SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

Rheumatoid arthritis is characterized by persistent synovial inflammation and progressive joint destruction, which are mediated by innate and adaptive immune responses. Cytokine blockade successfully treats some patient subsets; however, ∼50% do not respond to this approach. Targeting of pathogenic T lymphocytes is emerging as an effective alternative/complementary therapeutic strategy, yet the factors that control T cell activation in joint disease are not well understood. Tenascin-C is an arthritogenic extracellular matrix glycoprotein that is not expressed in healthy synovium but is elevated in the rheumatoid joint, where high levels are produced by myeloid cells. Among these cells, tenascin-C expression is most highly induced in activated dendritic cells (DCs). The aim of this study was to examine the role of tenascin-C in this cell type.

Methods

We systematically compared the phenotype of DCs isolated from wild-type mice or mice with a targeted deletion of tenascin-C by assessing cell maturation, cytokine synthesis, and T cell polarization.

Results

Dendritic cells derived from tenascin-C–null mice exhibited no defects in maturation; induction of the class II major histocompatibility complex and the costimulatory molecules CD40 and CD86 was unimpaired. Dendritic cells that did not express tenascin-C, however, produced lower levels of inflammatory cytokines than did cells from wild-type mice and exhibited specific defects in Th17 cell polarization. Moreover, tenascin-C–null mice displayed ablated levels of interleukin-17 in the joint during experimental arthritis.

Conclusion

These data demonstrate that tenascin-C is important in DC-mediated polarization of Th17 lymphocytes during inflammation and suggest a key role for this endogenous danger signal in driving adaptive immunity in erosive joint disease.

Rheumatoid arthritis (RA) occurs in 1% of the population worldwide and is associated with significant morbidity and increased mortality rates. The pivotal role of innate immunity in disease pathogenesis is highlighted by the success of proinflammatory cytokine blockade, including tumor necrosis factor α (TNFα), interleukin-1 (IL-1), and IL-6, which has significantly improved patient care. However, a relatively large subset of RA patients do not respond to these drugs, and those who initially respond well can become refractory over time (1).

In addition to dysregulated cytokine production, the RA joint also exhibits persistent infiltration of T lymphocytes, including Th1 and Th17 cells (2), along with impaired Treg cells (3). The contribution of adaptive immunity to RA has recently been clarified by the success of targeting CTLA-4–mediated T cell activation clinically and by the development of new T cell–directed therapies, exemplified most notably by the efficacy of IL-17 inhibition in suppressing RA symptoms (4). These approaches may offer valuable alternatives to patients who do not benefit from cytokine-based therapies. However, the molecular mechanisms that drive T cell activation during RA are not well understood.

T cell activation relies on the professional antigen-presenting capabilities of cells such as dendritic cells (DCs). Upon encountering immunologic danger, such as pathogenic invasion or tissue injury, the resultant DC maturation and migration to lymphoid tissues enables the activation of naive T cells by means of 3 distinct signals: antigen presentation, costimulatory molecule engagement, and release of a cocktail of polarizing cytokines (5). Mature DCs are present in the RA synovium in ectopic lymphoid structures (6, 7). However, it is not clear which factors activate DCs during RA, causing them to persistently polarize pathogenic T cells.

DC activation is mediated by their recognition of pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs), which are sensed by pattern-recognition receptors such as the Toll-like receptors (TLRs) (5). High levels of DAMPs, including intracellular molecules released by necrotic cells and extracellular matrix (ECM) molecules that are specifically up-regulated upon injury or degraded following tissue damage, are found in the destructive milieu of the RA joint (8, 9). By virtue of their ability to activate DCs, DAMPs form an important link between tissue injury and adaptive immunity. Persistently elevated levels of DAMPs in the RA joint may also stimulate aberrant T cell activation during RA, driving a chronic cycle of inflammation that mediates progressive tissue damage.

We have identified the ECM glycoprotein tenascin-C as an endogenous activator of inflammation (10). Its expression is specifically and transiently induced upon tissue injury, while persistent expression is associated with chronic inflammation (11). Little or no tenascin-C is detected in healthy synovium; however, the expression of this DAMP is significantly elevated in RA (12, 13), where levels correlate with the levels of inflammatory mediators (12). We demonstrated that tenascin-C is arthritogenic in vivo and that its expression is essential for persistent synovial inflammation and tissue destruction in mouse models of arthritis (10). We also assessed tenascin-C expression in immune myeloid cells and found an extremely high level of induction in activated DCs (12). The role of tenascin-C in DC function is not known; thus, we aimed to determine whether tenascin-C contributes to joint inflammation by promoting an adaptive immune response. Our data demonstrate that tenascin-C activates cytokine synthesis and T cell differentiation and drives pathogenic effector production during erosive joint disease.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Mice.

