IL-4 and IL-13 have been defined as anti-inflammatory cytokines that can counter myelin-reactive T cells and modulate experimental allergic encephalomyelitis. However, it is not known whether endogenous IL-4 and IL-13 contribute to the maintenance of peripheral tolerance and whether their function is coordinated with T regulatory cells (Tregs). In this study, we used mice in which the common cytokine receptor for IL-4 and IL-13, namely the IL-4Rα/IL-13Rα1 (13R) heteroreceptor (HR), is compromised and determined whether the lack of signaling by endogenous IL-4 and IL-13 through the HR influences the function of effector Th1 and Th17 cells in a Treg-dependent fashion. The findings indicate that mice-deficient for the HR (13R−/−) are more susceptible to experimental allergic encephalomyelitis than mice sufficient for the HR (13R+/+) and develop early onset and more severe disease. Moreover, Th17 cells from 13R−/− mice had reduced ability to convert to Th1 cells and displayed reduced sensitivity to suppression by Tregs relative to Th17 effectors from 13R+/+ mice. These observations suggest that IL-4 and IL-13 likely operate through the HR and influence Th17 cells to convert to Th1 cells and to acquire increased sensitivity to suppression, leading to control of immune-mediated CNS inflammation. These previously unrecognized findings shed light on the intricacies underlying the contribution of cytokines to peripheral tolerance and control of autoimmunity.
Autoimmunity develops when peripheral tolerance (1) is no longer able to keep self-reactive lymphocytes in check (2). T regulatory cells (Tregs) and anti-inflammatory cytokines are usually adept at containing aggressive lymphocytes and prevent the development of autoimmune diseases. However, whether these forms of tolerance coordinate their function and synergize their action against autoreactive lymphocytes has yet to be determined. IL-4 and IL-13 function as anti-inflammatory cytokines (3–7) and may serve alongside Tregs to preserve peripheral tolerance and prevent autoimmunity. In fact, we have previously shown that neonatal exposure to self-antigen, which induces responses dominated by IL-4–producing Th2 cells, confers resistance to experimental allergic encephalomyelitis (EAE) (8, 9). Alternatively, Tregs play a major role in keeping myelin-reactive T cells in check and preventing the development of EAE (10–12). In this study, we asked whether and how endogenous IL-4 and IL-13 synergize with Tregs to restrain myelin-reactive T cells and prevent the development of EAE.
IL-4 and IL-13 share the IL-4Rα/IL-13Rα1 (13R) heteroreceptor (HR) (13) and most likely carry out their anti-inflammatory function through its expression on APCs such as dendritic cells and macrophages, as T cells in adult mice do not express this receptor (14–16). Also, IL-4 does not signal through the conventional IL-4R (IL-4Rα/common γ-chain) in Th1 cells (17), and the conventional IL-13 receptor (IL-13Rα1/IL-13Rα2) serves rather as a decoy receptor (18). Thus, mice lacking IL-13Rα1 in which the conventional IL-4R is intact but the HR does not form (19–21) provide a suitable model to determine whether anti-inflammatory IL-4/IL-13 synergize with Tregs to maintain peripheral tolerance and contain EAE. This was indeed the case, as IL-13Rα1–deficient (13R−/−) mice which lack the HR (HR−/−) are more susceptible to EAE relative to 13R+/+ wild-type mice. Specifically, 13R−/− mice develop early onset and severe EAE when induced for disease with myelin oligodendrocyte glycoprotein 35–55 peptide (MOGp). This phenotype has been correlated with effects on Th17 to Th1 conversion (22) as well as interference with the sensitivity of these effectors to suppression by Tregs. Indeed, there was limited Th17 to Th1 conversion in 13R−/− mice relative to 13R+/+ animals. Also, although there was no effect on the development of Tregs in 13R−/− mice, both Th1 and Th17 cells displayed differential sensitivity to suppression by Tregs when compared with counterparts from 13R+/+ mice. These findings indicate that endogenous IL-4/IL-13 cytokines synergize their function with Tregs to control peripheral tolerance and restrain autoimmunity.
