TLR Tolerance as a Treatment for Central Nervous System Autoimmunity

Emily J. Anstadt, Mai Fujiwara, Nicholas Wasko, Frank Nichols and Robert B. Clark
J Immunol September 15, 2016, 197 (6) 2110-2118; DOI: https://doi.org/10.4049/jimmunol.1600876

Abstract

The role of TLR signaling in multiple sclerosis (MS) and experimental autoimmune encephalomyelitis (EAE) is unclear. This role is especially controversial in models of adoptive transfer EAE in which no adjuvant and no TLR ligands are administered. We recently reported that a microbiome-derived TLR2 ligand, Lipid 654 (L654), is present in healthy human serum but significantly decreased in the serum of MS patients. This suggested that microbiome products that gain access to the systemic circulation, rather than being proinflammatory, may normally play an immune-regulatory role by maintaining a state of relative TLR tolerance. Therefore, a loss of microbiome-mediated TLR tolerance, as suggested by lower serum levels of L654, may play a role in the pathogenesis of MS. As proof of concept we asked whether administering low-level TLR2 ligands in adoptive transfer EAE induces TLR2 tolerance and attenuates disease. We administered low-level Pam2CSK4 or L654 to mice receiving encephalitogenic cells and in doing so induced both TLR2 tolerance and attenuation of EAE. Disease attenuation was accompanied in the CNS by a decrease in macrophage activation, a decrease in a specific proinflammatory macrophage population, and a decrease in Th17 cells. In addition, disease attenuation was associated with an increase in splenic type 1 regulatory T cells. Kinetic tolerance induction studies revealed a critical period for TLR2 involvement in adoptive transfer EAE. Overall, these results suggest that inducing TLR tolerance may offer a new approach to treating CNS autoimmune diseases such as MS.

Introduction

Multiple sclerosis (MS) is a progressive demyelinating disease of the CNS that affects 1 in 1000 people in the United States. There is both a genetic and environmental component to the disease etiology, although these contributions have yet to be understood. Treatments for MS are currently based on a limited understanding of the disease pathogenesis, are broadly immunosuppressive, and largely target T and B cells. These treatments have the potential to decrease the frequency of relapses, but are not yet able to delay the progression of MS.

We have previously reported that a microbiome-derived TLR2 agonist, Lipid 654 (L654), is present in the circulation of healthy individuals. Importantly, L654 is recovered in significantly lower levels in the serum of patients with MS (1). Although pathogen-associated molecular patterns such as L654 are normally thought of as proinflammatory, we asked whether a decrease in TLR2 ligands could have relevance to the disease pathogenesis of MS.

TLR2 expression has been reported to be upregulated on oligodendrocytes and on PBMCs of MS patients (2, 3). The role of TLR2 signaling in MS patients, however, is controversial. Helminth-infected MS patients were found to have TLR2-dependent programming of dendritic cells that induced regulatory T cells (Tregs), suppressed production of proinflammatory cytokines, and was associated with upregulation of genes involved in retinoic acid biosynthesis and metabolism (4). In contrast, Nyirenda et al. (5) reported that TLR2 activation is involved in the reduced Treg function reported in MS. These authors found that stimulation with Pam3CSK, a TLR1/TLR2 agonist, reduced Treg function and induced Th17 skewing, and this effect was enhanced in MS patients compared with healthy controls. Interestingly, they reported that the TLR1/TLR2 heterodimer, but not the TLR6/TLR2 heterodimer, mediated these effects, and MS patients compared with healthy controls had a greater frequency of Tregs that expressed the TLR1/TLR2 heterodimer, but not the TLR6/TLR2 heterodimer.

In the mouse model of MS, experimental autoimmune encephalomyelitis (EAE), TLR2 expression has been reported to be upregulated in the spinal cord (6). Despite this increase in expression, the role of TLR2 has remained unclear. The effect of TLR2 deficiency in various EAE models is controversial. TLR2-deficient mice have been shown to variably develop comparable disease severity, attenuated disease severity, or no disease at all in both active disease and adoptive transfer models of EAE (6–10). As no adjuvant is used in most adoptive transfer models of EAE, this potential TLR2 dependence suggests that either endogenous damage-associated molecular patterns or microbiome-derived TLR2 ligands are involved in disease pathogenesis.

In contrast to the potential requirement for TLR2 in the EAE disease process, other studies have reported that administering relatively large amounts of TLR2 ligands can attenuate EAE through induction of various Treg populations (11–14). Thus, although the literature appears contradictory in suggesting both a requirement for and inhibitory capacity of TLR2 signaling in EAE, the concept of TLR tolerance can potentially resolve this paradox. It has been well documented that repeated ligation through TLRs results in subsequent signal dampening and a state of TLR tolerance, which both prevents successive ligations from signaling at normal magnitude and can change the downstream effector molecules to reflect a more regulatory phenotype (15). If TLR2 signaling is truly disease promoting in EAE, then a dampening of this signaling pathway through repeated ligation and tolerance development would lead to disease attenuation.

Consistent with this postulate, our finding of lower serum L654 levels in patients with MS (1) suggested that microbiome products gaining access to the systemic circulation might normally play an immune-regulatory role. Tonic interaction of TLRs with low-level microbiome-derived ligands may lead to relative TLR tolerance and establish the threshold and magnitude of systemic TLR signaling. A deficiency of microbiome-mediated TLR tolerance could then play a role in the pathogenesis of MS.

In proof-of-concept studies, we now show that administering a low dose of either a canonical TLR2 ligand or the human serum-accessing microbiome product, L654, induces TLR tolerance and attenuates adoptively transferred EAE. Additionally, utilizing kinetic tolerance induction studies, we show the requirement for TLR2 in adoptive transfer EAE occurs during a critical window between day +4 and +8 postcell transfer. Mechanistically, this disease attenuation induced by TLR2 tolerance induction is accompanied by alterations in CNS APCs, CNS T cells, and splenic regulatory cells. We suggest that TLR tolerance induction is a normal mechanism of microbiome-mediated immune regulation, deficient in MS, and a potential new therapeutic approach for CNS autoimmune disease.

Materials and Methods

Mice

Female C57BL/6 mice and SJL/J mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were maintained under specific pathogen-free conditions in accordance with the guidelines for the Center for Comparative Medicine at the University of Connecticut Medical School. All procedures were performed under Institutional Animal Care and Use Committee–approved protocols.

Reagents

Proteolipid protein (PLP)139–151 was synthesized by GenScript (Piscataway, NJ). Mycobacterium tuberculosis H37 Ra was purchased from BD Difco (Franklin Lakes, NJ). IFA was obtained from Sigma-Aldrich (St. Louis, MO). Purified Pam2CSK4 (Pam2Cys), purified lipoteichoic acid (LTA), oligodeoxynucleotide (ODN) 1668, and R848 were obtained from InvivoGen (San Diego, CA). L654 was derived from Porphyromonas gingivalis (American Type Culture Collection 33277 type strain) and purified, as previously described (16).