Mice deficient in tenascin-C (TN-C−/−) and their wild-type (TN-C+/+) littermates from heterozygous breeding pairs on the 129SvJ background were used for these experiments (14). All mice were female and were ages 8–12 weeks at the time of experimentation. All procedures were approved by the institutional ethics committee and the UK Home Office.

Generation of bone marrow–derived dendritic cells (BMDCs).

We generated BMDCs by flushing bone marrow from the tibias and femurs of TN-C+/+ and TN-C−/− mice. Red blood cells were depleted using red blood cell lysing buffer (Sigma). Live cells (5 × 106) were seeded in tissue culture plates in RPMI 1640 supplemented with 20 mM HEPES and L-glutamine (PAA), 10% fetal calf serum (Gibco), 1% antibiotic/antimycotic solution (PAA), and 50 μM β-mercaptoethanol (Invitrogen) containing 20 ng/ml of recombinant murine granulocyte–macrophage colony-stimulating factor (GM-CSF) (PeproTech), with or without 10 ng/ml of recombinant murine IL-4 (PeproTech) as indicated below. On day 3, fresh media containing 20 ng/ml of GM-CSF was added. Nonadherent and loosely adherent cells were harvested on day 7.

To induce BMDC maturation, 2 μg/ml of lipopolysaccharide (LPS; Enzo Life Sciences) was added on day 6 of the culture period, as described previously (15). Results of cell stimulation with 100 ng/ml, 10 ng/ml, and 1 ng/ml of LPS were not significantly different from the results of cell stimulation with 2 μg/ml. No effects on cell viability, which was assessed using the MTT assay (Sigma), were observed.

Fluorescence-activated cell sorter (FACS) analysis.

BMDCs were stained for 20 minutes at 4°C with Pacific Blue–conjugated anti-CD11c, allophycocyanin (APC)–conjugated anti-CD11b, phycoerythrin (PE)–conjugated anti–Ly-6G, fluorescein-isothiocyanate (FITC)–conjugated anti-CD3ε, FITC-conjugated anti-CD86, APC-conjugated anti-CD40, FITC-conjugated anti-CD45R (B220), FITC-conjugated anti–TLR-4/MD2 (all from eBioscience); PE-conjugated anti–class II MHC (I-A/I-E), FITC-conjugated anti–Ly-6C (both from BD Biosciences), PE-conjugated anti-F4/80 (Caltag MedSystems), and appropriate isotype controls in FACS buffer (2% bovine serum albumin [BSA], 2 mM EDTA, 0.02% NaN3, and phosphate buffered saline [PBS]). Cells were washed twice with FACS buffer and then fixed for 20 minutes at 4°C with 4% paraformaldehyde in PBS, followed by 2 washes with FACS buffer. Cells were analyzed using a FACSCalibur flow cytometer with BD FACSDiva software (both from BD Biosciences). Postacquisition data analysis was performed using FlowJo software version 7.6.1 (Tree Star).

Forward scatter and side scatter analyses were used to exclude dead cells and debris. No difference in cell size and granularity was observed between TN-C+/+ mice and TN-C−/− mice. Unstained controls to detect autofluorescence of the cells and isotype controls to assess nonspecific binding of antibodies were included in every experiment. No background staining was observed using the isotype controls, and so those data, rather than the data from the unstained samples, are illustrated below. The percentage of positive cells was determined, and the geometric mean fluorescence intensity (MFI) of positive cells was used as a semiquantitative measure of expression levels.

Assessment of BMDC cytokine synthesis.

LPS stimulation of BMDCs.

Live BMDCs (5 × 104) were plated in triplicate in 96-well plates (Becton Dickinson Labware) and were left untreated or were treated with 1 ng/ml of LPS for 4, 8, or 24 hours. Supernatants were collected and assayed for cytokine levels by Luminex analysis, as described below.

Luminex analysis.

Levels of TNFα, IL-6, keratinocyte-derived chemokine (KC), interferon-γ–inducible 10-kd protein (IP-10), and IL-10 in supernatants were analyzed using a Luminex multiplex bead-based assay. Capture antibodies (R&D Systems or PeproTech) were coupled to individual Luminex bead sets using an amine-coupling kit (both from Bio-Rad) according to the manufacturer's instructions. Bead sets were mixed with samples or standards (PeproTech), and plates were agitated overnight at 4°C in the dark. After washing, secondary antibodies (PeproTech) were incubated, with agitation, for 1 hour at room temperature in the dark. Streptavidin–PE (Europa Bioproducts) diluted in PBS plus 1% BSA was added to the wells, with agitation, for 30 minutes at room temperature in the dark. Beads resuspended in PBS were analyzed in a Luminex 100 analyzer using StarStation software (Applied Cytometry).

Assessment of BMDC-induced T cell differentiation.

BMDC and T cell coculture.

