Anti-Toll-like receptor 2 and 4 antibodies suppress inflammatory response in mice

Abstract

Toll-like receptors (TLRs) 2 and 4 recognize different endogenous and exogenous agonists and play a distinct role in infection and inflammation. However, their ultimate effect in a given infectious and inflammatory disease is less understood. We produced polyclonal anti-murine TLR2 and TLR4 antibodies and investigated their effect on cutaneous leishmaniasis and inflammatory arthritis. Administration of these antibodies to susceptible BALB/c mice, infected in the footpad with Leishmania major, reduced footpad swelling but not the parasite load compared with mice treated with control IgG. The antibodies synergistically reduced leishmanial-specific T-cell proliferation, T helper type 1 and type 2 cytokine production, specific IgG1 and IgG2a antibody synthesis, and T-cell receptor and co-stimulatory molecule expression on dendritic cells in infected mice. We then tested the effect of these antibodies on collagen-induced arthritis (CIA) in DBA/1 mice, a classic model of chronic inflammation. Both antibodies markedly suppressed the development of clinical parameters with concomitant reduction of pro-inflammatory cytokine production. These data therefore suggest that anti-TLR2 and 4 antibodies may have a synergistic therapeutic effect on inflammatory disease in vivo.

Introduction

Toll-like receptors (TLRs) are a class of pattern recognition receptors, originally identified in Drosophila.¹ At least 13 distinct TLRs have been identified in mammals.^2–4 TLRs sense conserved exogenous pathogen-associated molecular patterns and endogenous host damage-associated molecular patterns with target specificity.^2–4 For example, TLR2 selectively recognizes glycolipid of Mycobacteria,⁵ peptidoglycan of Gram-positive bacteria⁶ and Leishmania lipophosphoglycan,⁷ while TLR4 recognizes lipopolysaccharide (LPS) of Gram-negative bacteria,⁸ Leishmania proteoglycolipid complex (P8),⁹ and host saturated fatty acids, β-defensins and heat-shock protein 60.^3,4

TLR2 and TLR4 play important roles in a wide-range of infectious and inflammatory diseases^3,4 and are potential therapeutic targets in inflammatory diseases.¹⁰ However, their role in cutaneous leishmaniasis is less understood and controversial.^11,12 Leishmania parasites express both TLR2 and TLR4 ligands, lipophosphoglycan and proteoglycolipid complex (P8), respectively.^7,9,11,12 However, the ultimate effect of TLR2 and TLR4 in leishmaniasis in vivo is unknown.

Furthermore, most studies of TLRs in Leishmania infection used TLR-deficient mice on a C57BL/6 background, which are resistant to the infection by developing a T helper type 1 (Th1) immune response, whereas the role of TLRs in susceptible BALB/c mice, which develop non-healing lesions by generating a Th2 type response, is less studied.^11–13

We therefore generated anti-mouse TLR2 and TLR4 antibodies and investigated the effect and co-effect of these antibodies on Leishmania major infection in the highly susceptible BALB/c mice. We found that anti-TLR2 and anti-TLR4 antibody treatment reduced L. major-mediated tissue inflammation but not the parasite burden. We then determined the effect of these antibodies on collagen-induced arthritis (CIA) in DBA1 mice. Again, the antibodies synergistically suppressed the development and clinical parameters of CIA. These findings therefore extend the role of TLRs to parasitic infection in susceptible host and suggest that anti-TLR antibodies may have therapeutic potential against inflammatory diseases.

Materials and methods

Mice

BALB/c (female, 8–12 weeks old), and DBA/1 (male, 6–8 weeks old) mice were obtained from Harlan Olac (Bicester, UK). TLR2^−/− and TLR4^−/− mice have been described previously.¹⁴ Mice were housed in specific pathogen-free conditions at Glasgow University, UK, and procedures were in accordance with the UK Home Office animal experimentation guidelines.

