TGF-β signaling via smad4 drives IL-10 production in effector Th1 cells and reduces T cell trafficking in EAE

SUMMARY

Effector Th1 cells perpetuate inflammatory damage in a number of autoimmune diseases, including MS and its animal model EAE. Recently, a self-regulatory mechanism was described where effector Th1 cells produce the immunomodulatory cytokine IL-10 to dampen the inflammatory response in both normal and autoimmune inflammation. While the presence of TGF-β has been suggested to enhance and stabilize an IFN-γ⁺IL-10⁺ phenotype, the molecular mechanism is poorly understood. Additionally, in the context of adoptive transfer EAE, it is unclear if IL-10 acts on the transferred Th1 cells or on cells within the host. In the present study, using myelinspecific TCR-Tg mice, repetitive Ag stimulation of effector Th1 cells in the presence of TGF-β increased the population of IFN-γ⁺IL-10⁺ cells, which correlated with a decrease in EAE severity. Additionally, TGF-β signaling caused binding of smad4 to the IL-10 promoter, providing molecular evidence for TGF-β-mediated IL-10 production from Th1 effector cells. Lastly, this study demonstrates that IL-10 reduced encephalitogenic markers such as IFN-γ and T-bet on Th1 effector cells expressing the IL-10R, but also prevented recruitment of both transferred and host-derived inflammatory T cells. These data establish a regulatory mechanism by which highly activated Th1 effector cells modulate their pathogenicity through induction of IL-10.

INTRODUCTION

Th1 cells play a vital role in the adaptive immune response to intracellular pathogens; however, uncontrolled Th1 responses lead to excessive tissue inflammation and damage. As a result, immune regulatory mechanisms must fine-tune Th1 responses in a delicate balance of effective host-protection without deleterious damage. In autoimmune diseases such as psoriasis, rheumatoid arthritis, and MS, effector-memory Th1 cells directed toward self Ags are a major contributor to disease pathology [1–3]. Consequently, there is much interest in the biological function and regulation of effector-memory Th1 cells and how their effector mechanisms are regulated in both normal and autoimmune inflammation.

IL-10 is a multi-functioning cytokine that was first described as a product of Th2 cells that negatively regulated the activation and cytokine secretion of Th1 cells [4]. IL- 10 is now recognized to inhibit Th1 effector functions by reducing antigen presentation and cytokine production from antigen presenting cells [5], although IL-10 can also act directly on T cells to reduce proliferation and cytokine secretion [6–9]. In certain contexts, many immune cells can produce IL-10, including B cells, mast cells, macrophages, dendritic cells and other T cell lineages [5, 10, 11].

Recently, Th1 cells have been identified as a source of IL-10 (reviewed in [12, 13]) that can function to induce peripheral tolerance [14, 15], limit the Th1 response after infection [16], and reduce the encephalitogenic capacity of myelin-specific Th1 and Th17 cells in EAE [17, 18]. IL-27 was shown to induce production of IL-10 from Th1 cells, and TGF-β has also been implicated in the generation and maintenance of this population [17–19], although the molecular mechanisms are still not clear.

In the context of MS and EAE, IL-10 elicits beneficial effects on disease. In MS patients, IL-10 levels are increased in the serum during disease remission [20]. Additionally, IFN-β and glatiramer acetate, two widely used MS treatments, demonstrate efficacy in part by inducing IL-10 production from immune cells [20–23]. Genetic studies using the EAE model show that IL-10 deletion enhances EAE disease severity, while over-expression protects mice [24]. Thus, understanding what causes Th1 effector cells to increase IL-10 production and identifying regulatory functions of IL-10 may reveal potential therapeutic targets for the treatment of MS and other inflammatory Th1- mediated pathologies. In the present study, using MBP Ac1-11 TCR-Tg T cells in the adoptive transfer EAE model, we sought to assess the development and function of effector Th1 cells producing both IFN-γ and IL-10 throughout multiple Ag stimulations, modeling the episodic activation of distinct TCR-specific effector-memory T cells that is observed in MS [25, 26]. Additionally, since TGF-β is known to enhance IL-10 production from Th1 effector cells, we wanted to determine the molecular basis for this observation. Lastly, we wanted to determine how IL-10 from Th1 effector cells was mitigating EAE severity, since these cells still produce high levels of IFN-γ. The results from this study provide an enhanced understanding of the role of TGF-β in the induction of IL-10 from Th1 effector cells, while also providing a mechanism for the immunomodulatory role of IL-10 in EAE.

