Gomariz RP. Vasoactive intestinal peptide prevents experimental arthritis by downregulating both autoimmune and inflammatory components of the
disease. Nat Med 2001;7:563–8.
McInnes IB, Illei GG, Danning CL, Yarboro CH, Crane M,
Kuroiwa T, et al. IL-10 improves skin disease and modulates
endothelial activation and leukocyte effector function in patients
with psoriatic arthritis. J Immunol 2001;167:4075–82.
Kim KN, Watanabe S, Ma Y, Thornton S, Giannini EH, Hirsch R.
Viral IL-10 and soluble TNF receptor act synergistically to inhibit
collagen-induced arthritis following adenovirus-mediated gene
transfer. J Immunol 2000;164:1576–81.
Jorgensen C, Apparailly F, Canovas F, Verwaerde C, Auriault C,
Jacquet C, et al. Systemic viral interleukin-10 gene delivery
prevents cartilage invasion by human rheumatoid synovial tissue
engrafted in SCID mice. Arthritis Rheum 1999;42:678–85.
21. Jorgensen C, Apparailly F, Couret I, Canovas F, Jacquet C, Sany
J. Interleukin-4 and interleukin-10 are chondroprotective and
decrease mononuclear cell recruitment in human rheumatoid
synovium in vivo. Immunology 1998;93:518–23.
22. Marinova-Mutafchieva L, Taylor P, Funa K, Maini RN, Zvaifler
NJ. Mesenchymal cells expressing bone morphogenetic protein
receptors are present in the rheumatoid arthritis joint. Arthritis
Rheum 2000;43:2046–55.
23. Marinova-Mutafchieva L, Williams RO, Funa K, Maini RN,
Zvaifler NJ. Inflammation is preceded by tumor necrosis
factor–dependent infiltration of mesenchymal cells in experimental arthritis. Arthritis Rheum 2002;46:507–13.
24. Noel D, Gazit D, Bouquet C, Apparailly F, Bony C, Plence P, et al.
Short-term BMP-2 expression is sufficient for in vivo osteochondral differentiation of mesenchymal stem cells. Stem Cells 2004;
25. Takagi N, Mihara M, Moriya Y, Nishimoto N, Yoshizaki K,
Kishimoto T, et al. Blockage of interleukin-6 receptor ameliorates
joint disease in murine collagen-induced arthritis. Arthritis Rheum
26. Deans RJ, Moseley AB. Mesenchymal stem cells: biology and
potential clinical uses. Exp Hematol 2000;28:875–84.
27. Kraan MC, Smeets TJ, van Loon MJ, Breedveld FC, Dijkmans
BA, Tak PP. Differential effects of leflunomide and methotrexate
on cytokine production in rheumatoid arthritis. Ann Rheum Dis
28. Kawabata D, Tanaka M, Fujii T, Umehara H, Fujita Y, Yoshifuji
H, et al. Ameliorative effects of follistatin-related protein/TSC-36/
FSTL1 on joint inflammation in a mouse model of arthritis.
Arthritis Rheum 2004;50:660–8.
29. Nishimura K, Solchaga LA, Caplan AI, Yoo JU, Goldberg VM,
Johnstone B. Chondroprogenitor cells of synovial tissue. Arthritis
Rheum 1999;42:2631–7.
30. De Bari C, Dell’Accio F, Tylzanowski P, Luyten FP. Multipotent
mesenchymal stem cells from adult human synovial membrane.
Arthritis Rheum 2001;44:1928–42.
31. De Bari C, Dell’Accio F, Vandenabeele F, Vermeesch JR, Raymackers JM, Luyten FP. Skeletal muscle repair by adult human
mesenchymal stem cells from synovial membrane. J Cell Biol
32. Jones EA, English A, Henshaw K, Kinsey SE, Markham AF,
Emery P, et al. Enumeration and phenotypic characterization of
synovial fluid multipotential mesenchymal progenitor cells in
inflammatory and degenerative arthritis. Arthritis Rheum 2004;50:
33. Alsalameh S, Amin R, Gemba T, Lotz M. Identification of
mesenchymal progenitor cells in normal and osteoarthritic human
articular cartilage. Arthritis Rheum 2004;50:1522–32.
34. Pereira RF, Halford KW, O’Hara MD, Leeper DB, Sokolov BP,
Pollard MD, et al. Cultured adherent cells from marrow can serve
as long-lasting precursor cells for bone, cartilage, and lung in
irradiated mice. Proc Natl Acad Sci U S A 1995;92:4857–61.
35. Pereira RF, O’Hara MD, Laptev AV, Halford KW, Pollard MD,
Class R, et al. Marrow stromal cells as a source of progenitor cells
for nonhematopoietic tissues in transgenic mice with a phenotype
of osteogenesis imperfecta. Proc Natl Acad Sci U S A 1998;95:
36. Devine SM, Cobbs C, Jennings M, Bartholomew A, Hoffman R.
Mesenchymal stem cells distribute to a wide range of tissues
following systemic infusion into nonhuman primates. Blood 2003;
Без категории
Размер файла
120 Кб
stem, necrosis, factors, properties, induced, reversal, mesenchymal, immunosuppressive, arthritis, collagen, tumors, cells
Пожаловаться на содержимое документа