BMDC-induced T cell polarization was assessed as described previously (16). Single-cell suspensions were prepared from the spleens and lymph nodes of TN-C+/+ mice. After red blood cell depletion, CD4+ T cells were purified by positive selection using magnetic beads (Miltenyi Biotec) according to the manufacturer's protocol. BMDCs from TN-C+/+ and TN-C−/− mice were harvested on day 7 and plated in triplicate in 96-well U-bottomed plates (Becton Dickinson Labware) and stimulated with 10 ng/ml of LPS or with 50 μg/ml of Freund's complete adjuvant (CFA) containing Mycobacterium tuberculosis (Sigma). After 24 hours, the medium was replaced with medium containing freshly isolated CD4+ T cells with 100 ng/ml of anti-CD3 antibody (eBioscience). BMDCs and CD4+ T cells were cocultured at a 1:5 ratio for 5 days, and then the cytokine levels in the supernatants were determined by enzyme-linked immunosorbent assay (ELISA).

ELISA.

Antibodies and standards for ELISA, as well as streptavidin–horseradish peroxidase, were purchased from BD Biosciences and were used according to the manufacturer's instructions. Absorbance was read at 450 nm on a FluoStar Omega spectrophotometric ELISA plate reader, and data were analyzed using Omega software (both from BMG Labtech).

T cell proliferation.

To assess T cell proliferation, cocultures of BMDCs and T cells were established as above, except that the cells were cocultured for 72 hours before pulsing for 18 hours with 1 μCi of 3H-thymidine (PerkinElmer). 3H-thymidine incorporation was evaluated using a liquid scintillation counter (MicroBeta Jet; PerkinElmer).

Generation of the antigen-induced arthritis (AIA) mouse model.

AIA was induced in mice as previously described (10). Briefly, on day –7, the mice were sedated and immunized with 100 μg of methylated BSA (mBSA; Sigma) emulsified in 0.1 ml of CFA (Sigma) by subcutaneous injection at the base of the tail. On day 0, arthritis was induced by intraarticular injection of mBSA (200 μg in 10 μl of sterile PBS) into the right knee joint. Control mice were injected with 10 μl of PBS alone or were not injected. At 6 hours and at 1, 3, 5, and 7 days following intraarticular injection of mBSA, the mice were euthanized, and the knee joints were excised.

Real-time quantitative polymerase chain reaction (qPCR) analysis.

Total RNA was isolated from knee joints using an RNeasy Mini kit (Qiagen), and 500 ng of RNA was reverse transcribed using SuperScript III reverse transcriptase and oligo(dT)12–18 primers (Invitrogen). Expression of messenger RNA (mRNA) for IL-17A, interferon-γ (IFNγ), IL-6, IL-23, IL-23R, IFN regulatory factor 4 (IRF-4), runt-related transcription factor 1 (RUNX-1), and basic leucine zipper transcription factor ATF-like (B-ATF) was analyzed by real-time qPCR using TaqMan Gene Expression Master Mix and TaqMan probes and primers (gene expression assays from Applied Biosystems): for IL-17A, Mm00439619_m1; for IFNγ, Mm00801778_m; for IL-6, Mm00446190_m1; for IL-23p19, Mm00518984_m1; for IL-23R, Mm00519943; for IRF-4, Mm0051643_m1; for RUNX-1, Mm01213405_m1, and for B-ATF, Mm00479410_m1. Relative gene expression was analyzed using the ΔΔCt method (17). Hypoxanthine guanine phosphoribosyltransferase 1 (TaqMan gene expression assay Mm00446968_m1; Applied Biosystems) was used as an endogenous control gene, and the right knee joint from untreated mice was used as the calibrator. Real-time qPCR was performed in a Rotor-Gene 6000 instrument (Corbett Research).

Statistical analysis.

The mean ± SEM and statistical significance were calculated using GraphPad Prism 5 software. Multiple group means were analyzed by two-way analysis of variance, followed by the Bonferroni posttest where appropriate. Student's unpaired 2-tailed t-test was used for experiments involving only 2 groups and Student's paired 2-tailed t-test was used to compare 2 paired groups. P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Modest effect of tenascin-C on BMDC differentiation.

The number of DCs differentiated from bone marrow–derived myeloid precursor cells obtained from TN-C+/+ and TN-C−/− mice was determined by analysis of cell surface CD11c, which is widely used as a murine DC marker (18). FACS analysis revealed a single population of CD11c+ cells (Figure 1A). From TN-C+/+ mice, the percentage of cells expressing CD11c was routinely 91.08 ± 1.97% (mean ± SEM). A modestly, but significantly, reduced number of CD11c+ cells (85.52 ± 3.18%) was obtained from TN-C−/− mice (Figure 1B) that exhibited a slightly, but significantly lower MFI (mean ± SEM 1,913 ± 261.4) than did the CD11c+ cells from TN-C+/+ mice (2,357 ± 296.4) (Figure 1C). These data suggest that BMDC differentiation is slightly impaired in TN-C−/− mice.