RESULTS

Repetitive Ag stimulation in combination with TGF-β increases the IFN-γ⁺IL-10⁺ phenotype

Effector-memory CD4⁺ T cells exhibit enhanced activation and effector function when re-exposed to their specific Ag in order to mount a more rapid and effective immune response. We have previously demonstrated that a second Ag challenge induces regulatory signals by promoting IL-10 production from Th1 effector cells, which is enhanced with the addition of TGF-β [17]. To determine if repetitive cellular activation via Ag-stimulation (with or without TGF-β) enhances this effect, we began by differentiating Vα2.3/Vβ8.2 TCR-Tg splenocytes into a Th1 population with Ag and IL- 12. After primary stimulation, these cells are predominantly IFN-γ⁺T-bet⁺, but do not make IL-10 or IL-17 (Fig. 1A𠄼). These cells were then rested and restimulated with Ag alone or in combination with TGF-β, up to three times. After each round of stimulation, IFN-γ and IL-10 expression were analyzed by flow cytometry (Fig. 1D). Secondary stimulation of Th1 effector cells with Ag only induced a small population (3.2%) of cells making both IFN-γ and IL-10. While this population did not significantly change on the third round of stimulation, by the fourth round of stimulation it had nearly doubled in size to 5.7%. Adding TGF-β during each round of stimulation had a more dramatic effect on enhancing IL-10 production from Th1 effector cells. The percentage of IFN-γ⁺IL-10⁺ cells increased from 5.7% after secondary stimulation to 17% after fourth stimulation. To measure the amount of IL-10 being made by the IFN-γ⁺IL-10⁺ cells, supernatants were analyzed by ELISA (Fig. 1E). As expected, culture conditions with a higher percentage of IFN-γ⁺IL-10⁺ cells contained more IL-10 in the supernatant. These data demonstrate that repetitive Ag stimulation causes an increase in the population of IFN-γ⁺IL-10⁺ cells. This effect is significantly enhanced by the addition of TGF-β.

Figure 1.

Repetitive Ag stimulation in the presence of TGF-β promotes an IFN-γ⁺IL-10⁺ phenotype. Naïve Vα2.3/Vβ8.2 TCR-Tg splenocytes were differentiated with Ag and IL-12. Cells were collected at 72 h and analyzed by flow cytometry for IFN-γ, IL-10, and IL-17 (A) or T-bet (B). C, Supernatants were analyzed by ELISA for IFN-γ, IL-10, and IL-17. D, Cells from (A) were rested and restimulated multiple times with Ag only or Ag plus TGF-β. After each stimulation, cells were analyzed by flow cytometry for IFN-γ and IL-10. Numbers in quadrants indicate percentage of total CD4⁺CD44⁺ cells. E, Supernatants were analyzed after each round of stimulation by ELISA for IL-10 (mean ± SEM; student t test where significance was considered p<.05). Data are representative of three independent experiments.

TGF-β does not induce IL-17 or Foxp3 expression in Th1 effector cells

TGF-β is known to generate inducible Tregs (iTreg) from naïve T cells that are characterized by expression of Foxp3 and are able to suppress T cell effector functions [27]. To determine whether effector Th1 cells stimulated multiple rounds with Ag and TGF-β generated an iTreg phenotype, Foxp3 expression was examined after each round of stimulation using flow cytometric analysis. As shown in supplemental figure 1 no Foxp3 expression was detected in Th1 effector cells, despite the presence of TGF-β during stimulation.