In vitro tolerance induction in bone marrow–derived macrophages

Bone marrow–derived macrophages (BMDMs) were isolated from C57BL/6 mice by flushing marrow from femur and tibia of mice and culturing the cells obtained in 10% FBS and 10% L929 supernatant in DMEM. Cells were cultured for 24 h, and nonadherent cells were replated for another 6 d of culture. To obtain L929 supernatant, L929 cells from American Type Culture Collection (Manassas, VA) were cultured in 10% RPMI 1640 for 1 wk after confluence was reached. Assay media consisted of the BMDM differentiation media without L929 supernatant. BMDMs were pretreated with 500 pg/ml; 1 or 5 ng/ml Pam2Cys; and 1, 5, or 7 μg/ml L654 or vehicle control (VC) for 24 h. Cells were washed and a secondary TLR2 stimulus was provided by the addition of 5 ng/ml Pam2CSK4 (Pam2Cys). Cells were then cultured for another 24 h, and supernatant was collected. TNF-α in the supernatant was measured by the Ready-SET-Go! ELISA kit from Affymetrix (Santa Clara, CA).

In vivo tolerance induction studies

SJL/J mice were i.v. administered 0.35 μg Pam2Cys in PBS or PBS alone (VC) daily for 5 d. A total of 300 μg LTA was i.v. administered to mice 1, 3, or 5 d after Pam2Cys/VC treatment halted. Mice were bled 2 h later, and serum TNF-α was assayed by ELISA. In cross-tolerance studies, SJL/J mice were treated with Pam2Cys or VC as above, and 40 μg ODN 1668 was i.v. administered or 80 μg R848 was i.p. administered 1 d after the last Pam2Cys/VC administration. Mice were bled 1.5 h later, and serum TNF-α was assayed by ELISA.

T cell adoptive transfer EAE

The PLP139–151–specific model of adoptively transferred EAE in SJL/J mice was used in all EAE studies (17). Briefly, draining lymph node cells (LNCs) from mice immunized 9 d prior with PLP139–151 in CFA were restimulated in vitro with PLP139–151. After 4 d, the cultured cells (29 × 106 cells/mouse) were transferred i.p. to naive female SJL/J mice. No pertussis toxin is used in this protocol. Clinical grading of EAE, as previously described (18), was performed daily on all mice for 60 d.

Derivation of spinal cord mononuclear cells

Spinal cord mononuclear cells were derived, as previously described (19). Briefly, mice were perfused through the left cardiac ventricle, and the spinal cords were removed and weighed. Spinal cord tissue was digested with 2.5 mg/ml collagenase D (Roche Diagnostics, Indianapolis, IN) and 1 mg/ml DNase I (Sigma-Aldrich) at 37°C for 45 min. Mononuclear cells were isolated by passing the tissue through a 70-μm cell strainer, followed by a 70–37% Percoll gradient centrifugation. Mononuclear cells were removed from the interphase, washed, and resuspended in culture medium.

Flow cytometric analyses

Single-cell suspensions of splenocytes or spinal cord mononuclear cells were restimulated for 4 h with 200 ng/ml PMA and 1 μg/ml ionomycin in the presence of 5 μg/ml brefeldin A (Sigma-Aldrich). Cells were blocked with anti-mouse CD16/32 (101302; BioLegend, San Diego, CA) and stained with anti-mouse CD4-BV510, CD11b-FITC, CD11c-Alexa Fluor 700, and/or F4/80-PE/Cy7 (100559, 101206, 117319, 123113; BioLegend); CD39-PerCP-EF710 (46-0391, 56-5321-82; Affymetrix); CD80-PE (553769; BD Biosciences, San Jose, CA); L/D Near IR (L34975; Molecular Probes, Eugene, OR); and/or CD62L-PE (50-0621; Tonbo Biosciences, San Diego, CA). Cells were permeabilized and fixed (Fixation/Permeabilization Solution Kit, 554714; BD Biosciences) and stained with anti-mouse IFN-γ violetFluor450 (75-7311; Tonbo Biosciences); IL-17A PE-Cy7 (506921; BioLegend); and IL-10–allophycocyanin, GM-CSF–FITC, and/or TNF-α–PercP-eFluor710 (17-7101-81, 11-7331, 46-7321-80; Affymetrix). For Foxp3 staining, the Foxp3/Transcription Factor Staining Buffer Set Kit and protocol were used (00-5523-00; Affymetrix). Cells were analyzed by flow cytometry using a BD Biosciences LSR II. All data were obtained after gating on leukocytes, live cells, single cells, and nonautofluorescent cells. Remaining gating strategies are described in each figure.

Statistical analyses

Data are expressed as the mean ± SEs. Statistical testing included an ANOVA for comparisons between more than two normally distributed groups, Student t test for comparisons between two normally distributed groups, or Mann–Whitney U test for nonparametric analysis between two groups.

Results

TLR2 tolerance can be induced both in vitro and in vivo using a canonical TLR2 ligand or L654

We first confirmed that a canonical TLR2 ligand can tolerize the TLR2 response of mouse BMDMs. We also tested the ability of L654, a microbiome-derived TLR2 ligand that is capable of accessing the human circulation, to tolerize BMDMs (1). We opted to study BMDMs because we could not be sure whether macrophages, microglia, dendritic cells, B cells, or other cell types might be the most critical to (subsequently) tolerize in vivo so as to alter the disease course in EAE. We chose BMDMs in vitro as most likely reflecting the in vivo tolerance to macrophages, microglia, and also dendritic cells.

BMDMs were pretreated with Pam2Cys, L654, or VC for 24 h. The primary TLR2 stimulus was removed, the cells were washed, and a secondary TLR2 stimulus (Pam2Cys) was added for an additional 24 h. As TNF-α is a consistent readout for TLR tolerance induction (15), the concentration of TNF-α in the culture supernatants was then assayed by ELISA. The level of TNF-α was significantly decreased by both Pam2Cys and L654 pretreatment compared with pretreatment with VC (Fig. 1A, 1B).

FIGURE 1.
FIGURE 1.

TLR2 tolerance can be induced both in vitro and in vivo using a canonical TLR2 ligand or L654. (A) C57BL/6 BMDMs were treated with 0.5, 1, or 5 ng/ml Pam2Cys or a VC for 24 h. Cells were then cultured with a secondary stimulus of 5 ng/ml Pam2Cys for 24 h and supernatant TNF-α analyzed by ELISA; n = 4; ****p < 0.0001 by ANOVA. (B) BMDMs were treated with 1, 5, or 7 μg/ml L654 or VC for 24 h. Cells were then cultured with a secondary stimulus of 5 ng/ml Pam2Cys for 24 h and supernatant TNF-α analyzed by ELISA; n = 6; ****p < 0.0001 by ANOVA. (C) SJL/J mice were i.v. administered VC or 0.35 μg Pam2Cys once daily for 5 d. On day 6 all mice were i.v. administered 300 μg LTA and bled 2 h later, and serum TNF-α was analyzed by ELISA; n = 5–6 in each group; ****p < 0.0001 by Student t test. (D) SJL/J mice were i.v. administered VC, 5 μg L654, or 0.35 μg Pam2Cys once daily for 5 d. On day 6 all mice were i.v. administered 300 μg LTA and bled 2 h later, and serum TNF-α was analyzed by ELISA; n = 3–8 in each group; VC compared with 5 μg L654 or 0.35 μg Pam2Cys; *p = 0.0392, *p = 0.0377, respectively, by ANOVA. (E) SJL/J mice were i.v. administered VC or 0.35 μg Pam2Cys once daily for consecutive 5 d. On day 1, 3, or 5 after the final injection, the mice were i.v. administered 300 μg LTA and bled 2 h later, and serum TNF-α was analyzed by ELISA; n = 4–5 in each group; day 1, 3, and 5 VC compared with Pam2Cys: ****p < 0.0001, **p = 0.0045 by ANOVA.