Anti-TLR antibodies

Rabbits were immunized with the peptides corresponding to residues 287–311 of murine TLR2 (GenBank accession no. AF124741), or residues 252–274 of murine TLR4 (GenBank accession no. AF110133). The peptides were selected to have no homology to known proteins, spanning the hydrophilic region and within the extracellular domain of the TLRs. The rabbits were immunized subcutaneously with 100 μg of the peptide conjugated to keyhole limpet haemocyanin emulsified in complete Freund's adjuvant (CFA; Difco, Detroit, MI), and boosted sequentially with the peptide–keyhole limpet haemocyanin in incomplete Freund's adjuvant (Difco). Specificity and titres of the antibodies were tested with peptide-bound 96-well plates by ELISA. Total IgG was purified from the immunized and pre-immunized rabbit serum by ammonium sulphate precipitation and protein A column. Purity of the IgG was > 95% on SDS–PAGE by Coomassie blue staining. LPS content of the preparation was < 0·01 ng/μg protein by an amoebocyte Limulus test (E-toxate; Sigma, Poole, UK).

Flow cytometry analysis and cell culture

Bone-marrow-derived macrophages (BMDMs) were generated from the femurs of adult TLR2^−/−, TLR4^−/− and wild-type (WT) mice and cultured for 7 days with recombinant CSF-1 as described previously.¹⁵ BMDMs were plated out at 1·5 × 10⁶ cells/ml in complete medium [RPMI-1640 plus 10% fetal calf serum, penicillin/streptomycin and glutamine, all from Invitrogen (Paisley, UK)] in six-well plates and treated with LPS (100 ng/ml, Salmonella minnesota; Sigma) or were left untreated for 10 hr. Adherent cells were harvested by scraping the cells from culture wells in ice-cold PBS, resuspended in ice-cold PBS containing 2% fetal calf serum and incubated at 4° for 30 min with CD16/CD32 (BD Biosciences, Oxford, UK) to block the Fc receptor. The cells were then incubated with anti-TLR antibody or normal IgG (all 7 μg/ml) for 20 min at 4°. Cells were stained with FITC-conjugated anti-rabbit IgG monoclonal antibody and analysed on a dual laser (488 and 633 nm) FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA). Data analysis was performed using CellQuest Pro (Becton Dickinson).

To test the effect of anti-TLR antibodies on BMDM functions in vitro, BMDMs from WT mice were cultured with normal rabbit IgG, rabbit anti-TLR2 or anti-TLR4 antibody (25 μg/ml) for 1 hr. LPS (100 ng/ml, derived from Salmonella minnesota Re-595; Sigma) or peptidoglycan (10 μg/ml, derived from Staphylococcus aureus; Sigma) was then added, respectively and the cells were cultured for a further 10 hr. Culture supernatants were harvested and assayed for tumour necrosis factor-α (TNF-α) production level by ELISA.

Leishmanial infection and immunological analysis

Mice were infected in the left hind footpad with 5 × 10⁵ stationary phase promastigotes of L. major (LV39) and injected intraperitoneally with 250 μg/mouse of normal rabbit IgG, anti-TLR2, anti-TLR4, or a combination of the two antibodies (half the amount of each). The same amount of antibody was administered once a day for the next 2 days, and then every 3–4 days for 3 weeks. Footpad swelling (the difference between the infected left and uninfected right hind footpad) was measured at regular intervals. The maintenance of parasite infection and measurement of disease progression were as described previously.¹⁶ At the end of the experiment, mice were killed and footpads were removed to assay for parasite load by limiting dilution.¹⁶ Draining lymph node cells were harvested and cultured (2 × 10⁶ cells/ml of full medium) in vitro with parasite antigen prepared by freeze–thawed whole parasites (10⁵–10⁶ parasite/ml equivalent). Culture supernatant was harvested at 72 hr and assayed for cytokines by ELISA. Cellular proliferation was also analysed by [³H]thymidine incorporation.¹⁶ The results were read on a Perkin Elmer Micro Beta Trilux (PerkinElmer Life and Analytical Sciences, Shelton, CT), and expressed as counts per minute (c.p.m.).

Collagen-induced arthritis

The induction and analysis of CIA was as described previously.¹⁷ Briefly, male DBA/1 mice were immunized with 200 μg of bovine type II collagen (Sigma) emulsified in CFA by intradermal injection (day 0). Collagen (200 μg in PBS) was given again on day 21 by intraperitoneal injection. Anti-TLR2, anti-TLR4, or a combination of the two antibodies, or normal rabbit IgG (all 300 μg/mouse/day) were administered intraperitoneally daily from day 1 to 7. Paw thickness was measured daily from day 21 with a dial-calliper (Kroeplin, Schlüchtern, Germany). Clinical scores were assigned as follows: 0 – normal, 1 – erythema, 2 – erythema + swelling, 3 – extension/loss function; total score was the sum of the four limbs.

Cytokines and antibody titrations

Cytokine concentrations in culture supernatants were determined by ELISA, using paired antibodies (BD Biosciences) according to the manufacturer's instructions. Detection limits were: interleukin-6 (IL-6) and IL-12, 10 pg/ml; IL-4, IL-10, TNF-α, 30 pg/ml; interferon-γ (IFN-γ), 40 pg/ml. Anti-leishmanial antibody titres in the pooled sera (n = 5) were detected with biotin-conjugated anti-mouse IgG1 or IgG2 (MD Biosciences, St. Paul, MN) followed by conjugated avidin peroxidase (Sigma) and developed with tetramethylbenzidine substrate (Kirkegard & Perry, Gaithersburg, MD).

Quantitative PCR

RNA was purified from tissue samples using the RNeasy Mini Kit following the manufacturer's instructions (Qiagen, Manchester, UK). Reverse Transcription (RT) of RNA into cDNA was carried out using High-Capacity cDNA Reverse Transcription Kits (Applied Biosystems, Foster City, CA). Real-time PCR was performed using the specific probe and primers from Applied Biosystems and Fast SYBR Green master mix on a Prism 7900HT (Applied Biosystems). The cDNA levels during the linear phase of amplification were normalized against hypoxanthine ribosyltransferase controls.

Statistical analysis

Student's t-test was applied to in vitro studies. Analysis between individuals in groups in vivo was by analysis of variance followed by Student's t-test. Results are expressed as mean ± SD from five to 10 mice, and are representative of at least two individual experiments. P < 0·05 was considered significant.

Results

Generation and characterization of anti-TLR antibodies

We produced polyclonal antibodies against murine TLR2 and TLR4. The antibodies specifically stained TLR2 or TLR4, respectively, on BMDMs from WT mice by flow cytometry (Fig.1a). Although both antibodies positively stained cells from WT mice, anti-TLR2 antibody only stained cells from TLR4^−/− mice but not cells from TLR2^−/− mice. Conversely, anti-TLR4 antibody stained cells from TLR2^−/− mice but not cells from TLR4^−/− mice. Importantly, the intensity of staining by the antibodies on cells of the respective TLR-deficient mice was the same as that on cells from the WT mice double-stained with the two antibodies. This result suggests that binding of one antibody on a TLR on the cell surface did not affect the expression of the other TLR. Functionally, anti-TLR4 antibody but not anti-TLR2 antibody significantly suppressed the production of TNF-α by TLR4 agonist LPS-activated macrophages (Fig.1b). Conversely, anti-TLR2 antibody but not anti-TLR4 antibody markedly inhibited TNF-α production by TLR2 agonist peptidoglycan-activated BMDMs. It should also be noted that the inhibitory effect of these antibodies in vitro is unlikely to be a result of the direct cytotoxic activity of the antibodies, as the viability of the BMDMs remained unchanged at the end of the culture period (data not shown). Since LPS and peptidoglycan are established ligands for TLR4 and TLR2, respectively,¹⁴ these data not only demonstrate the specificity of the antibodies, they also suggest that these antibodies are functionally active, probably by blocking the receptor–ligand interaction of their respective TLRs.

Figure 1.

Characterization of anti-Toll-like receptor 2 (TLR2) and anti-TLR4 antibodies. (a). Flow cytometric analysis shows that both antibodies positively stained the surface of wild-type bone-marrow-derived macrophages (BMDMs). Anti-TLR4 antibody (thick line) but not the anti-TLR2 antibody (thin line) stained the BMDMs of TLR2^−/− mice. Conversely, anti-TLR2 antibody but not anti-TLR4 antibody stained BMDMs from TLR4^−/− mice. Dotted line represents staining with control normal rabbit IgG. (b) Anti-TLR4 but not anti-TLR2 antibody significantly inhibited lipopolysaccharide (LPS) -induced tumour necrosis factor-α (TNF-α) production by wild-type BMDMs. Conversely, anti-TLR2 but not anti-TLR4 antibody inhibited peptidoglycan (PGN) -induced TNF-α synthesis by BMDMs. Results of ELISA are mean ± SD, n = 3, and are representative of three experiments. *P < 0.05 compared with IgG control.