TGF-β is also known to play a role in Th17 cell biology, although recent evidence has demonstrated this is an indirect mechanism through inhibition of Th1 differentiation that subsequently enhances the Th17 phenotype [28–31]. However, to determine whether TGF-β induced IL-17 expression from Th1 effector cells during multiple rounds of Ag stimulation, supernatants were collected and analyzed by ELISA. No IL-17 was detected, even when high-dose (5ng/ml) TGF-β was added to the cultures (data not shown).

IFN-γ⁺IL-10⁺ effector Th1 cells reduce EAE disease severity

Myelin-specific TCR-Tg T cells are capable of inducing EAE when activated in vitro with Ag and adoptively transferred into naïve recipient mice [32, 33]. Since multiple studies have demonstrated that IL-10 from a variety of cellular sources is able to reduce EAE severity [34–36], we sought to determine if an increase in the percentage of IFN-γ⁺IL-10⁺ Th1 effector cells reduced disease severity. Myelin-specific Th1 effector cells were generated and stimulated (two or three rounds) in vitro as previously described and transferred into mice. Figure 2A details each culture condition and is a graphical representation derived from flow cytometry data of the percentage of IFN-γ⁺IL-10⁺ Th1 effector cells, illustrating that the percentage of IFN-γ⁺IL-10⁺ T cells correlates with the number of times the cells were exposed to TGF-β.

Figure 2.

Increasing the percentage of IFN-γ⁺IL-10⁺ cells transferred reduces EAE severity. Myelin-specific Th1 effector cells were generated and restimulated with Ag alone or in combination with TGF-β for a second (2°), third (3°), or fourth (4°) stimulation. A, After each round of stimulation, cells were analyzed by flow cytometry for IFN-γ and IL-10. Percentages shown are from the CD4⁺CD44⁺ gate. B-D, Cells collected in (A) were transferred into mice and EAE disease course and survival were monitored. Data are representative (A, D) or combined (B, C) from three independent experiments.

Adoptive transfer of Th1 effector cells after the third round of Ag only stimulation resulted in the most severe disease and poorest survival rate (Fig. 2B,C). This corresponds to the culture condition with the fewest IFN-γ⁺IL-10⁺ cells (Fig. 2A). Conversely, Th1 effector cells that were restimulated in the presence of TGF-β during both second and third rounds of stimulation transferred a milder disease with 100% survival of the mice, despite 100% incidence of disease. To ensure we were not simply inducing an anergic state, proliferation was measured with a ³H-thymidine incorporation assay and demonstrated the ability of these cells to continue to proliferate in response to Ag stimulation (data not shown).

To further enhance the IFN-γ⁺IL-10⁺ population, cells were stimulated a fourth time (Fig 2A). Transferring these cells into mice resulted in mild EAE (Fig. 2D). It should also be noted that the Th1 cells receiving Ag only four times had twice the number of IFN-γ⁺IL-10⁺ cells compared to the Th1 cells that had Ag only stimulation three times (Fig 2A), and this correlated with changes in disease course (closed circles in Fig. 2B vs. open circles in Fig. 2D). This illustrates the biological significance of the IL-10 producing Th1 cells, even in the absence of exogenous TGF-β.

Silencing IL-10 in Th1 effector cells restores encephalitogenicity

TGF-β is a multi-functioning cytokine that induces numerous effects in T cells with specificity based on the differentiation state of the target cell [37]. To verify that TGF-β-induced IL-10 production from effector Th1 cells is responsible for the reduction in EAE severity observed after tertiary stimulation, effector Th1 cells were transfected with either siRNA specific for IL-10 (IL10-siRNA) or a nonsense siRNA (NS-siRNA), activated with Ag alone or in combination with TGF-β, then transferred into mice. Supernatants were analyzed for IL-10 by ELISA to demonstrate the efficacy of IL10-siRNA (Fig. 3A). As seen in figure 3B silencing IL-10 restored disease severity, indicating that IL-10 was mediating the amelioration of disease.