thumbnail image

Figure 1. Modest defect in bone marrow–derived dendritic cell (BMDC) differentiation in tenascin-C–null mice. BMDCs from TN-C+/+ mice and TN-C−/− mice were stained with anti-CD11c (A–C) or with anti-CD11b (D–F) antibodies. A, Representative histograms of CD11c+ cells from the two groups. Red histograms show isotype control; blue histograms show anti–CD11c staining. B, Percentage of CD11c+ cells in the two groups. C, Geometric mean fluorescence intensity (MFI) of CD11c expression in the two groups. D, Representative histograms of CD11b+ cells from the two groups. Red histograms show isotype control; blue histograms show anti–CD11b staining. E, Percentage of CD11b+ cells in the two groups. F, MFI of CD11b expression in the two groups. Values are the mean ± SEM of 8 TN-C+/+ mice and 6 TN-C−/− mice in B and C and of 3 mice per group in E and F. = P < 0.05; ∗∗ = P < 0.01. NS = not significant.

Download figure to PowerPoint

Cells were also stained for CD11b, which is widely used as a myeloid cell marker (19). The majority of cells expressed CD11b at high levels (Figure 1D). TN-C−/− mice generated significantly less CD11b+ cells (mean ± SEM 89.90 ± 0.95%) than did the TN-C+/+ mice (95.50 ± 0.92%) (Figure 1E), although the MFI of CD11b cells was not significantly different between genotypes (Figure 1F). These findings confirm that lower numbers of myeloid cells are generated in TN-C−/− mice.

Effective maturation of BMDCs in the absence of tenascin-C.

Murine BMDCs generated by GM-CSF–driven differentiation of bone marrow–derived myeloid precursor cells are in an immature state. They express low levels of class II MHC molecules and low levels of costimulatory molecules, including CD86 and CD40. Treatment with inflammatory stimuli triggers BMDC maturation, resulting in the up-regulation of cell surface expression of class II MHC and costimulatory molecules (20). Accordingly, in the absence of LPS, class II MHC staining delineated 3 distinct populations in BMDCs from both TN-C+/+ and TN-C−/− mice: one population that was negative for class II MHC (MHCII–), one that expressed intermediate levels of class II MHC (MHCIIintermediate), and one that expressed high levels of class II MHC (MHCIIhigh) (Figure 2A).

thumbnail image

Figure 2. No defect in bone marrow–derived dendritic cell (BMDC) maturation in tenascin-C–null mice. BMDCs were left untreated or were treated with 2 μg/ml of lipopolysaccharide (LPS) for 24 hours. Cells were double-stained with anti-CD11c and anti–class II major histocompatibility complex (anti–class II MHC) antibodies (A–C) or with anti-CD11c and anti-CD40 antibodies (D–F). A, Representative histograms of CD11c+MHCII+ cells from the two groups. Red histograms show isotype control; blue histograms show MHCII staining. B, Percentage of MHCII+ cells in the two groups. C, Geometric mean fluorescence intensity (MFI) of MHCII expression in the CD11c+MHCII+ population from the two groups. D, Representative histograms of CD11c+CD40+ cells from the two groups. Red histograms show isotype control; blue histograms show CD40 staining. E, Percentage of CD40+ cells in the two groups. F, MFI of CD40 expression in the CD11c+CD40+ population from the two groups. Values are the mean ± SEM of 8 TN-C+/+ mice and 6 TN-C−/− mice in the absence of LPS and of 3 mice per group in the presence of LPS in B and C and of 4 TN-C+/+ mice and 5 TN-C−/− mice in the absence of LPS and of 3 mice per group in the presence of LPS in E and F. NS = not significant. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Download figure to PowerPoint

Quantification of MHCII expression by CD11c+ cells revealed no significant differences in the percentages of CD11c+MHCII– and CD11c+MHCII+ cells between TN-C+/+ and TN-C−/− mice (Figure 2B). There were also no significant differences in the MFI of these cell populations from TN-C+/+ mice as compared with TN-C−/− mice (Figure 2C). Upon LPS-induced maturation, staining for class II MHC revealed 2 cell populations: one MHCII– and the other with intermediate to high MHCII expression (Figure 2A). Treatment with LPS did not result in a significant change in the percentage of MHCII+ cells (Figure 2B). However, levels of MHCII expression were markedly up-regulated upon LPS treatment. No difference in the levels of MHCII expression was observed between TN-C+/+ and TN-C−/− mice after LPS stimulation (Figure 2C).

In the absence of LPS, cells did not express CD40, but this maturation marker was induced by stimulation with LPS (Figure 2D). Approximately 70% of CD11c+ BMDCs expressed CD40 after LPS treatment, with no significant differences being observed between cells derived from TN-C+/+ mice and those derived from TN-C−/− mice (Figure 2E). No differences were observed in CD40 expression levels between TN-C+/+ and TN-C−/− mice (Figure 2F).