Effect of anti-TLR antibodies on L. major infection

The role of TLRs in host susceptibility to Leishmania infection is less studied. We next investigated the effect of these antibodies on L. major infection in the highly susceptible BALB/c mice. We initially determined the kinetic expression of TLR2 and TLR4 in the L. major-infected footpads. As shown in Fig.2(a), TLR2 expression was induced 3 days after the infection and declined on day 21. In contrast, TLR4 expression was reduced on day 3 but then enhanced on day 21 (Fig.2a). Furthermore, although both TLRs were detected in the infected tissues, the expression levels of TLR4 were lower than TLR2.

Figure 2.

Effect of anti-Toll-like receptor 2 (TLR2) and anti-TLR4 antibodies on Leishmania major infection of BALB/c mice. Groups of 10 BALB/c mice were infected in the footpads with 5 × 10⁵ stationary phase promastigotes and treated with the antibodies or control IgG on days 1 and 2 and then every 3–4 days for 3 weeks as indicated. (a) The kinetic expression of TLR2 and TLR4 in infected footpads. The antibodies significantly reduced the footpad swelling of the mice (b), but not the parasite load in the infected footpad (c) compared with control. Data are mean ± SD, n = 10 and are representative of three experiments. *P < 0.05, **P < 0.01 compared with IgG control.

We next determined the effect of anti-TLR antibodies on Leishmania infection in vivo. Mice were treated with anti-TLR2, anti-TLR4 or a combination of the two antibodies. Control mice were injected with normal IgG. Mice injected with anti-TLR2 or anti-TLR4 antibody showed a modest but significant reduction in footpad swelling. In contrast, mice treated with a combination of the two antibodies developed markedly reduced footpad swelling (Fig.2b). At the end of experiment (terminated as required by Home Office guidelines when control mice developed ulcerated lesions), parasite load in the footpad was estimated. Surprisingly, despite a significant decrease in footpad swelling, there was no difference in the parasite load in the footpad among all groups of mice (Fig.2c).

Anti-TLR antibodies reduced Leishmania-specific T- and B-cell response in vivo

We sought next to understand how the footpad swelling in infected mice was reduced by determining Leishmania-specific T- and B-cell response in the mice. Draining lymph node cells were harvested from the mice, stimulated in vitro with leishmanial antigen, and T-cell proliferation and cytokine production in the supernatant measured. Mice treated with the antibody showed significant reduction in the specific T-cell proliferative response compared with the control mice (Fig.3a). Cells from mice treated with anti-TLR2 produced significantly reduced levels of both Th1 and Th2 cytokines; IFN-γ, TNF-α, IL-4 and IL-6 (Fig.3b). The effect of anti-TLR4 was variable. It reduced the production of IL-4 and IL-6, but had no effect on IFN-γ or TNF-α secretion. Mice treated with a combination of the two antibodies also showed a marked reduction in all the cytokines tested, in general the cytokine reduction was even more profound than that induced by anti-TLR2 alone. While IL-10 plays a key role in the non-healing response to L. major infection,¹³ the antibody treatments did not significantly change the production level of IL-10 compared with IgG control in the infected mice (data not shown). The combined antibody (but not single antibody) treatment also suppressed serum anti-leishmanial antibody synthesis of both IgG1 and IgG2a isotypes (Fig.3c).

Figure 3.

Anti-Toll-like receptor 2 (TLR2) and anti-TLR4 antibodies suppress T helper type 1 (Th1), Th2 and pro-inflammatory cytokine response of Leishmania major-infected BALB/c mice. The draining lymph node cells and serum from mice depicted in Fig.2 were collected. Cells were cultured with 10⁶ organisms equivalent of frozen-thawed promastigotes. The antibody treatment also markedly reduced the leishmanial antigen-specific T-cell proliferation of the draining lymph node cells (a). Supernatants collected at 72 hr were analysed for cytokine production by ELISA (b). Leishmania major-specific IgG1 and IgG2a antibodies titres in the pooled serum were also determined (c). Results are mean ± SD, n = 10. *P < 0.05, **P < 0.01 compared with control IgG-treated group.