Materials and Methods
C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). IL-13Rα1−/− C57BL/6 (13R−/−) mice were previously described (19). IL-17acre mice obtained from Dr. B. Stockinger (The Francis Crick Institute) were also previously described (22). To generate 13R−/− IL-17acre–enhanced yellow fluorescent protein (eYFP) mice, we first bred the IL-17acre mice with the 13R−/− mice and then to B6.129 × 1-Gt(ROSA)26Sortm1(EYFP)Cos/J(ROSA26-YFP) mice (obtained from The Jackson Laboratory). 13R−/−.Foxp3-GFP mice were generated by breeding 13R−/− C56BL/6 mice to 13R+/+.Foxp3-GFP C57BL/6 mice. Because both 13R and Foxp3 are located on the X chromosome, the breeding was done by speed congenic technology. Only age-matched female mice were used in the experiments. All animals were maintained in our animal care facility for the duration of the experiments. All experimental procedures were performed according to the guidelines of the University of Missouri Animal Care and Use Committee.
MOGp and tetramer
MOGp encompassing amino acid residues 35–55 of MOG, which is encephalitogenic in C57BL/6 mice (23), was purchased from EZBiolab (Westfield, IN). I-Ab tetramer containing MOG aa 38–49 (MOGtet) was obtained from the National Institutes of Health Core Tetramer Facility (Emory University, Atlanta, GA).
Induction of EAE
EAE was induced with MOGp as described (23, 24). Briefly, female mice (6–8 wk old) were induced for EAE by s.c. injection of a 200-μl IFA/PBS (v/v) solution containing 300, 100, or 60 μg of MOGp and 200 μg of Mycobacterium tuberculosis H37Ra (Difco) in the footpads and at the base of the limbs. Six hours later, the mice were given 500 ng of Bordetella pertussis toxin (List Biological Laboratories) i.v. A second injection of B. pertussis toxin was given after 48 h. The mice were then scored daily for clinical signs of EAE as follows: 0, no clinical score; 1, tail weakness; 2, loss of tail tone; 3, hindlimb weakness; 4, hindlimb paralysis; 5, forelimb paralysis; and 6, moribund or death. A cumulative disease score (CDS) was calculated by adding the daily scores that the mice received during the monitoring period divided by the number of mice per group.
Detection of cytokines by ELISA
The levels of IFN-γ and IL-17 in culture supernatant were determined by ELISA as previously described (25). The Abs used in this assay were: IFN-γ, R4-6A2 (capture), and biotinylated XMG1.2; IL-17, TC11-18H10, and biotinylated TC11-8H4 (BD Biosciences). The OD405 was measured using a SpectraMax 340 microplate reader (Molecular Devices, Menlo Park, CA) and analyzed with SoftMax Pro software v3.1.1. Graded amounts of IFN-γ and IL-17 (PeproTech, Rocky Hill, NJ) were used to generate standard curves. The linear portion of the standard curve was used to calculate the concentration of cytokines in culture supernatants.
Cells from spleen (SP), lymph nodes (LN; axillary, inguinal, and popliteal: draining LN), or CNS were first incubated with mouse IgG to block FcγRs. Subsequently, allophycocyanin-conjugated MOGtet and anti-CD4 (RM4-5; BD Biosciences) or isotype control Ab were added for 30 min on ice. The cells were then washed, fixed, and permeabilized using an eBioscience intracellular fixation and permeabilization buffer according to the manufacturer’s instructions. For detection of intracellular cytokine, the cells were incubated with anti–IL-17A Ab (TC11-18H10), anti–IFN-γ (XMG1.2; BD Biosciences), or isotype control for 30 min on ice. For detection of transcription factors, the cells were incubated with anti–retinoic acid–related orphan receptor (ROR)γt (AFKJS-9), anti-Tbet (4B10), anti-Foxp3 (FJK-16s) Ab, or isotype controls (from eBioscience) also for 30 min on ice.
Data were collected on a CyAn (Beckman Coulter,Brea, CA) and analyzed using FlowJo version 10.1 (Tree Star) or Summit software version 4.0 (Dako).
For the conversion experiments using IL-17acre–eYFP mice, the cells were prefixed with 1% paraformaldehyde for 15 min prior to the use of the intracellular fixation and permeabilization buffer.