Untreated cells also exhibited a negative-to-low surface expression of CD86. There was no difference in the percentage of cells expressing CD86 in BMDCs derived from TN-C+/+ and TN-C−/− mice, nor was any difference in CD86 expression levels observed. LPS treatment did not affect the percentage of CD86-expressing CD11c+ BMDCs. However, LPS stimulation profoundly up-regulated CD86 expression levels in mice of both genotypes, although no differences in CD86 expression levels between TN-C+/+ and TN-C−/− mice were observed (data not shown).

Treatment of BMDCs with IL-4 also induced BMDC maturation, and the levels of expression of MHCII, CD86, and CD40 were elevated to an equal extent in BMDCs from TN-C+/+ mice and TN-C−/− mice (Table 1). Taken together, these data demonstrate that tenascin-C does not play a role in BMDC maturation upon LPS or IL-4 stimulation.

Table 1. IL-4–induced maturation of BMDCs from TNC+/+ and TNC−/− mice*
 BMDCs cultured without IL-4BMDCs cultured with IL-4
TNC+/+TNC−/−TNC+/+TNC−/−
  • *

    Interleukin-4 (IL-4)–induced maturation of bone marrow–derived dendritic cells (BMDCs) was analyzed by fluorescence-activated cell sorting. Values are the mean ± SEM. MFI = mean fluorescence intensity.

  • P < 0.05.

  • P < 0.01.

% positive    
 CD11c+ BMDCs91.08 ± 1.9785.52 ± 3.1865.03 ± 11.0264.18 ± 11.63
 BMDC maturation state    
  CD11c+MHCII+83.23 ± 7.3079.78 ± 8.6284.93 ± 6.5287.01 ± 2.08
  CD11c+CD86+84.00 ± 8.1179.05 ± 10.7577.61 ± 7.7875.63 ± 9.49
  CD11c+CD40+0.29 ± 0.280.64 ± 0.4049.20 ± 1.1552.32 ± 5.23
MFI    
 CD11c+ BMDCs2,357 ± 296.41,913 ± 2 61.43,199 ± 712.13,019 ± 464.0
 BMDC maturation state    
  CD11c+MHCII+4,089 ± 664.13,710 ± 809.64,208 ± 18905,174 ± 2458
  CD11c+CD86+571.8 ± 46.71610.8 ± 28.591,693 ± 338.11,879 ± 226.4
  CD11c+CD40+148.7 ± 22.68177.8 ± 26.571,155 ± 75.661,301 ± 132.9

Tenascin-C promotion of BMDC cytokine synthesis.

DCs fulfill 2 key roles: they are an important source of proinflammatory cytokines, and they are the professional antigen-presenting cells of the immune system. To assess the impact of tenascin-C on DC function, we first examined cytokine synthesis upon activation with LPS. Unstimulated BMDCs from TN-C+/+ and TN-C−/− mice did not express TNFα, IL-6, or KC at any time point examined (Figure 3A). LPS treatment induced the expression of TNFα, IL-6, and KC in BMDCs from both TN-C+/+ and TN-C−/− mice, which increased over 24 hours. At all time points, cytokine levels were significantly lower in BMDCs from the TN-C−/− mice than in those from the TN-C+/+ mice (Figure 3A), suggesting that tenascin-C contributes to TNFα, IL-6, and KC production in BMDCs. Only low levels of IL-10 and IP-10 were detected after LPS stimulation, and no consistent differences between TN-C+/+ and TN-C−/− mice were found (data not shown).

thumbnail image

Figure 3. Reduced synthesis of proinflammatory cytokines in tenascin-C–null mice. A, Bone marrow–derived dendritic cells (BMDCs) from TN-C+/+ and TN-C−/− mice were left unstimulated or were stimulated with 1 ng/ml of lipopolysaccharide (LPS) for 4, 8, or 24 hours, and levels of tumor necrosis factor α (TNFα), interleukin-6 (IL-6), and keratinocyte-derived chemokine (KC) were quantified and normalized to cytokine expression in cells from TN-C+/+ mice at 24 hours. Values are the mean ± SEM of 7 mice per group for TNFα and IL-6 and of 11 mice per group for KC. = P < 0.05; ∗∗ = P < 0.01. B–E, BMDCs were double-stained with anti-CD11c and anti–Toll-like receptor 4 (anti–TLR-4) antibodies. Representative histograms (B) and dot plots (C) of data for cells from TN-C+/+ and TN-C−/− mice are shown. In B, the red histograms show isotype control; blue histograms show TLR-4 staining. In C, the numbers in each quadrant are the percentage of positive cells. The percentage of CD11c+TLR-4+ cells (D) and the geometric mean fluorescence intensity (MFI) of TLR-4 expression (E) in the two groups are also shown. Values in D and E are the mean ± SEM of 3 mice per group. NS = not significant. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Download figure to PowerPoint

LPS induces cytokine expression via activation of TLR-4 signaling. Cell surface expression of TLR-4 in BMDCs from TN-C+/+ and TN-C−/− mice was examined by FACS analysis. The majority of cells expressed TLR-4 at intermediate levels (Figures 3B and C). There was no significant difference in the percentages of TLR-4–expressing CD11c+ cells (Figure 3D) or in the expression levels of TLR-4 (Figure 3E) in BMDCs derived from TN-C+/+ mice and TN-C−/− mice. These data indicate that the impaired cytokine synthesis in TN-C−/− mice is not due to lower TLR-4 surface expression.