With the goal of developing an in vivo TLR2-induced tolerance protocol that might mimic that induced by microbiome products accessing the circulation, we administered a low dose of either the canonical TLR2 ligand Pam2Cys, L654, or VC once daily for consecutive 5 d to SJL/J mice. On day 6, another TLR2 ligand, LTA, was administered and serum TNF-α was assayed by ELISA as a readout for systemic TLR tolerance induction (15). We determined that the lowest dose of Pam2Cys that efficiently induced TLR2 tolerance in vivo using our 5-d protocol was 0.35 μg/d (Fig. 1C, data not shown). Attempting to recapitulate the concentration of L654 recovered in human serum by administering L654 to mice is difficult for many reasons, including the fact that bacterially-derived L654 requires sonication for solubilization. Nevertheless, we found that the lowest dose of L654 that efficiently induced TLR2 tolerance in vivo using our 5-d protocol was 5 μg/d (Fig. 1D, data not shown). To determine the length of time that in vivo induced TLR2 tolerance persists, mice were administered VC or Pam2Cys once daily for consecutive 5 d. On day 1, 3, or 5 after the last injection of Pam2Cys, mice were administered LTA, and 2 h later serum was obtained and assayed for TNF-α. At 1, 3, and 5 d following the last injection of Pam2Cys, mice demonstrated statistically significant tolerance to LTA stimulation (Fig. 1E). This suggests that TLR2 tolerance persists for at least 5 d after Pam2Cys administration has ended.

Administration of low-dose Pam2Cys attenuates adoptively transferred EAE

To test the effect of TLR2 tolerance on adoptive transfer EAE, we used the well-characterized model of EAE induced by the adoptive transfer of PLP-activated LNCs (PLP-LNCs) into naive SJL/J mice (17). This model does not involve administration of exogenous TLR ligands as an adjuvant. Therefore, any TLR2 requirement demonstrated would suggest a role for TLR2 after the T cell–priming stage of disease and through interaction with endogenous or microbiome-derived TLR ligands.

We began by administering low-dose Pam2Cys (0.35 μg/d) from day −5 before transfer of PLP-LNCs through day +8 postcell transfer. Disease severity was recorded daily for 60 d after the adoptive transfer of the PLP-LNCs. We initially found that although 0.35 μg/d Pam2Cys was well tolerated by most mice for all 14 d, a small percentage of mice died of non-EAE causes. For this reason, we adjusted this longer protocol to include a 5-d period of 0.35 μg/d Pam2Cys to induce tolerance, followed by 4 d of even lower dose Pam2Cys (0.1 μg/d) to maintain tolerance, and then reinstituted the 0.35 μg/d Pam2Cys dose for the final 5 d. This protocol was well tolerated by all mice. The 14-d protocols resulted in very significant inhibition of disease (p < 0.0001; Fig. 2). These results suggest that a chronic exposure to low-dose microbiome products can very efficiently inhibit CNS autoimmune disease.

FIGURE 2.
FIGURE 2.

Administration of low-dose Pam2Cys attenuates adoptively transferred EAE. SJL/J mice were injected i.p. with 29 × 106 PLP139–151–reactive LNCs on day 0. EAE disease severity was recorded daily for each mouse and compiled as a daily mean disease score for each cohort: 1 = tail paralysis, 2 = unsteady gait, 3 = hind leg paralysis, 4 = front leg paralysis, and 5 = death. Injections of VC or Pam2Cys began at day −5 and continued through day +8 postcell transfer; two independent experiments were performed and compiled in the figure; mice received 0.35 μg Pam2Cys for all 14 d, or mice received 0.35 μg Pam2Cys for 5 d, 0.1 μg Pam2Cys for the next 4 d, and 0.35 μg Pam2Cys for the final 5 d; n = 6–10 per group; disease incidence was VC: 9/10 and P2C: 1/6; p < 0.0001 by Mann–Whitney U test.

Kinetics of TLR2 tolerance induction and function of L654 in the inhibition of EAE

To begin to determine when TLR2 played a role in the pathogenesis of adoptive transfer EAE, we used our 5-d tolerance induction protocol and initiated these 5-d treatments at various time points during adoptively transferred EAE development (Fig. 3A). Recipient mice were i.v. administered VC or low-dose Pam2Cys once daily for consecutive 5 d. Injections were given to recipient mice beginning at day −2 before PLP-LNC cell transfer, day +4 postcell transfer, or beginning at day +8 postcell transfer. Disease severity was recorded daily for 60 d.

FIGURE 3.
FIGURE 3.

Kinetics of TLR2 tolerance induction and function of L654 in the inhibition of EAE. Mice received PLP139–151–reactive LNCs as in Fig. 2. Mice received VC (black squares) or 0.35 μg Pam2Cys (gray squares) i.v. daily for consecutive 5 d. (A) VC or Pam2Cys injections were initiated at various time points before or after adoptive transfer of PLP-LNCs. (B) Injections of VC or Pam2Cys began at day −2 and continued for 5 d; two independent experiments performed, one representative experiment shown, n = 4 per group; disease incidence was VC: 3/4 and P2C: 4/4. (C) Injections of VC or Pam2Cys began at day +8 and continued for 5 d; two independent experiments performed, one representative experiment shown, n = 4 per group; disease incidence was VC: 3/4 and P2C: 4/4. (D) Injections of VC or Pam2Cys began at day +4 and continued for 5 d; two independent experiments performed, one representative experiment shown, n = 5 per group; disease incidence was VC: 4/5 and P2C: 2/5; p < 0.0001 by Mann–Whitney. (E) Injections of VC or 5 μg L654 began at day +4 and continued for 5 d; two independent experiments performed; one representative experiment shown, n = 5 per group; disease incidence was VC: 5/5 and P2C: 4/5; p < 0.0001 by Mann–Whitney U test. (F) Injections of VC or Pam2Cys were given on day +4, +5, and +6 posttransfer of the PLP-LNCs. Mice received 300 μg LTA i.v. on day +7, and serum TNF-α was analyzed by ELISA; two independent experiments performed, n = 6 in each group; **p = 0.0027 by Student t test.