Anti-TLR antibodies reduced the expression of TCR and co-stimulatory molecules on DCs in vivo

Since the treatment with anti-TLR antibodies led to markedly suppressed both T- and B-cell responses we further determined whether the antibodies affected antigen-presenting cells by assessing their expression of TCR and co-stimulating molecules. We collected draining lymph nodes from mice 50 days after infection with L. major and treated with a combination of anti-TLR2 and anti-TLR4 antibodies or the control IgG as described above. The draining lymph nodes were assayed for the expression of CD86, CD80 and MHC class II on CD11c⁺ DCs by flow cytometry. The percentage of dendritic cells expressing CD86, CD80 and MHC class II from the antibody-treated L. major-infected mice was significantly lower than cells from mice treated with the control IgG (Table1).

Table 1.

Anti-Toll-like receptor 2 (TLR2) and anti-TLR4 antibodies reduced the expression of co-stimulatory molecules on CD11c⁺ dendritic cells in vivo

Molecules	IgG-treated (%)	Antibody-treated (%)	Fold reduction
CD86	2·75	1·80*	1·5
CD80	4·25	2·10*	2·0
MHCII	3·67	2·54*	1·5

Draining lymph nodes from mice infected with Leishmania major and treated with normal IgG or with a combination of anti-TLR2 and anti-TLR4 antibody (see Fig.2) were collected on day 50 after infection and assayed for the cell surface markers by flow cytometry. Per cent of total cells double-stained for CD11c and the markers indicated was calculated.

Results are pooled from two experiments,

^*P < 0·05 compared with respective IgG control.

Effects of anti-TLR antibodies on collagen-induced arthritis

Since anti-TLR2 and anti-TLR4 antibodies are apparently anti-inflammatory in Leishmania infection, we further investigated their potential role in a well-studied chronic inflammatory model, CIA. The induction of CIA entails intradermal priming of DBA/1 mice with bovine type II collagen (CII) emulsified in CFA. Anti-TLR2 and anti-TLR4 antibodies were administered intraperitoneally to mice from day 1 to 7 relative to the induction of CIA (on day 0). The mice were boosted intraperitoneally with CII in PBS on day 21 and the joint swelling and disease were scored daily for 24 days.¹⁷ As shown in Fig.4, the control mice treated with normal rabbit IgG exhibited the expected disease profile, which was apparent on day 5 and peaked around day 20 post-CII challenge. In contrast, mice treated with the antibodies exhibited minimal arthritic disease and irrespective of whether the antibodies were used individually or in combination. Importantly, serum from the antibody-treated mice contained significantly less pro-inflammatory cytokines IFN-γ, IL-12 and TNF-α compared with serum from mice treated with control IgG (Fig.4). These data therefore demonstrate that the anti-TLR antibodies are potently anti-inflammatory.

Figure 4.

Anti-Toll-like receptor (TLR) antibodies inhibit the induction of collagen-induced arthritis (CIA). Groups of DBA/1 mice were primed intradermally with CII in complete Freund's adjuvant and challenged intraperitoneally 21 days later with CII in PBS. Mice were injected intraperitoneally with anti-TLR antibodies or control normal IgG from the day before priming and then daily for 7 days. (a) Arthritic disease index was recorded daily after challenge injection. (b) Serum samples were collected on day 24 post challenge and assayed for cytokines by ELISA. Data are mean ± SD, n = 8 and are representative of two experiments. *P < 0.05, **P < 0.01 compared with IgG control.

Discussion

Data reported here demonstrate that anti-TLR2 and anti-TLR4 antibodies markedly suppress the inflammatory response elicited by infection by the protozoan parasite L. major, and CIA. These results enhance current understanding of the role of TLRs in Leishmania infection and inflammatory arthritis. Furthermore, our findings suggest that the anti-TLR antibodies may have potential therapeutic effects against inflammatory diseases.

The anti-TLR2 and anti-TLR4 antibody treatment synergistically reduced footpad swelling in L. major-infected BALB/c mice, this is commensurate with previous reports that both TLR2 and TLR4 signals are closely associated with Leishmania infection.^11,12 The synergistic effect between anti-TLR2 and anti-TLR4 antibodies observed here is also consistent with reports showing synergistic signalling pathways among TLRs in vitro.^18–20

While synergistically reducing footpad swelling, these antibody treatments had no significant effect on parasite load in L. major-infected BALB/c mice. This suggests that the decreased footpad swelling in the infected mice, following antibody treatment, reflects the reduced inflammatory reaction associated with reduced T- and B-cell responses in the infected tissue.

It is currently unclear why the antibody treatment reduced footpad swelling but not parasite load in infected BALB/c mice. Leishmania parasites express both TLR2 and TLR4 agonists.^7,9 Available evidence generated from TLR-deficient mice in resistant C57BL/6 background suggests that TLR2 and TLR4 signals play a contradictive role in the growth of Leishmania parasites in mice.¹¹ Although TLR4-deficient mice increased the growth of L. major parasites,²¹ TLR2-deficient mice reduced number of Leishmania braziliensis and Leishmania amazonensis compared with WT mice.^22,23 However, the ultimate effect of TLR2 and TLR4 on Leishmania infection in a given context is unknown. Our results suggest that blocking both TLR2 and TLR4 signalling pathways may abolish the contradictive TLR2- and TLR4-mediated effect in Leishmania infection, and so had no influence on the parasites' growth/survival in the susceptible BALB/c mice.

We found that the antibody treatments reduced both type 1 and type 2 immune responses and pro-inflammatory cytokine production. It is well known that Th1 cells can protect against L. major infection by inducing IFN-γ,^13,24 whereas Th2 cells promote the infection by secreting IL-4 and IL-13.¹³ Th1 and Th2 cells can also counter-regulate each other's function via the cytokines they produce.^13,24 It is also known that unbalanced Th1 and Th2 response determines the outcome of L. major infection in mice.^13,24,25 The lack of influence of the antibody treatment on parasite loading may be a result of the simultaneously reduced but rebalanced Th1 and Th2 responses after treatment in this context. Current evidence suggest that IL-10 produced by regulatory T cells also contributes to the non-healing response to L. major infection.¹³ Since anti-TLR treatments did not significantly alter IL-10 levels in the Leishmania-infected mice, it is unlikely that IL-10 is responsible for the TLR-mediated effect in this context.

The mechanism by which anti-TLR antibodies suppress the specific T- and B-cell-mediated immune response is still largely unknown. Since dendritic cell maturation is required for optimal antigen-presentation of dendritic cells, and TLR signals can promote dendritic cell maturation, it is most likely that the antibody treatments could diminish Leishmania-elicited immune response by blocking TLR2- and TLR4-mediated dendritic cell maturation and thereby interfere with the presentation of leishmanial antigen by these professional antigen-presenting cells.^26–28 Consistent with this, we have found that CD11c⁺ dendritic cells from the infected mice treated with a combination of the two antibodies significantly reduced the expression of dendritic cell maturation markers CD86, CD80 and MHC II compared with cells from mice treated with the control IgG. These molecules are well documented and are crucial to the induction of adaptive immune responses.

The suppression of CIA by both anti-TLR2 and anti-TLR4 antibodies is intriguing; the anti-TLR4 antibody is as effective as the anti-TLR2 antibody in this model. Furthermore, the antibody treatments reduced the production of key pro-inflammatory cytokines, IL-12 and TNF-α, which play a pathogenic role in many pro-inflammatory and autoimmune diseases.¹⁷ This is consistent with previous reports that TLR2 and TLR4 signals are generally pro-inflammatory in other inflammatory arthritis models and suggests that the anti-inflammatory effect observed here may extend beyond CIA to other inflammatory disorders.^29–31 While the detailed mechanisms are still unknown, the neutralizing antibodies could block the TLRs-mediated effects by directly antagonizing the ligand/TLRs interaction in the arthritic context. The possible TLR2 ligands may include the endogenous gp96 and serum amyloid A and extraneous mycobacteria components present in the complete adjuvant.^3,4,30 Possible TLR4 ligands may be the endogenous fibrinogen, hyaluronic acid and HMGB-1 and exogenous LPS, which are present in vivo and proven important for the induction of CIA.^3,4,31

In conclusion, this study demonstrates the feasibility of modulating the inflammatory immune response in vivo using both anti-TLR2 and anti-TLR4 antibodies. This finding will facilitate the analysis of the biological functions of TLRs in vivo and may lead to novel therapeutic approaches against inflammatory diseases.

Glossary

BMDMs

bone marrow-derived macrophages

CFA

complete Freund's adjuvant

CIA

collagen-induced arthritis

CII

bovine type II collagen

IFN-γ

interferon-γ

IL-6

interleukin-6

LPS

lipopolysaccharide

Th1

T helper type 1

TLR

Toll-like receptor

TNF-α

tumour necrosis factor-α

WT

wild-type

Disclosures

The authors have no financial conflicts of interest.