Figure 3.

Silencing IL-10 during tertiary stimulation of Th1 effector cells restores encephalitogenicity. Effector Th1 cells stimulated two rounds with Ag alone or in combination with TGF-β were transfected prior to third stimulation with siRNA specific for IL-10 or a nonsense siRNA. A, Supernatants were analyzed by ELISA for IL-10. B, Cells were transferred into mice and disease course was monitored. Data are representative of two independent experiments.

TGF-β directly induces IL-10 production via Smad4

The molecular mechanisms driving IL-10 production from Th1 effector cells are still being elucidated. Previous studies have demonstrated the importance of IL-27 signaling via Stat-3 to induce IL-10 production [18, 19]. These studies also imply a role for TGF-β in stabilizing IL-10 production, but fail to show molecular evidence. To determine the role of IL-27 and TGF-β in IL-10 production from Th1 effector cells, Th1 cells were stimulated with Ag and combinations of TGF-β, or neutralizing Abs to IL-27 and TGF-β (Fig. 4A). In our system, IL-27 was dispensable for IL-10 production, indicating that TGF-β was sufficient to drive IL-10 production in Th1 effector cells.

Figure 4.

Smad4 directly binds to the IL-10 promoter in TGF-β treated Th1 effector cells. A, Effector Th1 cells were stimulated with Ag alone or in combination with TGF-β or neutralizing Abs to TGF-β or IL-27 and IL-10 production was measured by ELISA. B, Cells after primary Th1 stimulation and secondary Th1 + TGF-β or Th1 + α-TGF-β stimulation were harvested for use in a chromatin immunoprecipitation assay performed with an Ab specific for Smad4. Bound DNA was purified and used as a template in standard PCR. The IL-10 promoter was specifically amplified in Th1 effector cells stimulated in the presence of TGF-β. Data are representative of three independent experiments.

Because the possibility still existed that TGF-β indirectly induced IL-10 production through effects on cellular activation or by other unknown mechanisms, we performed a ChIP assay to determine whether TGF-β-induced Smad4 was directly bound to the IL-10 promoter in Th1 effector cells stimulated with Ag and TGF-β. Th1 cells after primary stimulation were used as a negative control (in addition to a rabbit-IgG negative control) since these cells do not produce any IL-10, as shown in figure 1C. As demonstrated in figure 4B, immunoprecipitation with anti-Smad4 pulled down the DNA segment of the IL-10 promoter containing the predicted Smad4 binding site [38], suggesting a direct regulation of IL-10 gene expression via TGF-β signaling. Additionally, neutralizing TGF-β during secondary stimulation abolished smad4 binding to the IL-10 promoter, further verifying the importance of TGF-β signaling in IL-10 production.

IL-10 reduces both IFN-γ and T-bet expression

Since enhancing IL-10 production from Th1 effector cells reduced EAE severity, we next sought to determine how IL-10 affected Th1 effector cell phenotype. To do this, effector Th1 cells were generated and restimulated in the presence of IL-10 either in an Ag/APC setting or on plates coated with anti-CD3 in the absence of APCs. Since IL-10 is known to effect APC function [5], this allowed us to look at its effect on T cells independent of APCs. Intracellular staining for IFN-γ and T-bet expression as markers for T cell encephalitogenicity was performed. Interestingly, when gating on total CD4⁺CD44⁺ cells, we saw a trend but not a significant reduction in IFN-γ and T-bet expression (Supplemental Fig. 2), indicating that IL-10 modulation may be on a more specific subset of Th1 effector cells.