Initiating the 5-d TLR2 tolerance protocol at day −2 or day +8 did not affect the time of disease onset or the overall disease severity (Fig. 3B, 3C). In contrast, the day +4 protocol resulted in a significant delay of disease onset that lasted ∼5 d (p < 0.0001; Fig. 3D). Interestingly, this approximate 5-d delay in onset of clinical disease using the day +4 tolerance induction protocol was consistent with our documented kinetics of tolerance persistence (Fig. 1E). In addition to this delay in onset, we also noted a significant attenuation in disease severity over the entire 60-d course in the mice treated using the day +4 protocol (Fig. 3D). Therefore, TLR2 seems to contribute to the pathogenesis of adoptive transfer EAE specifically during the day +4 through +8 postcell transfer window.

We suspect that the day +4 through day +8 time period is critical in that it most likely affects the phase of EAE pathogenesis that involves T cell entry and restimulation in the CNS. Interestingly, our laboratory has previously published a study using a SJL adoptive transfer model of EAE and labeled T cell lines (20). Tracking the transferred T cells in that study revealed that day 4 after adoptive transfer appeared to be the key time point at which encephalitogenic T cells were restimulated in the CNS. Thus, the kinetic window of tolerance that we have defined as relevant for TLR2 tolerance in the current study (day 4 through day 8) is consistent with our previous results. These kinetic studies suggest that using TLR2 tolerance to inhibit CNS autoimmune disease will be most effective when initiated during disease remissions in relapsing-remitting MS. Importantly, evidence for a role of TLR2 signaling in the remyelination defect seen in progressive forms of MS (3) suggests that our approach may also be effective in treating these as yet untreatable forms of MS.

We next tested the concept that L654 might normally be contributing to immune homeostasis and prevention of human CNS autoimmune disease through TLR2 tolerance induction. We asked whether L654 was capable of attenuating EAE when administered using our 5-d tolerance induction protocol during the day +4 through day +8 window. Similar to Pam2Cys, L654 was able to both delay disease onset and attenuate the severity of the chronic phase of the disease (p < 0.0001; Fig. 3E). We found that L654 usually resulted in a shorter delay in disease onset compared with that seen using Pam2Cys, and this is consistent with the relative strength of TLR2 tolerance induced by these ligands (Fig. 1D). These results suggest that L654, as a microbiome-derived molecule that accesses the human circulation, may normally aid in regulating TLR2 responses through low-level tonic signaling and tolerance induction. In this paradigm, the decreased recovery of L654 in MS serum samples may indicate a deficiency in this tonic TLR regulatory function in MS.

Finally, to confirm that our protocol was inducing TLR2 tolerance in the context of EAE, LTA was administered on day 7 to mice that had been treated with either VC or Pam2Cys from day 4 through day 6 after PLP-LNC transfer. Mice treated with Pam2Cys demonstrated a significant reduction in serum TNF-α in response to LTA (Fig. 3F). This indicates that our protocol indeed induces systemic TLR2 tolerance in the context of the developing adoptively transferred EAE.

Induction of TLR2 tolerance cross-tolerizes to the other TLRs implicated in EAE pathogenesis

It has been documented that inducing tolerance to a specific TLR can also induce cross-tolerance to other TLRs (21, 22). Therefore, we asked whether our 5-d TLR2 tolerance protocol induces cross-tolerance to signaling through other TLR molecules. A number of TLRs, including TLR9, TLR7, and TLR4, have been suggested to be important in the pathogenesis of EAE (6, 8, 23–25). Because there has been much evidence that TLR2 and TLR4 usually cross-tolerize each other (22, 26), but less documentation that TLR2 can cross-tolerize TLR7 and TLR9, we specifically asked whether our TLR2 tolerance induction protocol cross-tolerizes mice to TLR7 and TLR9.

We administered VC or low-dose Pam2Cys to naive SJL mice for consecutive 5 d and, on day 6 we administered either ODN 1668, a TLR9 ligand, or R848, a TLR7/8 ligand. Mice pretreated with the 5-d Pam2Cys protocol demonstrated a significantly decreased serum TNF-α response to both ODN 1668 and R848 (Fig. 4). This indicates that our TLR2 tolerance protocol induces cross-tolerance to other TLRs of potential relevance in EAE and suggests that we cannot implicate tolerance only through TLR2 as underlying the disease attenuation we have noted. Nevertheless, our results confirm a role for TLR signaling in the pathogenesis of adoptive transfer EAE and, to our knowledge, for the first time identify a specific time interval during disease development in which this function is most relevant. Moreover, our results suggest that inducing tolerance with even a single TLR2 ligand may be an effective approach for attenuating CNS autoimmune disease.

FIGURE 4.
FIGURE 4.

Induction of TLR2 tolerance cross-tolerizes to the other TLRs implicated in EAE pathogenesis. Naive SJL/J mice were administered VC or 0.35 μg Pam2Cys i.v. for 5 d. (A) On day 6, 40 μg ODN 1668 (a TLR9 ligand) was administered i.v., and serum TNF-α was assayed 1.5 h later; n = 6; ****p < 0.0001 by Student t test. (B) On day 6, 60–80 μg R848 (a TLR7/8 ligand) was administered i.p., and serum TNF-α was assayed 1.5 h later; n = 5–6; ***p = 0.0003 by Student t test.

Regulatory effects of TLR2 tolerance induction are observed in the CNS

To determine the disease-relevant immune effector and regulatory alterations associated with induction of TLR2 tolerance in EAE, recipient mice were i.v. administered VC or Pam2Cys for 5 d on day +4 through +8 postcell transfer. Mice were then sacrificed on day +10 or +12, and spleen and spinal cord mononuclear cells were harvested and prepared for flow cytometric analysis using standard approaches. At days +10 and +12, the tolerized mice showed either no clinical signs or had clinical signs that were significantly less than those seen in the VC-treated cohort.

We found that the tolerized cohort demonstrated, on average, fewer mononuclear cells and fewer CD4+ T cells infiltrating the spinal cord compared with control mice (data not shown). Most importantly, in tolerized mice we noted significant differences in the frequencies of three CNS cellular populations compared with the control mice. These differences were also reflected in the absolute numbers of these cell populations.

First, the tolerized mice demonstrated a significant decrease in the frequency of CNS F4/80+ macrophages expressing the costimulatory molecule CD80 (*p = 0.0248; Fig. 5A). A decrease in CD80-expressing F4/80+ macrophages is indicative of diminished activation of the macrophage population and is consistent with what has been shown in previous studies of TLR tolerance (27, 28). A state of diminished activation of F4/80+ macrophages is associated with decreased Ag-presentation ability and thus is most likely an important mechanism underlying the decreased spinal cord cellular inflammation and decreased disease manifestations seen in the tolerized mice. Specifically, we postulate that the delay in disease onset we have noted is a result of a diminished activation state of CNS macrophages that are involved in reactivating pathogenic T cells that have trafficked into the CNS during the day +4 through day +8 interval.

FIGURE 5.
FIGURE 5.

Regulatory effects of TLR2 tolerance induction are observed in the CNS. SJL/J mice were injected with PLP-LNCs as in Fig. 2 and then i.v. administered VC or 0.35 μg Pam2Cys daily for consecutive 5 d starting on day +4 postcell transfer. Mice were sacrificed on day +10 or +12 postcell transfer, and their spleens and spinal cords were harvested and prepared for flow cytometric analysis. (A) Percentage of spinal cord F4/80+ cells that are positive for CD80 expression; n = 9; *p = 0.0248 by Student t test. (B) Percentage of spinal cord mononuclear cells that are F4/80+CD11c+; n = 13; ****p < 0.0001 by Student t test. (C) Percentage of spinal cord CD4+ T cells that are positive for IL-17 (Th17 cells); n = 7–8; **p = 0.0068 by Student t test. (D) Percentage of spinal cord CD4+ T cells that are positive for IFN-γ (Th1 cells); n = 7–8; p = 0.8223 by Student t test. (E) Percentage of spinal cord CD4+ T cells that are positive for both IL-17 and IFN-γ; n = 7–8; p = 0.0989 by Student t test. (F) Percentage of spinal cord CD4+ T cells that are Foxp3+; n = 10; p = 0.4770 by Student t test.

Second, tolerized mice demonstrated a significant decrease in the frequency of CNS F4/80+CD11c+ cells (****p < 0.0001; Fig. 5B). The role of CNS F4/80+CD11c+ cells in EAE has not been reported, but in other contexts cells with this phenotype have been reported to serve either an inflammatory or regulatory function (29–31). We found these cells exhibited a proinflammatory phenotype in both control and tolerized cohorts as evidenced by their production of TNF-α (via intracellular staining; data not shown). Thus, we believe that this finding represents a decrease in a CNS proinflammatory monocytic population of potential pathogenic significance in EAE. Interestingly, proinflammatory F4/80+CD11c+ cells have been shown to infiltrate hypertrophic adipose tissue and to be activated through TLR2 and TLR4 (31). Thus, the decrease in this population that we have noted may be a direct effect of TLR2 tolerance induction.

Third, the tolerized mice demonstrated a significant decrease in the frequency of CNS CD4+ T cells that were Th17 cells (**p = 0068; Fig. 5C). Interestingly, this decrease was not observed in CNS Th1 cells (Fig. 5D), in GM-CSF–producing T cells, or in T cells that produced both IL-17 and IFN-γ (data not shown and Fig. 5E). Similarly, no statistically significant difference was observed in the percentage of CNS CD4+Foxp3+ T cells (Fig. 5F) or other regulatory cell types discussed below (data not shown). CD4+ Th17 T cells are believed to be the most pathogenic subset of CD4+ T cells in EAE, so finding a decrease in this subset is consistent with tolerized mice having significantly less disease than the control mice. Reynolds et al. (9) have previously demonstrated that optimal Th17 T cell generation is dependent on TLR2 expression. Thus, it is possible that the decrease in CNS Th17 T cells noted in the tolerized mice is a direct result of TLR2 tolerance induced in the encephalitogenic T cells in the transferred PLP-LNC population. However, these T cells have been activated initially in vivo and then again for 4 d in vitro. Therefore, it is likely that most of the differentiation of the PLP-specific CD4+ T cells into Th1 or Th17 subsets occurs before they are exposed to the in vivo tolerance-inducing protocol initiated 4 d after transfer. However, it is conceivable that inducing TLR2 tolerance in the adoptively transferred Th17 cells has effects on their migration or survival in the recipient mice that have not yet been reported.

Finally, it should be noted that one mechanism often reported to be involved in TLR tolerance is the production of IL-10 (15). We approached the relevance of IL-10 in our model by performing intracellular staining for IL-10 in T cells, macrophages, dendritic cells, and B cells derived from the spinal cords (and spleens, see below). Using this approach, we did not identify enhanced intracellular IL-10 in any cells derived from the spinal cords.

Regulatory effects of TLR2 tolerance induction are observed in the spleen

Previous reports in which TLR2 ligands were administered or TLR2 was genetically deleted in mice with EAE showed an increase in CD39-expressing, CD62L-expressing, or IL-10–producing Tregs (8, 13, 14). It should be noted that in these studies significantly larger concentrations of TLR2 ligands were administered (100 μg) than were used in our tolerance studies (0.35 μg Pam2Cys and 5 μg L654). Nonetheless, we asked whether Tregs may also be contributing to the attenuation of EAE seen in TLR2-tolerized mice. We found no significant increase in the percentages of Foxp3+CD4+ T cells, CD39+Foxp3+CD4+ T cells, CD39+CD4+ T cells, or IL-10–producing CD4+ T cells in the spleens of tolerized versus control mice (Fig. 6A–C and data not shown). Similarly, no significant increase was seen in the CD62L+Foxp3+CD4+ populations in the spleens of tolerized mice (Fig. 6D). However, we did observe a significant increase in the percentage of CD4+ Foxp3− IFN-γ+ IL-10+ type 1 regulatory T cells (Tr1) in the spleens of the tolerized mice (p = 0.0056; Fig. 6E). This difference was not seen in the spinal cords of these mice (data not shown). Thus, our tolerance protocol did not recapitulate an increase in the specific CD39+ or CD62L+ Treg populations reported previously, but did result in an increase in the population of splenic Tr1 Tregs. Given that Th17 T cells have been shown to be most efficiently regulated by IL-10–producing T cells such as Tr1 Tregs (32), it is possible that this increase in Tr1 Tregs is at least partially responsible for both the decreased percentage of CNS Th17 T cells and the attenuation of chronic disease severity observed in tolerized mice.

FIGURE 6.
FIGURE 6.

Regulatory effects of TLR2 tolerance induction are observed in the spleen. SJL/J mice were injected with PLP-LNCs as in Fig. 2 and then i.v. administered VC or 0.35 μg Pam2Cys daily for consecutive 5 d starting on day +4 postcell transfer. Mice were sacrificed on day +10 or +12 postcell transfer, and their spleens and spinal cords were harvested and prepared for flow cytometric analysis. (A) Percentage of spleen CD4+ T cells that are Foxp3+; n = 10 per group; p = 0.44 by Student t test. (B) Percentage of spleen CD4+Foxp3+ T cells that are CD39+; n = 10 per group; p = 0.001 by Student t test. (C) Percentage of spleen CD4+ T cells that are CD39+; n = 10 per group; p = 0.1228 by Student t test. (D) Percentage of spleen CD4+Foxp3+ T cells are CD62L+; n = 10 per group; p = 0.6081 by Student t test. (E) Percentage of spleen CD4+ T cells are positive for both IFN-γ and IL-10 (Tr1 cells); n = 13–15 per group; **p = 0.0056 by Student t test.