Th17 cell polarization driven by tenascin-C in T cell coculture assays.

We also examined the efficacy of BMDCs in promoting T cell proliferation and polarization into the specific subclasses Th1, Th2, Th17, and Treg cells, as defined by their ability to produce IFNγ, IL-5, IL-17, and IL-10, respectively. No significant difference was observed in IFNγ, IL-5, and IL-10 levels induced by BMDCs from TN-C+/+ and TN-C−/− mice. However, BMDCs from TN-C−/− mice induced the synthesis of significantly reduced levels of IL-17 as compared to those from TN-C+/+ mice (Figure 4A). Stimulation of BMDCs from TN-C−/− mice with either LPS or M tuberculosis generated significantly decreased levels of Th17 cell polarization (Figure 4A). No significant difference in T cell proliferation was induced by BMDCs from TN-C+/+ and TN-C−/− mice, suggesting that tenascin-C does not play a role in BMDC-induced T cell proliferation (Figure 4B). These data indicate a specific defect in BMDC cell induction of Th17 cells in the absence of tenascin-C.

thumbnail image

Figure 4. Defects in Th17 polarization in tenascin-C–null mice. Bone marrow–derived dendritic cells (BMDCs) from TN-C+/+ and TN-C−/− mice were left unstimulated or were stimulated with 10 ng/ml of lipopolysaccharide (LPS) or with 50 μg/ml of Mycobacterium tuberculosis (Mtb). CD4+ T cells were isolated from spleen and lymph nodes of TN-C+/+ mice and cocultured at a 1:5 ratio with stimulated or unstimulated TN-C+/+ and TN-C−/− mouse BMDCs in the presence of anti-CD3 (100 ng/ml) for 5 days (A) or 72 hours (B). A, Supernatants were collected, and levels of interferon-γ (IFNγ), interleukin-5 (IL-5), IL-10, and IL-17 were analyzed by enzyme-linked immunosorbent assay. TN-C+/+ expression for each cytokine was normalized to 100%. Values are the mean ± SEM of 7 mice per group. = P < 0.05; ∗∗ = P < 0.01. NS = not significant. B, Cells were pulsed for 18 hours with 1 μCi of 3H-thymidine per well. Cells were then harvested, and 3H-thymidine incorporation was measured. Values are the mean ± SEM of 4 mice per group.

Download figure to PowerPoint

IL-17 synthesis driven by tenascin-C during joint inflammation in vivo.

We have previously shown that intraarticular injection of tenascin-C results in joint inflammation and that TN-C−/− mice are protected from sustained and erosive joint inflammation in the AIA model (10). In this model, mice develop an IL-17–dependent chronic erosive monarticular arthritis that closely resembles human RA, including the development of synovial hyperplasia, infiltration of inflammatory cells, and articular bone and cartilage destruction (21). In the present study, we assessed tissue levels of IL-17 during AIA in both TN-C+/+ and TN-C−/− mice. Wild-type mice exhibited a rapid induction of IL-17, peaking 24 hours following intraarticular injection of mBSA; however, little IL-17 was observed in the knee joints of TN-C−/− mice (Figure 5A). In contrast, while IFNγ was induced in the joints of TN-C+/+ mice, peaking at 6 hours and declining over time, IFNγ was also expressed in TN-C−/− mice, but with a delayed induction, peaking at 3 days (Figure 5B). These data suggest that production of IFNγ can occur in the absence of tenascin-C, but that tenascin-C is absolutely required for IL-17 synthesis during joint disease.

thumbnail image

Figure 5. Lower levels of interleukin-17 (IL-17) in tenascin-C–null mice during joint inflammation and destruction in vivo. TN-C+/+ and TN-C−/− mice were immunized with methylated bovine serum albumin (mBSA) emulsified in Freund's complete adjuvant (day –7) and 7 days later (day 0), antigen-induced arthritis (AIA) was generated by intraarticular injection of mBSA into the right knee joint. At 6 hours and at 1, 3, 5, and 7 days after AIA induction, knee joints were excised, tissue was homogenized, and total RNA was isolated. Levels of expression of mRNA for A, IL-17A, B, interferon-γ (IFNγ), C, IL-6, D, IL-23, E, IL-23 receptor (IL-23R), F, IFN regulatory factor 4 (IRF-4), G, B basic leucine zipper transcription factor ATF-like (B-ATF), and H, runt-related transcription factor 1 (RUNX-1) were analyzed by real-time quantitative polymerase chain reaction. Hypoxanthine guanine phosphoribosyltransferase 1 was used as endogenous control, and knees from untreated mice were used as calibrator (ΔΔCt method). Values are the mean ± SEM fold induction in arthritic joints relative to that in joints from untreated TN-C+/+ mice (n = 4 mice per group). = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001.