We then wanted to determine what proportion of effector Th1 cells expressed the IL-10R, indicating they were able to respond to IL-10. Effector Th1 cells were stimulated with Ag only or Ag + TGF-β, and IL-10R expression on CD4⁺CD44⁺ cells was analyzed (Fig. 5A). Interestingly, only a small percentage of cells expressed the IL-10R when stimulated with Ag only (7.7%) or Ag + TGF-β (13%). We then analyzed IFN-γ and Tbet on IL-10R⁺ and IL-10R^neg populations. As figure 5A demonstrates, the IL-10R⁺ cells displayed a marked reduction in both IFN-γ and T-bet expression, demonstrating the ability of IL-10 to directly alter the phenotype of Th1 effector cells.

Figure 5.

IL-10R expression is associated with a non-encephalitogenic effector cell phenotype. Effector Th1 cells were stimulated with Ag alone or in combination with TGF-β for 48h. A, Cells were analyzed by flow cytometry for IL-10R, IFN-γ, and T-bet expression on CD4⁺CD44⁺ cells. B, Th1 effector cells were stimulated with TGF-β and an IL-10R neutralizing antibody for 48h, then transferred into mice and EAE disease course was monitored. Data are representative of four independent experiments (A) or combined from two independent experiments (B).

Because the IL-10R is only expressed on a minority of Th1 effector cells, we wanted to determine the importance of this population in the adoptive transfer of myelin-specific Th1 effector cells. Prior to restimulation, Th1 effector cells were cultured with an Ab to the IL-10R that has demonstrated neutralization effects [39]. Forty-eight hours after stimulation with Ag and TGF-β, cells were adoptively transferred into mice. As expected, since relatively few cells express the IL-10R, neutralization led to only a mild increase in EAE disease severity (Fig. 5B). Thus, while IL-10 produced by Th1 effector cells is able to reduce the encephalitogenicity of a small percentage of the cells transferred, IL-10 released in the host environment may play a more important role in altering disease course.

IL-10 from Th1 effector cells prevents host-derived inflammatory Th1 and Th17 cell recruitment

The small percentage of Th1 effector cells expressing the IL-10R pointed towards a role for IL-10 in the host. To determine the effect of IL-10 from Th1 effector cells on the host during EAE, myelin-specific Th1 effector cells were generated and re-stimulated with Ag only or Ag + TGF-β and transferred into mice. As previously shown, Th1 effector cells restimulated in the presence of TGF-β had reduced EAE (Fig. 6A). At day 10 post-transfer, spinal cords were removed and mononuclear cells isolated for phenotypic analysis.

Figure 6.

Mice receiving Th1 effector cells stimulated with Ag + TGF-β have reduced trafficking and host-recruitment of inflammatory cells into the spinal cord. Effector Th1 cells were stimulated with Ag only or Ag + TGF-β for 48h and adoptively transferred into mice (3 × 10⁶ cells/mouse). A, EAE disease course was monitored. B-D, Day 10 post-transfer, spinal cords were removed and mononuclear cells isolated for flow cytometric analysis. B-D, After isolation, CNS mononuclear cells were stained directly ex vivo and analyzed by flow cytometry. E, CNS mononuclear cells were cultured in media alone for 18h, then stimulated with PMA-Ionomycin for five hours and stained for IFN-γ and IL-17. All data are from two independent experiments with groups of pooled mice within each experiment (mean ± SEM).