Discussion

We have previously reported that L654, a microbiome-derived TLR2 ligand, can be identified in the serum of all healthy individuals, but is significantly lower in the serum of MS patients (1). L654 stimulates cells through TLR2, does not stimulate through TLR4, and fails to stimulate in the absence of TLR2 (16). These findings led us to ask why a TLR ligand, normally thought to be proinflammatory, would be lower in the setting of an autoimmune disease such as MS. The concept of TLR tolerance offered one possible mechanism for the association of a lower level of TLR ligands with an autoinflammatory disease.

We postulated that microbiome-derived products, like L654, which gain access to the systemic circulation, may normally regulate the systemic immune system by inducing a state of relative TLR tolerance. As such, when circulating levels of microbiome-derived products are deficient, the normal induction of TLR tolerance may be inadequate, resulting in a lower threshold of initiation and higher magnitude of responses in innate immune cells. This may then contribute to autoinflammatory diseases such as MS. Consistent with this postulate, the relevance of TLR tolerance as a mechanism of immune regulation in human disease has been reported in the context of both asthma and Crohn’s disease (33, 34).

High levels of environmental endotoxin have been demonstrated to have a protective effect against the development of allergy and asthma in both farming and nonfarming households. Schuijs et al. (34) recently reported that chronic exposure to low-dose endotoxin or farm dust protects mice from developing house dust mite–induced asthma. These authors provided evidence that these environmental protective factors influenced the threshold for allergen recognition by suppressing the activation of epithelial and dendritic cells. Furthermore, loss of the TLR regulatory molecule A20 in lung epithelium abolished this protective effect. To validate their findings in humans, they found that a polymorphism in the A20 gene was associated with an increased risk for asthma. Furthermore, the protective effect that growing up on a farm exerted on the risk for asthma was much weaker in children with a polymorphism affecting a functional domain in the A20 molecule (34). Similarly, a defect in TLR tolerance induction has been suggested as underlying the association of polymorphisms in NOD2 with Crohn’s disease. In this mechanism, the normal NOD2 induction of cross-tolerance to TLR2 and TLR4 is defective in the subset of Crohn’s patients with these NOD2 polymorphisms, resulting in chronic enhanced inflammatory responses in the gastrointestinal tract (33). These studies emphasize the relevance of TLR tolerance as a mechanism regulating the development of human inflammatory diseases.

Although T cells and the adaptive immune system have been the major research focus in EAE and MS, the role of TLRs has not been extensively studied in these diseases. In studies that use TLR2-deficient mice, a role for TLR2 has been demonstrated, albeit inconsistently, in both actively induced and passive models of EAE (7–10). In actively induced EAE, adjuvants containing numerous TLR ligands are required. However, in most T cell adoptive transfer models, no exogenous TLR ligands are administered. Therefore, it is surprising that a TLR2 dependence has been demonstrated in these adoptive transfer models (8). Additionally, TLR2-deficient mice have been shown to have fewer systemic Th17 cells, a T cell subset known to contribute to EAE disease pathogenesis (9). Overall, these studies, as well as our present results, suggest that either endogenous damage-associated molecular patterns or microbiome-derived TLR ligands are playing an as yet unidentified role in the pathogenesis of adoptive transfer EAE and, by extension, in MS.

In contrast to the disease-promoting effect of TLR2 in EAE, other studies have shown that TLR2 ligands can inhibit EAE. The TLR2 ligands, polysaccharide A and zymosan, when administered to mice during EAE (in the range of ∼100 μg per day), significantly reduced disease severity via induction of various subsets of Tregs (11–14). In the present studies, we postulated that TLR tolerance could explain the TLR2 paradox in EAE as well as offer a potential mechanism for microbiome-mediated immune regulation. Our study differs from the polysaccharide A and zymosan studies (11–14) in that we show TLR tolerance induction using a very low dose of a canonical TLR2 ligand and the microbiome-derived TLR ligand known to access human serum, L654, attenuate EAE. Additionally, our study differs in our sole use of adoptive transfer EAE and in the effector and regulatory alterations we found associated with TLR tolerance-induced disease attenuation. Although we did not find the induction of the specific Treg populations described in those studies, we instead found alterations in CNS innate immune cell populations, a specific decrease in CNS Th17 T cells, and an enhancement in splenic Tr1 Tregs in the tolerized mice.

TLR tolerance in autoimmunity, although investigated in a murine model of type 1 diabetes (28), has rarely been studied in the context of either TLR2 or EAE (23, 28). In the only study examining this concept in EAE, tolerance induction using TLR7 ligands was shown to attenuate active immunization (i.e., not adoptive transfer) EAE models (23). However, in that study, it was also demonstrated that the TLR7 ligands cross-tolerized to TLR2. In the current study, our goal was to test, as proof of concept, TLR tolerance as a mechanism underlying microbiome-mediated immune regulation. We focused on testing this concept in a model for MS and utilizing adoptive transfer EAE in which no exogenous TLR ligands are required. Specifically, our goal was to test this mode of microbiome-mediated immune regulation using a very low dose of both a canonical TLR2 ligand (Pam2Cys) and a documented microbiome-derived TLR2 ligand, L654.

Our results now demonstrate the following: 1) low concentrations of the canonical TLR2 ligand Pam2Cys can induce in vivo TLR2 tolerance, and in doing so attenuate adoptive transfer EAE; 2) similarly, a microbiome-derived molecule, L654, recovered in the serum of all healthy individuals, but found in significantly lower levels in the serum of patients with MS, can also function both to tolerize mice to TLR2 and attenuate EAE; 3) this TLR tolerance approach clearly identifies a requirement for TLR signaling in the pathogenesis of adoptive transfer EAE; 4) there is a critical kinetic window after encephalitogenic T cell transfer during which the induction of TLR2 tolerance is most efficient in attenuating EAE; 5) mechanistically, TLR2 tolerance and attenuated EAE are associated with significant decreases in CD80-expressing CNS macrophages, proinflammatory CNS F4/80+CD11c+ macrophages, and CNS-infiltrating Th17 T cells, as well as a significant increase in splenic Tr1 Tregs; 6) induction of TLR2 tolerance in vivo cross-tolerizes to other relevant TLRs, which most likely significantly enhances the efficiency of this mode of immune regulation as a potential new therapeutic approach for MS.

Finally, our tolerance approach utilizes a very low dose of TLR ligands, thereby not only offering a potentially safe treatment, but also theoretically mimicking and enhancing a natural function of the microbiome. We believe that L654 is most likely just one example of what is potentially many other microbiome-derived TLR tolerance-inducing molecules. This low-level administration of TLR ligands theoretically allows for the regulatory effects of TLR tolerance while avoiding the proinflammatory effects. This study should now provide impetus for further investigation of the relevance of TLR tolerance induction as a potential new therapeutic approach for MS and other autoimmune diseases and as a mechanism underlying the normal microbiome regulation of systemic immune responses.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Dr. Andrei Medvedev for expert advice and suggestions regarding TLR tolerance and Drs. Michael Murphy, Goutham Pattabiraman, and Tissa Manavalan for expert technical assistance.

Footnotes

  • This work was supported by National Multiple Sclerosis Society Grant RG4070B7/2 (to R.B.C.).

  • Abbreviations used in this article:

    BMDM
    bone marrow–derived macrophage
    EAE
    experimental autoimmune encephalomyelitis
    L654
    Lipid 654
    LNC
    lymph node cell
    LTA
    lipoteichoic acid
    MS
    multiple sclerosis
    ODN
    oligodeoxynucleotide
    PLP
    proteolipid protein
    Tr1
    type 1 regulatory T cell
    Treg
    regulatory T cell
    VC
    vehicle control.

  • Received May 20, 2016.
  • Accepted July 10, 2016.

References

  1. ↵
    1. Farrokhi V.,
    2. R. Nemati,
    3. F. C. Nichols,
    4. X. Yao,
    5. E. Anstadt,
    6. M. Fujiwara,
    7. J. Grady,
    8. D. Wakefield,
    9. W. Castro,
    10. J. Donaldson,
    11. R. B. Clark
    . 2013. Bacterial lipodipeptide, Lipid 654, is a microbiome-associated biomarker for multiple sclerosis. Clin. Transl. Immunology 2: e8.
    OpenUrlCrossRef
  2. ↵
    1. Hasheminia S. J.,
    2. S. H. Zarkesh-Esfahani,
    3. S. Tolouei,
    4. V. Shaygannejad,
    5. H. Shirzad,
    6. M. Hashemzadeh Chaleshtory
    . 2014. Toll like receptor 2 and 4 expression in peripheral blood mononuclear cells of multiple sclerosis patients. Iran. J. Immunol. 11: 74–83.
    OpenUrlPubMed
  3. ↵
    1. Sloane J. A.,
    2. C. Batt,
    3. Y. Ma,
    4. Z. M. Harris,
    5. B. Trapp,
    6. T. Vartanian
    . 2010. Hyaluronan blocks oligodendrocyte progenitor maturation and remyelination through TLR2. Proc. Natl. Acad. Sci. USA 107: 11555–11560.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Correale J.,
    2. M. F. Farez
    . 2013. Parasite infections in multiple sclerosis modulate immune responses through a retinoic acid-dependent pathway. J. Immunol. 191: 3827–3837.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Nyirenda M. H.,
    2. E. Morandi,
    3. U. Vinkemeier,
    4. D. Constantin-Teodosiu,
    5. S. Drinkwater,
    6. M. Mee,
    7. L. King,
    8. G. Podda,
    9. G. X. Zhang,
    10. A. Ghaemmaghami,
    11. et al
    . 2015. TLR2 stimulation regulates the balance between regulatory T cell and Th17 function: a novel mechanism of reduced regulatory T cell function in multiple sclerosis. J. Immunol. 194: 5761–5774.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Prinz M.,
    2. F. Garbe,
    3. H. Schmidt,
    4. A. Mildner,
    5. I. Gutcher,
    6. K. Wolter,
    7. M. Piesche,
    8. R. Schroers,
    9. E. Weiss,
    10. C. J. Kirschning,
    11. et al
    . 2006. Innate immunity mediated by TLR9 modulates pathogenicity in an animal model of multiple sclerosis. J. Clin. Invest. 116: 456–464.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Herrmann I.,
    2. M. Kellert,
    3. H. Schmidt,
    4. A. Mildner,
    5. U. K. Hanisch,
    6. W. Brück,
    7. M. Prinz,
    8. R. Nau
    . 2006. Streptococcus pneumoniae infection aggravates experimental autoimmune encephalomyelitis via Toll-like receptor 2. Infect. Immun. 74: 4841–4848.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Miranda-Hernandez S.,
    2. N. Gerlach,
    3. J. M. Fletcher,
    4. E. Biros,
    5. M. Mack,
    6. H. Korner,
    7. A. G. Baxter
    . 2011. Role for MyD88, TLR2 and TLR9 but not TLR1, TLR4 or TLR6 in experimental autoimmune encephalomyelitis. J. Immunol. 187: 791–804.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Reynolds J. M.,
    2. B. P. Pappu,
    3. J. Peng,
    4. G. J. Martinez,
    5. Y. Zhang,
    6. Y. Chung,
    7. L. Ma,
    8. X. O. Yang,
    9. R. I. Nurieva,
    10. Q. Tian,
    11. C. Dong
    . 2010. Toll-like receptor 2 signaling in CD4(+) T lymphocytes promotes T helper 17 responses and regulates the pathogenesis of autoimmune disease. Immunity 32: 692–702.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Shaw P. J.,
    2. M. J. Barr,
    3. J. R. Lukens,
    4. M. A. McGargill,
    5. H. Chi,
    6. T. W. Mak,
    7. T. D. Kanneganti
    . 2011. Signaling via the RIP2 adaptor protein in central nervous system-infiltrating dendritic cells promotes inflammation and autoimmunity. Immunity 34: 75–84.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Li H.,
    2. P. Gonnella,
    3. F. Safavi,
    4. G. Vessal,
    5. B. Nourbakhsh,
    6. F. Zhou,
    7. G. X. Zhang,
    8. A. Rostami
    . 2013. Low dose zymosan ameliorates both chronic and relapsing experimental autoimmune encephalomyelitis. J. Neuroimmunol. 254: 28–38.
    OpenUrlCrossRefPubMed
    1. Manicassamy S.,
    2. R. Ravindran,
    3. J. Deng,
    4. H. Oluoch,
    5. T. L. Denning,
    6. S. P. Kasturi,
    7. K. M. Rosenthal,
    8. B. D. Evavold,
    9. B. Pulendran
    . 2009. Toll-like receptor 2-dependent induction of vitamin A-metabolizing enzymes in dendritic cells promotes T regulatory responses and inhibits autoimmunity. Nat. Med. 15: 401–409.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Manoharan I.,
    2. Y. Hong,
    3. A. Suryawanshi,
    4. M. L. Angus-Hill,
    5. Z. Sun,
    6. A. L. Mellor,
    7. D. H. Munn,
    8. S. Manicassamy
    . 2014. TLR2-dependent activation of beta-catenin pathway in dendritic cells induces regulatory responses and attenuates autoimmune inflammation. J. Immunol. 193: 4203–4213.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Wang Y.,
    2. K. M. Telesford,
    3. J. Ochoa-Repáraz,
    4. S. Haque-Begum,
    5. M. Christy,
    6. E. J. Kasper,
    7. L. Wang,
    8. Y. Wu,
    9. S. C. Robson,
    10. D. L. Kasper,
    11. L. H. Kasper
    . 2014. An intestinal commensal symbiosis factor controls neuroinflammation via TLR2-mediated CD39 signalling. Nat. Commun. 5: 4432.
    OpenUrlPubMed
  15. ↵
    1. Biswas S. K.,
    2. E. Lopez-Collazo
    . 2009. Endotoxin tolerance: new mechanisms, molecules and clinical significance. Trends Immunol. 30: 475–487.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Clark R. B.,
    2. J. L. Cervantes,
    3. M. W. Maciejewski,
    4. V. Farrokhi,
    5. R. Nemati,
    6. X. Yao,
    7. E. Anstadt,
    8. M. Fujiwara,
    9. K. T. Wright,
    10. C. Riddle,
    11. et al
    . 2013. Serine lipids of Porphyromonas gingivalis are human and mouse Toll-like receptor 2 ligands. Infect. Immun. 81: 3479–3489.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Brocke S.,
    2. C. Piercy,
    3. L. Steinman,
    4. I. L. Weissman,
    5. T. Veromaa
    . 1999. Antibodies to CD44 and integrin alpha4, but not L-selectin, prevent central nervous system inflammation and experimental encephalomyelitis by blocking secondary leukocyte recruitment. Proc. Natl. Acad. Sci. USA 96: 6896–6901.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Adams C. O.,
    2. W. J. Housley,
    3. S. Bhowmick,
    4. R. E. Cone,
    5. T. V. Rajan,
    6. F. Forouhar,
    7. R. B. Clark
    . 2010. Cbl-b(−/−) T cells demonstrate in vivo resistance to regulatory T cells but a context-dependent resistance to TGF-beta. J. Immunol. 185: 2051–2058.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Korn T.,
    2. J. Reddy,
    3. W. Gao,
    4. E. Bettelli,
    5. A. Awasthi,
    6. T. R. Petersen,
    7. B. T. Bäckström,
    8. R. A. Sobel,
    9. K. W. Wucherpfennig,
    10. T. B. Strom,
    11. et al
    . 2007. Myelin-specific regulatory T cells accumulate in the CNS but fail to control autoimmune inflammation. Nat. Med. 13: 423–431.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Barbarese E.,
    2. H. Soares,
    3. S. Yang,
    4. R. B. Clark
    . 1992. Comparison of CNS homing pattern among murine TH cell lines responsive to myelin basic protein. J. Neuroimmunol. 39: 151–162.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Lehner M. D.,
    2. S. Morath,
    3. K. S. Michelsen,
    4. R. R. Schumann,
    5. T. Hartung
    . 2001. Induction of cross-tolerance by lipopolysaccharide and highly purified lipoteichoic acid via different Toll-like receptors independent of paracrine mediators. J. Immunol. 166: 5161–5167.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Sato S.,
    2. F. Nomura,
    3. T. Kawai,
    4. O. Takeuchi,
    5. P. F. Muhlradt,
    6. K. Takeda,
    7. S. Akira
    . 2000. Synergy and cross-tolerance between Toll-like receptor (TLR) 2- and TLR4-mediated signaling pathways. J. Immunol. 165: 7096–7101.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Hayashi T.,
    2. C. S. Gray,
    3. M. Chan,
    4. R. I. Tawatao,
    5. L. Ronacher,
    6. M. A. McGargill,
    7. S. K. Datta,
    8. D. A. Carson,
    9. M. Corr
    . 2009. Prevention of autoimmune disease by induction of tolerance to Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 106: 2764–2769.
    OpenUrlAbstract/FREE Full Text
    1. Park J. H.,
    2. S. Y. Jeong,
    3. A. J. Choi,
    4. S. J. Kim
    . 2015. Lipopolysaccharide directly stimulates Th17 differentiation in vitro modulating phosphorylation of RelB and NF-κB1. Immunol. Lett. 165: 10–19.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Reynolds J. M.,
    2. G. J. Martinez,
    3. Y. Chung,
    4. C. Dong
    . 2012. Toll-like receptor 4 signaling in T cells promotes autoimmune inflammation. Proc. Natl. Acad. Sci. USA 109: 13064–13069.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Diterich I.,
    2. C. Rauter,
    3. C. J. Kirschning,
    4. T. Hartung
    . 2003. Borrelia burgdorferi-induced tolerance as a model of persistence via immunosuppression. Infect. Immun. 71: 3979–3987.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Cole T. S.,
    2. M. Zhang,
    3. T. J. Standiford,
    4. M. Newstead,
    5. J. Luther,
    6. J. Zhang,
    7. C. C. Chen,
    8. J. Y. Kao
    . 2012. IRAK-M modulates expression of IL-10 and cell surface markers CD80 and MHC II after bacterial re-stimulation of tolerized dendritic cells. Immunol. Lett. 144: 49–59.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Kim D. H.,
    2. J. C. Lee,
    3. S. Kim,
    4. S. H. Oh,
    5. M. K. Lee,
    6. K. W. Kim,
    7. M. S. Lee
    . 2011. Inhibition of autoimmune diabetes by TLR2 tolerance. J. Immunol. 187: 5211–5220.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Kokubu Y.,
    2. K. Tabu,
    3. N. Muramatsu,
    4. W. Wang,
    5. Y. Murota,
    6. I. Nobuhisa,
    7. M. Jinushi,
    8. T. Taga
    . 2016. Induction of protumoral CD11c(high) macrophages by glioma cancer stem cells through GM-CSF. Genes Cells 21: 241–251.
    OpenUrlCrossRefPubMed
    1. McNelis J. C.,
    2. J. M. Olefsky
    . 2014. Macrophages, immunity, and metabolic disease. Immunity 41: 36–48.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Nguyen M. T.,
    2. S. Favelyukis,
    3. A. K. Nguyen,
    4. D. Reichart,
    5. P. A. Scott,
    6. A. Jenn,
    7. R. Liu-Bryan,
    8. C. K. Glass,
    9. J. G. Neels,
    10. J. M. Olefsky
    . 2007. A subpopulation of macrophages infiltrates hypertrophic adipose tissue and is activated by free fatty acids via Toll-like receptors 2 and 4 and JNK-dependent pathways. J. Biol. Chem. 282: 35279–35292.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Huber S.,
    2. N. Gagliani,
    3. E. Esplugues,
    4. W. O’Connor Jr..,
    5. F. J. Huber,
    6. A. Chaudhry,
    7. M. Kamanaka,
    8. Y. Kobayashi,
    9. C. J. Booth,
    10. A. Y. Rudensky,
    11. et al
    . 2011. Th17 cells express interleukin-10 receptor and are controlled by Foxp3⁻ and Foxp3+ regulatory CD4+ T cells in an interleukin-10-dependent manner. Immunity 34: 554–565.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Hedl M.,
    2. J. Li,
    3. J. H. Cho,
    4. C. Abraham
    . 2007. Chronic stimulation of Nod2 mediates tolerance to bacterial products. Proc. Natl. Acad. Sci. USA 104: 19440–19445.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Schuijs M. J.,
    2. M. A. Willart,
    3. K. Vergote,
    4. D. Gras,
    5. K. Deswarte,
    6. M. J. Ege,
    7. F. B. Madeira,
    8. R. Beyaert,
    9. G. van Loo,
    10. F. Bracher,
    11. et al
    . 2015. Farm dust and endotoxin protect against allergy through A20 induction in lung epithelial cells. Science 349: 1106–1110.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section