Download figure to PowerPoint

Th17 cell polarization requires a unique cocktail of cytokines, as well as the coordinated activation of a complex network of transcription factors (for review, see refs.22 and23). Both IL-6 and IL-23, cytokines that are central to driving Th17 cell polarization and maintaining a differentiated phenotype, respectively, were rapidly induced in wild-type mice, peaking at 24 hours. This induction was significantly attenuated in TN-C−/− mice (Figures 5C and D). Moreover, expression of IL-23R was also significantly reduced in TN-C−/− mice as compared to TN-C+/+ mice (Figure 5E). In addition, while expression of the master transcription factor for Th17 cells, retinoic acid receptor–related orphan nuclear receptor γt (RORγt), was below the level of detection in the joints of both wild-type and tenascin-C–null mice (data not shown), the expression of other transcription factors that work together with RORγt and are required for optimal activation of the IL-17 gene (23) could be detected.

Expression of IRF-4, RUNX-1, and B-ATF was induced 6 hours after disease onset in both TN-C−/− and TN-C+/+ mice. There was no significant difference in the level of expression of B-ATF or RUNX-1 in either genotype over time. However, TN-C−/− mice exhibited a second peak of IRF-4 expression on day 5 that did not occur in TN-C+/+ mice (Figures 5F–H). No induction of cytokines, cytokine receptors, or transcription factors was observed in control mice.

Taken together, these data show that tenascin-C is important in inducing the production of cytokines that are essential for driving Th17 cell polarization, as well as factors that maintain the growth and stabilization of the Th17 cell phenotype, and may alter the balance of transcription networks that define the T cell phenotype during the pathogenesis of experimental RA.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Within the last decade, IL-17 has emerged as a major player in joint disease, driving synovial inflammation and proliferation, and directly inducing tissue destruction (4). This destructive capacity is corroborated by the correlation of high levels of IL-17 in RA patients with high levels of joint damage (24) and a strong dependence on IL-17 during the erosive phase of AIA and streptococcal cell wall–induced arthritis (25, 26). Evidence also suggests that IL-17–mediated inflammation may drive autonomous, TNFα-independent erosive disease (27, 28) and therefore could account for patients who do not respond to currently available treatments. This is supported by observations that TNFα therapy prevents erosions less effectively in RA patients with high IL-17 levels (4). Combination or sequential therapy that targets TNFα and IL-17 may therefore be effective, and patients with advanced joint damage may benefit from IL-17 blockade. Determining the mechanisms that promote IL-17 synthesis in the arthritic joint will aid our understanding of RA disease progression and may potentially highlight new therapeutic targets. In the present study, we identified a novel role of tenascin-C in driving IL-17–mediated immunity in joint disease.

Prompted by the high expression of tenascin-C in activated DCs, we compared the phenotype and function of BMDCs derived from TN-C+/+ and TN-C−/− mice. We found that BMDCs from TN-C−/− mice secreted lower levels of proinflammatory cytokines than did BMDCs from TN-C+/+ mice after activation with LPS. In addition, Th17 polarization was impaired in response to BMDCs derived from TN-C−/− mice. Moreover, during AIA, levels of IL-17 were significantly reduced.

BMDCs derived from TN-C−/− mice exhibited a significantly reduced ability to induce the differentiation of naive T cells into Th17 cells in vitro. This process was impaired regardless of whether DCs were activated with LPS or with M tuberculosis. BMDCs derived from TN-C−/− mice were, however, able to induce Th1, Th2, and Treg cell polarization to the same extent as BMDCs from wild-type mice, and no defect in BMDC stimulation of T cell proliferation was observed, suggesting a specific defect in Th17 polarization in the absence of tenascin-C. Consistent with these data, up-regulation of MHCII, CD86, and CD40, which are crucial for efficient T cell activation, was not impaired in BMDCs from TN-C−/− mice upon maturation. However, T cell differentiation also requires a cocktail of polarizing cytokines uniquely tailored for each effector class (2). Tenascin-C may induce a cytokine environment that is favorable for Th17 polarization. Key cytokines specific for Th17 cell differentiation include IL-6, which acts on naive T cells to initiate Th17 polarization (29), and IL-23, which via the activation of IL-23R in primed T cells, is necessary for the full differentiation of T cells into effector Th17 cells. IL-23 also promotes the proliferation and survival of already differentiated Th17 cells (30). During experimental arthritis, TN-C−/− mice produced significantly lower levels of IL-6, IL-23, and IL-23R than did TN-C+/+ mice, indicating that tenascin-C drives the production of cytokines required to induce and maintain Th17 cell polarization.