To determine the composition of mononuclear cells recovered from the spinal cord, flow cytometry was used to analyze CD11b and CD45 (Fig. 6B). This allows for the general identification of resting microglia (CD11b⁺CD45^lo), activated microglia/macrophages (CD11b⁺CD45^hi) and lymphocytes (CD11b⁻CD45⁺). While minor differences were seen in these three populations, none were significant. To determine if the adoptively transferred cells infiltrated into the spinal cord, flow cytometry was used to analyze CD4⁺ T cells expressing the Vβ8.2 TCR that is present on all transferred cells. Spinal cords from mice given Th1 effector cells stimulated with Ag + TGF-β had fewer infiltrating CD4⁺Vβ8.2⁺ T cells (Fig. 6C), as well as very few host-derived infiltrating CD4⁺ T cells (Fig. 6B,D); while the Th1 effector cells stimulated with Ag only caused the recruitment of host-derived CD4⁺ T cells (Fig. 6B,D). Since it is possible that a small percentage of host derived T cells could express the Vβ8.2 TCR chain, we also stained for the Vα2.3 chain and found all Vβ8.2⁺ cells were also Vα2.3⁺ (data not shown). We then analyzed IFN-γ and IL-17 expression to determine differences in cytokine production between transferred Vβ8.2⁺ cells and host-derived Vβ8.2^neg cells (Fig. 6E). While Vβ8.2⁺ T cells from each group demonstrated similar IFN-γ and IL-17 expression, the group given Th1 effector cells stimulated with Ag only had a population of host-derived cells that also produced IFN-γ. Interestingly, we recovered no IL-10 producing Th1 cells from the spinal cord. Since IL-10 can have profound effects on microglia/macrophage populations, we also characterized phenotypic changes in these populations, but observed no significant differences. Taken together, these results suggest that Th1 effector cells stimulated in the presence of TGF-β reduce the recruitment of host-derived inflammatory cells into the spinal cord. This is another potential mechanism for the observed decrease in EAE severity.

DISCUSSION

Th1/Th17-mediated autoimmune diseases are characterized by episodic inflammation in the target organ. In relapsing-remitting MS, inflammation within the CNS damages the myelin sheath and leads to neurodegeneration with long-term neurological consequences. The presence of episodic inflammation suggests that immune regulatory mechanisms function to modulate immune cells, even when they are directed at self-Ags.

We have previously demonstrated that TGF-β, a cytokine present in the inflamed CNS, was capable of enhancing IL-10 production from Th1 effector cells during secondary stimulation as a self-regulatory mechanism [17]. In the present study, we have provided evidence for TGF-β-mediated induction of IL-10 from Th1 effector cells via smad4 signaling throughout multiple rounds of Ag stimulation. Furthermore, using the adoptive transfer EAE model, we have demonstrated the in vivo relevance of IFN-γ⁺IL-10⁺ cells in reducing the encephalitogenicity of myelin-specific Th1 effector cells. IL- 10-mediated effects are observed on both the transferred Th1 effector cells and in the host recruitment of inflammatory cells.

For some time, repetitive Ag stimulation has been used to induce peripheral tolerance in CD4⁺ T cells to render them unresponsive to their specific Ag. This strategy has also been implemented in MS clinical trials to render the myelin-specific autoreactive T cells anergic, but has had mixed outcomes [40–42]. More recently, peripheral tolerance studies have demonstrated that Th1 effector cells begin to make IL-10, which coincides with reduced T cell proliferation, suggesting the development of an anergic state [14, 15]. In our study, we observe an increase in proliferation, cellular activation, and cytokine secretion with each subsequent round of stimulation, demonstrating that the effector Th1 cells are still responsive to Ag and not merely being induced to enter an anergic state (Fig. 1 and data not shown). This is in agreement with a recent study by Saraiva et al that postulated IL-10 production from Th1 cells is in response to situations with severe inflammation [43].

After determining that repetitive Ag stimulation increased the percentage of IFN-γ⁺IL-10⁺ Th1 effector cells and increasing this population decreased EAE severity (Fig. 2,3), we sought to determine the molecular mechanism driving TGF-β-induced IL-10 production. IL-27, a cytokine in the IL-12 family has a demonstrated ability to signal through Stat-3 and induce IL-10 production from Th1 cells [18, 19]. However, in our experiments, neutralizing IL-27 did not significantly affect the level of IL-10 production. This led us to explore the potential role of TGF-β in directly inducing IL-10 from Th1 effector cells. TGF-β classically signals through Smad molecules, of which Smad4 is known to bind the promoters of target genes [44]. Kitani et al. demonstrated that retroviral over-expression of TGF-β caused Smad4-mediated production of IL-10 from numerous T cell subtypes [38]. Additionally, they identified a Smad4 binding site in the IL-10 promoter, and using an artificial system, demonstrated the ability of Smad4 to bind to this cloned sequence. We have built on these observations by using a ChIP assay that allows for in vivo identification of transcription factor binding to DNA in the natural genomic state [45], and demonstrate that Th1 effector cells stimulated in the presence of TGF-β contain Smad4 bound to IL-10 promoters (Fig. 4). Interestingly, even when neutralizing TGF-β during stimulation of Th1 effector cells, we still observe small levels (<1ng/ml) of IL-10 produced. This may be explained by a recent study that demonstrated IFN-γ-induced NF-δB led to small amounts of downstream IL-10 production by macrophages [46], which may also be the case for Th1 effector cells.

IL-10 has been well described to suppress naïve T cell activation by reducing the antigen presenting capabilities of professional APCs by down-regulating co-stimulatory molecules and cytokine secretion [5]; however, the ability of IL-10 to influence Th1 effector cells is not as clear. Interestingly, when we looked at our Th1 effector cell population as a whole, we did not observe significant reductions in IFN-γ or T-bet as a result of IL-10, although a mild trend existed (Supplemental Fig. 2). However, when we refined our analysis based on the presence or absence of the IL-10R, we demonstrated that IL-10R⁺ cells had a dramatic reduction in both IFN-γ and T-bet expression (Fig 5). Because of the small percentage of cells expressing the IL-10R, this cannot fully explain the reduction in EAE disease course and suggests that IL-10 may play an additional role in the host.

In our EAE model system, myelin-specific effector Th1 cells enter the CNS where they perpetuate inflammation. It has also been shown that the recruitment of hostderived inflammatory cells enhances inflammatory damage [47–49]. In our experiments, effector Th1 cells stimulated in the presence of TGF-β had a reduced ability to recruit host inflammatory cells into the spinal cord. This leaves open a number of possibilities as to the exact site of IL-10 importance. IL-10 produced by Th1 effector cells could be functioning in the periphery to reduce chemokine expression, and/or functioning at the level of the blood brain barrier by inhibiting adhesion molecule expression and subsequent trafficking of lymphocytes.

Surprisingly, we did not recover any IL-10-producing Th1 cells from the spinal cords. This could be a result of the transient nature of IL-10 expression and may reflect the time point at which the spinal cords were removed, or may indicate these cells remain in the periphery. We also characterized the phenotype of microglia/macrophages within the spinal cord, but only observed minor differences in IL-6, TNF, and CD86 expression. This also points to a role for IL-10 in the periphery, but may also be due to the fact that the majority of microglia recovered would come from non-lesioned areas and may have diluted out any observable differences.

In conclusion, this study demonstrates that repetitive Ag stimulation in the context of TGF-β increases the self-regulation of effector Th1 cells by inducing IL-10 production through Smad4 signaling. IL-10 from the effector Th1 cells reduced EAE disease severity, in part through down-regulation of encephalitogenic markers like T-bet and IFN-γ, and in part by preventing the trafficking of both transferred and host-derived inflammatory cells into the spinal cord. Taken together, these results demonstrate a mechanism of self-regulation used by the immune system to modulate highly activated effector Th1 cells.

MATERIALS AND METHODS

Mice and adoptive transfer EAE

Vα2.3/Vβ8.2 MBP Ac1-11 TCR-Tg B10.PL mice have been previously described [50]. B10.PL mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice were bred and maintained in a specific pathogen-free animal facility at The Ohio State University Medical Center (Columbus, OH). The Institutional Animal Care and Use Committee at The Ohio State University approved all animal protocols.

EAE was induced by i.p. injection of 3–10 × 10⁶ cells/mouse in 200µl PBS. EAE disease course was scored on a scale of 0–6: 0, no clinical disease; 1, limp tail; 2, moderate hind limb weakness; 3, severe hind limb weakness; 4, complete hind limb paralysis; 5, quadriplegia or premoribund state; 6, death due to EAE.

Splenocyte cultures and restimulation

Spleens from 5–10 week-old TCR-tg mice were removed and single-cell splenocyte suspensions prepared. Tg splenocytes were cultured at 1 × 10⁶ cells/well with irradiated B10.PL splenocytes at 5 × 10⁶/well. Th1 effector cells were generated by adding MBP Ac1-11 (2 µg/ml) and IL-12 (0.5 ng/ml) (R&D Systems, Minneapolis, MN) to naïve splenocyte preparations. After 72 h of primary stimulation, cells were washed with PBS and rested in media alone with irradiated B10.PL splenocytes for 96 h. Cells were then restimulated with MBP Ac1-11 (1 µg/ml) alone or in combination with TGF-β (1 ng/ml) (R&D Systems, Minneapolis, MN) for 48 h, followed by another resting period of 96 h in media alone. This rest/restimulation protocol was repeated for either two or three rounds as indicated in each experiment.

Cytokine ELISA

Supernatants were collected for each culture condition described and analyzed for IFN-γ, IL-10, and IL-17 as previously described [17]. Supernatants were diluted to ensure values were within the standard curve. ELISA was performed using purified anti-mouse capture antibodies and biotinylated rat anti-mouse detection antibodies (BD Biosciences, San Jose, CA). Cytokine concentrations were calculated by generating a standard curve using recombinant proteins (R&D Systems, Minneapolis, MN) and analyzed using SoftMax Pro Software (Molecular Devices, Sunnyvale, CA).

Flow cytometry and intracellular cytokine staining

Flow cytometric analysis was performed to evaluate cytokine production (IFN-γ, IL-10, and IL-17) and T-bet and Foxp3 expression as previously described [17]. In each culture condition, GolgiPlug (1 µg/ml; BD Biosciences) was used to block cytokine secretion for the last 4 h prior to staining. 10⁵ live cell events were acquired on a FACSCanto II (BD biosciences) and analyzed using FlowJo software (Tree Star Inc.). The following antibodies were purchased from BD biosciences: APC-conjugated anti-IFN-γ, PE-conjugated anti-IL-17, PE-conjugated anti-IL-10, PerCP-conjugated anti-CD4, V450-conjugated anti-CD11b, PerCP-conjugated anti-CD45, PE-conjugated anti-IL-10R, FITC-conjugated anti-Vβ8.2, and PE-conjugated anti-Vα2.3. FITC-conjugated anti-T-bet was purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and APC-conjugated anti-Foxp3 and eFlour-450-conjugated anti-CD44 were purchased from eBioscience (San Diego, CA).

Chromatin-immunoprecipitation (ChIP) assay

For ChIP experiments, the EZ-ChIP kit (Millipore) was used and manufacturer’s protocol was followed. Rabbit anti-mouse Smad4 Ab and Rabbit IgG control Ab (Cell Signaling Technologies) were used for the IP. The primer set for the Smad4 site for PCR amplification of the IL-10 promoter was as follows: forward, ACAGCCCGGGAGTGTACCCTC and reverse TGTCCCAGCCTTGGAGACGTGT.

Statistical Analysis

GraphPad Prism software (GraphPad, La Jolla, CA) was used to perform all statistical analyses. The Mann-Whitney U test was performed for all clinical EAE experiments by analyzing each mouse at each time-point. This is a nonparametric test that accounts for the fact that EAE scores are ordinal and not interval scaled. Other statistical analyses were performed using a two-tailed Student t test. A p value < 0.05 was considered significant.