Isolation of CNS mononuclear cells
Mice were perfused through the left cardiac ventricle with 50 ml of ice-cold PBS. The brain was removed and the spinal cord was flushed out by hydrostatic pressure and placed in PBS. Tissue was minced and cells were processed through a 70-μm filter. Homogenized tissue was resuspended in 8 ml of 37% Percoll, overlaid onto 4 ml of 70% Percoll, and centrifuged at 1800 rpm for 20 min at room temperature with no brake. Mononuclear cells were collected from the interphase of the 37 and 70% Percoll gradient and used for subsequent experiments.
Stimulation with PMA and ionomycin
Isolated draining LN or CNS mononuclear cells (1 × 106 per well) were stimulated in vitro with 50 ng/ml PMA and 500 ng/ml ionomycin in the presence of 10 μg/ml brefeldin A for 2 h. The cells were then washed and used for intracellular cytokine and transcription factors staining as above.
Isolation of Tregs and T effector cells
Female 13R+/+ or 13R−/−.Foxp3-GFP mice were immunized s.c. with 200 μl of IFA/PBS (v/v) solution containing 100 μg of MOG peptide and 200 μg of M. tuberculosis H37Ra (Difco) in the footpads and at the base of the limbs.
For isolation of Tregs the LN were harvested on day 10 postimmunization. Cells were gated on CD4 and CD25 and sorted on the basis of GFP (Foxp3+) expression (97% purity) using a Beckman Coulter MoFlo XDP.
For isolation of effector Th1 and Th17 cells, SP cells were harvested on day 10 postimmunization. CD4+ T cells were isolated from the SP by negative selection of lineage-specific cells using a CD4+ T cell isolation kit (Miltenyi Biotec, San Diego, CA). The CD4+ T cells were stimulated with PMA and ionomycin for 3 h to induce cytokine secretion optimal for isolation of bulk Th17 and Th1 cells by IL-17 and IFN-γ secretion assay/cell enrichment and detection kits, respectively (Miltenyi Biotec). Only cells with 98% purity are used for in vitro and in vivo suppression assays.
In vitro Treg suppression assays
Tregs and effector T cells were cocultured at different Treg/T effector ratios for 3 d and the culture supernatants were used to measure residual IFN-γ or IL-17 by ELISA.
The 50% inhibition ratio and sensitivity index
The percentage residual cytokine production represents the amount of cytokine produced in the presence of Tregs over the amount of cytokine produced in the absence of Tregs times 100. The 50% inhibition ratio (IR50) is the Treg/T effector ratio at which the residual cytokine production is reduced by 50%. It is calculated by dividing the number of Tregs by the number of T effector cells at the 50% inhibition point. The sensitivity index (SI) is the intrinsic decline in cytokine production potential per effector T cell, which represents the absolute amount of cytokine that an effector T cell could not produce due to suppression by Tregs. It is calculated as the ratio whereby the numerator is the amount of cytokine (femtograms per milliliter) produced at the 50% inhibition point and the denominator is the absolute number of T effector cells per milliliter times the IR50. The SI unit is femtograms.
Evaluation of the impact of effector T cell sensitivity to Treg suppression on EAE
Effector Th1 or Th17 cells from 13R+/+ or 13R−/− mice were separately mixed with Tregs isolated from 13R+/+ and transferred i.v. into Rag2−/− C57BL/6 mice. Two days later the hosts were induced for EAE with 100 μg of MOGp and then monitored daily for clinical signs of EAE.
Data were analyzed using either an unpaired, two-tailed Student t test or Mann–Whitney U test as indicated. All statistical analyses were performed using Prism software v6 (GraphPad Software, La Jolla, CA).
13R−/− mice develop early onset and more severe EAE
The onset of EAE in C57BL/6 mice induced for disease with 300 μg of MOGp usually manifests around day 10 postimmunization (23). However, when a similar disease induction regimen was applied to 13R−/− mice, the clinical signs of paralysis manifested as early as day 6 postimmunization whereas 13R+/+ control mice had disease onset around day 10 as is typical in our mouse colony (Fig. 1A). Specifically, the average day of onset for 13R−/− mice was 6.6 ± 0.2 compared with 10.7 ± 0.3 for 13R+/+ animals (Fig. 1A). Discrepancies also manifest at disease incidence level, as 100% of 13R−/− mice displayed clinical signs of paralysis at day 8 postimmunization whereas 13R+/+ mice remained symptom free (Fig. 1B). Despite differences in disease onset, both strains had similar clinical scores at the peak of disease and resolved their symptoms in a comparable manner (Fig. 1C). This indicates that 13R contributes regulatory mechanisms that influence the onset of disease. To begin exploring the mechanisms by which 13R influences disease manifestation, mice were induced for EAE with graded doses of MOGp and the animals were monitored daily for signs of paralysis. The findings indicate that 100 μg of MOGp, although preserving early disease onset in 13R−/− mice, does not significantly influence the maximal severity or resolution of disease relative to 13R+/+ mice (Fig. 1D). However, when the mice were induced for EAE with 60 μg of MOGp, the 13R−/− animals developed more severe paralytic signs at the plateau of disease and were unable to resolve their symptoms for the 40-d monitoring period in comparison with 13R+/+ mice (Fig. 1E). Neither strain developed disease with 20 μg of MOGp (Fig. 1F). Taken together, these results indicate that 13R serves to control disease manifestation. To determine how 13R influences T cell responses to hasten disease onset, we analyzed the frequency of, as well as cytokine production by, Th1 and Th17 effector cells at the initial phase of EAE. Ex vivo analysis of T cell responses indicate that in the 13R-deficient mice there were more Th17 but less Th1 cells among Ag-specific MOGtet+ LN cells relative to 13R-sufficient mice (Fig. 1G, 1H). This is significant whether the results are presented as percentages (Fig. 1G) or absolute number (Fig. 1H). Restimulation of LN cells with MOG peptide in vitro shows that there was higher IL-17 production but lower IFN-γ secretion by cells from 13R−/− versus 13R+/+ mice (Fig. 1I). These significant differences are Ag-specific, as they correlate with the dose of MOGp. Taken together, these data suggest that the increased susceptibility and early onset of EAE in 13R−/− mice is related to increased frequency of Th17 rather than Th1 effector cells.
13R controls Ag-driven Th17 to Th1 conversion
To begin analyzing the mechanism by which 13R controls manifestation of EAE, we sought to profile the responses of Th1 and Th17 cells both in the LN and CNS. Accordingly, mice were induced for EAE with MOGp and their LN and CNS ex vivo Ag-specific T cell responses were analyzed daily during the early phase of disease. The data show that in the absence of 13R MOGtet+ Th17 cells were more frequent in the LN and CNS as assessed by percentage (Fig. 2A) and absolute number (Fig. 2B, left panel) early during disease onset (days 3 and 4). However, MOGtet+ Th1 cells were present in the LN and CNS at similar percentages (Fig. 2A) and absolute number (Fig. 2B, middle panel) in both 13R−/− and 13R+/+ mice. Interestingly, a significant number of MOGtet+ IFN-γ/IL-17 double-positive cells were observed in 13R+/+ but not 13R−/− mice (Fig. 2B, right panel). In all, these findings indicate that 13R plays a key role in Th17 dynamics and perhaps suggest a process of Ag-driven Th17 to Th1 conversion (22, 26) highlighted by the presence of MOGtet+ IFN-γ/IL-17 double-positive cells (Fig. 2B). This could explain the lower frequency of Th17 cells in the 13R+/+ mice during the early phase of disease (Fig. 1G, 1H), as the process of conversion is possibly in motion.
To determine whether Ag-driven Th17 to Th1 conversion is occurring and whether 13R plays a role in this process, we used IL-17acre (22) and ROSA26-YFP reporter mice to generate 13R+/+ IL-17acre–eYFP and 13R−/− IL-17acre–eYFP strains in which IL-17 production will trigger expression of YFP that persists even when IL-17 expression is terminated as a result of conversion to Th1 cells. In an initial experiment, we assessed the penetrance of the YFP tag to IL-17 production. The results show that in both 13R+/+ and 13R−/− mice approximately half of Th17 cells that were induced upon induction of EAE were tagged with YFP (Fig. 3A), a finding that agrees with the original study in the IL-17acre–eYFP reporter system (22). Furthermore, when Th1 cells arising alongside Th17 cells were tested for YFP expression, a significant number were YFP+, indicating that conversion is occurring in both strains (Fig. 3B).
To correlate the presence of transitional MOGtet+ IFN-γ/IL-17 double-positive cells with conversion, the IL-17acre–eYFP reporter mice induced for EAE with 100 μg of MOGp were examined for expression of signature transcription factors (RORγt for Th17 and Tbet of Th1 cells) between day 4 and 6 after disease induction, a time period when transition is in motion. The data show that although most CD4+MOGtet+YFP+ cells expressed both RORγt and Tbet in the LN of 13R+/+ mice on day 5 after disease induction, no significant percentages of these cells were observed in the 13R−/− mice during the day 4–6 monitoring period (Fig. 3C). Again, in the CNS, although most (79.6%) of cells were double-positive in the 13R+/+ mice, only 9.3% had both transcription factors in 13R−/− mice at day 5 after disease induction (Fig. 3D). The absolute number of CD4+MOGtet+YFP+ cells expressing both RORγt and Tbet also were significantly higher in the LN (Fig. 3E) and CNS (Fig. 3F) of 13R+/+ versus 13R−/−, peaking at day 5 (120 h) after disease induction. Interestingly, CD4+MOGtet+YFP+ Tbet single-positive cells begin to accumulate significantly both in the LN (Fig. 3G) and CNS (Fig. 3H) 6 h past day 5 (120 h) as the number of Tbet/RORγt double-positive cells decline, indicating that Th17 to Th1 conversion goes through this transitional stage that is under the influence of 13R.
13R does not influence the frequency or function of Tregs
13R seems to foster Th17 to Th1 conversion and accumulation of the latter in the CNS. In light of this observation, however, the early onset and severe EAE observed in 13R−/− mice is puzzling. It is possible, though, that 13R affects the frequency and function of Tregs, which would explain the differential disease outcome in the two strains. To test this premise, we began by crossing Foxp3-GFP reporter mice to the 13R−/− strain and then analyzed the frequency of CD4+Foxp3+(GFP) Tregs in different lymphoid organs from both 13R+/+ and 13R−/−.Foxp3-GFP reporter strains. The results indicate that the frequency of Tregs in the thymus, SP, and LN are similar in both strains and do not show any significant difference (Fig. 4A–C). We then thought that the function of Tregs might be compromised in the 13R−/− mice. To evaluate this postulate, Tregs were sorted from 13R+/+ and 13R−/− mice on day 10 postimmunization with MOGp and tested for suppression of Th1 and Th17 effector cells isolated from the same mice. The findings show that Tregs from 13R−/− mice are as effective as Tregs from 13R+/+mice whether the targets are Th1 or Th17 effector cells from 13R+/+ (Fig. 4D) or 13R−/− (Fig. 4E) mice. These data indicate that there is no defect in Treg function in 13R−/− mice.
13R promotes diminished sensitivity of Th1 cells to suppression by Tregs
Given that the frequency and function of Tregs are not affected by 13R deficiency, one would envision that the early onset and severe EAE observed in 13R−/− mice is related to effects of the receptor on effector T cells. Because both Th1 and Th17 cells contribute to EAE (27–30), the shift in cell frequency with more Th17 and less Th1 cells in 13R−/− mice, although it can be interpreted as compensatory, should not exacerbate the disease. This would logically imply that sensitivity to Treg suppression, rather than frequency, may be influenced by 13R. To test this postulate, splenic Th1 cells were sorted from 13R+/+ and 13R−/− mice on day 10 postimmunization with MOGp, and their sensitivity to suppression by LN Tregs was evaluated. The findings indicate that Th1 cells from 13R−/− mice ceased IFN-γ secretion at both 1:1 or 1:4 Treg/Th1 ratios whereas Th1 cells from 13R+/+ mice continued to produce significant levels of the cytokine (Fig. 5A). These findings suggest that Th1 cells from 13R+/+ mice are less sensitive to Treg suppression than those from 13R−/− mice. To determine the degree of sensitivity to Treg suppression the experiment was carried out with a broad range of Treg/Th1 ratios. Again, the results indicate that Th1 cells from 13R+/+ mice are less sensitive to suppression whether the Tregs are from 13R+/+ or 13R−/− mice (Fig. 5B). In fact, Th1 cells from 13R+/+ mice had their effector cytokine production suppressed by 50% with a ratio of 1:2.3 (Treg to Th1) whereas Th1 cells from 13R−/− mice reached 50% cytokine suppression with a much lower ratio of 1:33 (Fig. 5C). These ratios, referred to as IR50 when expressed as numerical values, again indicate that the Th1 cells from 13R−/− mice are more sensitive to suppression than are Th1 cells from 13R+/+ mice whether the Tregs are from 13R+/+ (compare 0.43 for 13R+/+ to 0.03 for 13R−/− Th1 cells) or 13R−/− (compare 0.33 for 13R+/+ to 0.02 for 13R−/− Th1 cells) mice (Fig. 5C). These IR50 values were then used to define the SI for suppression of Th1cells by Tregs. The SI was calculated as the ratio whereby the numerator is the amount of cytokine (femtograms per milliliter) produced at the 50% inhibition point and the denominator is the absolute number of T effector cells per milliliter times the IR50. We define the SI as the intrinsic decline in cytokine production potential per effector T cell. In essence, the SI represents the absolute femtogram amount of cytokine that an effector T cell could not produce due to suppression by Tregs. Accordingly, a Th1 cell from 13R−/− mice was unable to produce 883 fg of IFN-γ whereas a Th1 cell from 13R+/+ mice could not produce 63 fg of IFN-γ, indicating that the 13R−/− Th1 cells are much more sensitive to Treg suppression than are 13R+/+ Th1 cells (Fig. 5C). Similar sensitivity patterns were observed when the Tregs were from 13R−/− mice (Fig. 5C). Overall, the finding indicates that 13R plays a role in the sensitivity of effector Th1 cells to suppression by Tregs, as Th1 cells from 13R+/+ were less sensitive to Treg suppression than were Th1 cells from 13R−/− mice.
13R promotes increased sensitivity of Th17 cells to Treg suppression
In light of the fact that Th1 cells in 13R−/− mice are highly sensitive to suppression by Tregs, the observation of early onset and severe EAE in 13R−/− mice is perplexing. One likely possibility is that 13R displays a reverse effect on the susceptibility of Th17 cells to Treg suppression. To test this premise, splenic Th17 cells were sorted from 13R+/+ and 13R−/− mice on day 10 postimmunization with MOGp and their sensitivity to suppression by LN Tregs was measured. The findings indicate that Th17 cells from 13R−/− mice continue to produce higher levels of IL-17 cytokine at both 1:1 or 1:4 Treg/Th1 ratios whereas Th17 cells from 13R+/+ mice produce significantly less IL-17 at both ratios (Fig. 6A). These findings indicate that Th17 cells from 13R+/+ mice are more sensitive to Treg suppression than are those from 13R−/− mice. Similar to experiments with Th1 cells, experiments were carried out with a broad range of Treg/Th1 ratios to determine the degree of Th17 sensitivity to Treg suppression. The findings show that Th17 cells from 13R+/+ mice are more sensitive to suppression whether the Tregs are from 13R+/+ or 13R−/− mice (Fig. 6B). In fact, Th17 cells from 13R+/+ mice had their effector cytokine production suppressed by 50% with a ratio of 1:5.2 (Treg to Th1) whereas Th17 cells from 13R−/− mice reached 50% cytokine suppression with a much higher ratio of 1:1.4. The IR50 for Th17 cells from 13R−/− mice is higher (0.69) than that of Th17 cells from 13R+/+ mice (0.19), indicating that the Th17 cells from 13R−/− mice are less sensitive to Treg suppression than are Th17 cells from 13R+/+ mice (Fig. 6C). Similar IR50 patterns were observed when the Tregs were from 13R−/− mice. Moreover, the SI for Th17 cells from 13R−/− mice was 11 fg whereas the SI for Th17 cells from 13R+/+ mice was 45 fg, again indicating that the 13R−/− Th17 cells are much less sensitive to Treg suppression than 13R+/+ Th17 cells (Fig. 6C). Similar sensitivity patterns were observed when the Tregs were from 13R−/− mice (Fig. 6C). Overall, the findings indicate that 13R increases the sensitivity of effector Th17 cells to suppression by Tregs.
13R-mediated resistance of Th1 effector cells to Treg suppression leads to exacerbation of EAE
Given that in vitro coculture assays indicate that Th1 cells from 13R+/+ mice are less sensitive to Treg suppression than are those derived from 13R−/− mice, one would envision that such differential sensitivity translates into discrepancy in EAE manifestation. To test these premises, effector Th1 cells from MOGp-immunized 13R+/+ or 13R−/− mice were transferred into Rag2−/− C57BL/6 animals alongside LN Tregs from MOGp-immunized 13R+/+.Foxp3-GFP reporter mice. The hosts were then induced for EAE with MOGp and monitored daily for disease progression. Note that transfer of bulk Th1 cells induced weak clinical signs of EAE, which would be readily suppressible by Tregs. Immunization with MOGp was therefore used to recall Ag-specific Th1 effectors and maximize disease severity to assess for Treg suppression in a rigorous manner. The results indicate that mice that were recipients of Th1 effectors from 13R+/+ mice alongside Tregs had delayed disease onset relative to those given effectors without Tregs (Fig. 7A). In contrast, mice that were recipients of Th1 effectors from 13R−/− mice alongside Tregs had delayed disease onset and more severe paralytic signs of EAE in comparison with hosts that received effector Th1 cells without Tregs (Fig. 7B). Thus, Th1 effectors from 13R+/+ mice are less sensitive to Treg suppression and drive severe EAE whereas Th1 effectors from 13R−/− mice are more sensitive to Treg suppression and drive milder EAE (Fig. 7C). In fact, although the onset of disease was similar for Th1 effectors from 13R+/+ and 13R−/− mice (days 16 and 17, respectively), the CDS was much higher for Th1 effectors from 13R+/+ mice (111.8 ± 7.8) than those from 13R−/− mice (68.7 ± 6.4). Also, the mean maximal disease score (mmds) for Th1 effectors from 13R+/+ mice was 3.7 and occurred at three episodes (days 32–34, 37–39, and 42–44), which is higher than the mmds for Th1 effectors from 13R−/− mice (2.5), which occurred during one episode only (days 40–43). In all, the 13R supports resistance of Th1 cells to Treg suppression, leading to severe EAE.
13R-mediated sensitivity of Th17 effector cells to Treg suppression leads to reduced EAE severity
In vitro coculture suppression assays indicated that Th17 cells from 13R+/+ mice are more sensitive to Treg suppression than those derived from 13R−/− mice. We then set up experiments to test whether such differential sensitivity is operative in vivo and manifests in the severity of EAE. Accordingly, effector Th17 cells from MOGp-immunized 13R+/+ and 13R−/− mice were transferred into Rag2−/−C57BL/6 mice alongside LN Tregs from MOGp-immunized 13R+/+.Foxp3-GFP reporter mice. The hosts were then induced for EAE with MOGp and monitored daily for disease progression. Note that transfer of bulk Th17 cells induced weak clinical signs of EAE, which would be readily suppressible by Tregs. Immunization with MOGp was therefore used to recall Ag-specific Th17 effectors and maximize disease severity to assess for Treg suppression in a rigorous manner. The results indicate that mice recipient of Th17 effector cells from 13R+/+ mice alongside Tregs had less severe disease relative to those given effectors without Tregs (Fig. 8A). In contrast, mice that received Th17 effectors from 13R−/− mice alongside Tregs had a similar disease pattern in comparison with hosts that received effector Th17 cells without Tregs (Fig. 8B). Thus, Th17 effectors from 13R+/+ mice are sensitive to Treg suppression and drive milder EAE whereas Th17 effectors from 13R−/− mice are less sensitive to Treg suppression and drive more severe EAE (Fig. 8C, Table I). In fact, although the onset of disease was similar for Th17 effectors from 13R+/+ and 13R−/− mice (day 17), the CDS was lower for Th17 effectors from 13R+/+ mice (70.2 ± 9.8) than those from 13R−/− mice (144.8 ± 24.6). Also, the mmds for Th17 effectors from 13R+/+ mice was 2.7 and occurred at two episodes (days 35–37 and 41–46), which is lower than the mmds for Th17 effectors from 13R−/− mice (4.6), which occurred for a longer episode (days 32–46). In all, the 13R supports sensitivity of Th17 cells to Treg suppression, leading to mild EAE.
|Sensitivity to Treg Suppression .||Th17 to Th1 Conversion|
|13R−/−||**||*||+||Early onset and severe|
|Sensitivity to Treg Suppression .||Th17 to Th1 Conversion|
|13R−/−||**||*||+||Early onset and severe|
, No sensitivity to Treg suppression; **, sensitive to Treg suppression ***, highly sensitive to Treg suppression as defined by EAE manifestation; +, little or no conversion; +++, pronounced conversion.
Both IL-4 and IL-13 cytokines have been shown to play an anti-inflammatory role in EAE (6, 31). Because both cytokines share the HR, IL-4 and IL-13 may exercise such a function through the HR. Indeed, this was the case, as mice lacking the HR developed early disease onset and more severe clinical signs of paralysis upon EAE induction with MOGp. This correlates with an increased frequency of Ag-specific Th17 cells in the LN and the CNS early on day 4 after disease induction, a phenomenon that bodes well with the notion that these cells migrate to the CNS to attract neutrophils and initiate the inflammatory process (32). Accumulation of Th1 cells within the CNS was comparable to 13R+/+ mice, suggesting that the differential disease onset and severity are rather related to Th17 cells. In line with this thought is the observation that there was lower Th17 to Th1 conversion in 13R−/− mice where the onset of disease is early and the severity is more pronounced (Table I). In contrast, there was greater conversion of Th17 to Th1 cells particularly in the CNS of 13R+/+ mice, which developed rather typical EAE (Table I). The process of Ag-driven Th17 to Th1 conversion has previously been described (33) and, similar to IL-23R–deficient mice (34), proceeds through Tbet/RORγt double-positive transitional intermediate T cells producing both IFN-γ and IL-17 cytokines. The HR and its ligands IL-4 and IL-13 likely influence this process, as the 13R−/− mice had rather less pronounced Th17 to Th1 conversion. In addition to conversion, there was a differential sensitivity to Treg suppression by the effector T cells in 13R+/+ versus 13R−/− mice, which correlates with the type of EAE observed in the two strains (Table I). Indeed, in 13R−/− mice the Th17 cells that accumulate in the CNS due to limited conversion display little sensitivity to suppression by Tregs, further supporting early onset and severe EAE. In contrast, in 13R+/+ mice where Th17 to Th1 conversion was prominent, the residual Th17 cells were more sensitive to Treg suppression, hence the milder EAE observed in the 13R+/+ strain. Despite that differential sensitivity to Treg suppression was observed among 13R+/+- and 13R−/−-derived Th1 cells, it could not influence the disease outcome, perhaps because of similar accumulation in the CNS where the disease is impacted by Th17 cells. Although sensitivity of effector T cells to Tregs has previously been reported to affect cell proliferation and survival (35, 36), perhaps through regulation of intrinsic factors (37–40), it was not clear whether both Th1 and Th17 effectors display similar patterns of sensitivity. This study, however, shows that Th1 and Th17 cells display different patterns of sensitivity to Treg suppression in both 13R+/+ and 13R−/− mice. Because Th1 and Th17 cells do not express the HR even in 13R+/+ mice, it is likely that the sensitivity to Treg suppression is guided by APCs whose function had been shaped by the HR (15, 16, 20, 21). It is possible that the interaction of naive T cells with HR-conditioned APCs not only guides subset differentiation but also prompts expression of surface molecules that control the sensitivity of the effector to Treg suppression (41–44). Additionally, because IL-21 has been shown to play a role in effector resistance to Treg suppression (45), it is possible that the differential sensitivity to Treg suppression among Th1 and Th17 cells is related to production of the cytokine by the latter and not the former. Overall, both conversion and sensitivity to Treg suppression point to Th17 cells as the culprit of disease severity regulated by the HR. Specifically, Th17 accumulation due to reduced conversion and increased Th17 cell resistance to Treg suppression explain the severity of EAE in 13R−/− mice. The reverse would explain the milder EAE observed in 13R+/+ mice.
In all, the HR and its cytokines IL-4 and IL-13 play a role in Th17 to Th1 conversion as well as the differential sensitivity of the effectors to Treg suppression, leading to control of EAE.
This work was supported by National Institute of Neurological Disorders and Stroke Grant R01 NS057194 (to H.Z.) and by the J. Lavenia Edwards Endowment. M.M.M. was supported by National Institute of General Medical Sciences T32 Training Grant GM008396.
Abbreviations used in this article:
cumulative disease score
experimental allergic encephalomyelitis
enhanced yellow fluorescent protein
50% inhibition ratio
mean maximal disease score
myelin oligodendrocyte glycoprotein 35–55 peptide
retinoic acid–related orphan receptor
T regulatory cell.
The authors have no financial conflicts of interest.