The mechanism by which tenascin-C aids cytokine synthesis is not known. Secreted tenascin-C can bind to a range of cell surface receptors, including integrins and heparan sulfate proteoglycans, in order to modulate cell spreading and cytoskeletal organization (31). Alterations in cell shape and internal architecture as a result of tenascin-C deficiency might impair vesicle transport or protein secretion in DCs. However, the fact that Th1, Th2, and Treg cell polarization proceeds normally in the absence of tenascin-C speaks against a global secretory dysfunction and in favor of a more specific cellular defect. Tenascin-C may induce the synthesis of factors specific for Th17 polarization by activating receptors on the BMDC surface and induce signaling pathways that result in de novo cytokine synthesis and secretion. Tenascin-C stimulates TNFα, IL-6, and IL-8 expression in primary human macrophages via activation of TLR-4 (10). TLR-4 activation has been shown to be required for the development of murine IL-17–dependent arthritis (32). Tenascin-C also stimulates IL-1α and IL-6 expression in murine macrophages via activation of α9 integrins (33) suggesting that a number of receptors may mediate tenascin-C–induced cytokine synthesis, either individually or cooperatively.

Also essential for Th17 cell polarization is a complex network of transcription factors that work synergistically with the master regulator of Th17 polarization, RORγt, to mediate IL-17 synthesis. B-ATF, RUNX-1, and IRF-4 are required for sustained RORγt expression in Th17 cells and are essential for optimal IL-17 gene transcription, IRF-4 via a mechanism or mechanisms unknown (34), B-ATF through direct interaction with the IL-17 gene promoter (35), and RUNX-1 via interactions with RORγt (36). However, a great deal of transcriptional cross-talk regulates T cell polarization in a context-dependent manner. While IRF-4 is critical for early Th17 differentiation, it is also expressed in other subsets of activated T cells: it drives Th2 cell polarization by inducing IL-4, IL-5, and IL-13 synthesis, is implicated in Treg cell differentiation by inducing IL-10 synthesis, and can negatively regulate Th1 cell differentiation by suppressing IFNγ synthesis (37). Our data show that in the presence of tenascin-C, IRF-4 expression is down-regulated during the latter stages of experimental joint disease. This may enable the suppression of IRF-4–mediated Th2 cell– or Treg cell–polarizing cytokines to ensure prolonged Th17 cell dominance in the joint.

Differentiation of bone marrow cells from TN-C−/− mice generated modestly (6%) lower numbers of CD11c+ cells as compared to cells from TN-C+/+ mice. This may be caused by the reduced capacity of TN-C−/− mice to generate DC precursors. Tenascin-C is expressed in the bone marrow by stromal cells (38), and TN-C−/− mice produced markedly lower numbers of hematopoietic progenitor cells (39). This was due to a defect in early stromal cell–mediated hematopoiesis (39), where tenascin-C may promote the adhesion of hematopoietic cells to stromal cells (38, 40). Consistent with these data, we observed that TN-C−/− mice yielded significantly lower numbers of total bone marrow cells than did TN-C+/+ mice (data not shown) as well as fewer BMDCs. Lower levels of cytokine production may result from slightly fewer CD11c+ BMDCs derived from the TN-C−/− mice or from the lower surface expression levels of CD11c. However, there was no difference in the maturation state of BMDCs generated from TN-C+/+ mice and TN-C−/− mice, making this explanation unlikely.

Several studies have implied a role of tenascin-C in adaptive immunity by regulating T cell behavior. For example, tenascin-C can directly affect lymphocyte function by promoting migration (41), potentially via anti-adhesive effects (42). In contrast, glioma cells can inhibit the transmigration of T cells by the expression of tenascin-C (43). Our data suggest a specific role for tenascin-C in promoting DC-mediated Th17 cell differentiation. This is corroborated by the lack of significant induction of IL-17 in vivo during AIA in mice lacking tenascin-C.

IL-17 blockade reduces disease severity in murine collagen- and adjuvant-induced arthritis and protects against cartilage damage (26, 44–46). This approach also inhibits cytokine synthesis in human RA synovial tissue (47, 48) and, in phase I clinical trials, was shown to suppress RA symptoms (49, 50). However, global blockade of IL-17 may not be without drawbacks; it is a key part of the immune response, and its inhibition may confer unwanted side effects, including suppression of neutrophil infiltration upon infection (4). Defining disease-specific stimuli that drive IL-17 synthesis in the RA joint would enable the exclusive targeting of pathologic inflammation without affecting physiologic host defense.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

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. Midwood 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. Kong, Midwood.

Acquisition of data. Ruhmann, Piccinini, Kong.

Analysis and interpretation of data. Ruhmann, Piccinini, Kong, Midwood.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We thank Emma Timms, Katrina Blazek, and Fiona McCann for technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES