Tim-1 stimulation of dendritic cells regulates the balance between effector and regulatory T cells

Summary

We show that Tim-1, initially reported to be expressed on CD4⁺ T cells, is constitutively expressed on dendritic cells (DC) and that its expression further increases after DC maturation. Tim-1 signaling into DC upregulates costimulatory molecule expression and proinflammatory cytokine production, thereby promoting effector T cell responses, while inhibiting Foxp3⁺ Treg responses. By contrast, Tim-1 signaling in T cells only regulates Th2 responses. Using a high-avidity/agonistic anti-Tim-1 antibody as a co-adjuvant enhances the immunogenic function of DC, decreases the suppressive function of Treg cells, and substantially increases proinflammatory Th17 responses in vivo. The treatment with high-but not low-, avidity anti-Tim-1 not only worsens experimental autoimmune encephalomyelitis (EAE) in susceptible mice but also breaks tolerance and induces EAE in a genetically resistant strain of mice. These findings indicate that Tim-1 has an important role in regulating DC function, and thus shifts the balance between effector and regulatory T cells towards an enhanced immune response. By understanding the mechanisms by which Tim-1 regulates DC and T cell responses, we may clarify the potential utility of Tim-1 as a target of therapy against autoimmunity, cancer and infectious diseases.

Introduction

The T-cell immunoglobulin mucin (Tim) family of proteins are expressed on various immune cells and regulate immune responses [1-3]. Tim-1 was first identified as a hepatitis A virus cellular receptor 1 (HAVCR1) [4, 5] and later as a kidney injury molecule, KIM-1 [6, 7]. Interestingly, HAV infection is associated with a reduced risk of developing asthma, and Tim-1 has been genetically linked to Th2-driven murine airway hypersensitivity [8, 9]. Similarly, allelic variants of TIM-1 in humans have been associated with susceptibility to asthma and other atopic diseases as well as susceptibility to autoimmune diseases, suggesting that Tim-1 may have a role in regulating both autoimmune and allergic diseases [10]. In the immune system, Tim-1 is expressed on CD4⁺ T cells upon activation [11]. Under polarizing conditions, its expression was sustained preferentially on Th2 cells, but not on Th1 or Th17 cells [11-13]. Recent studies suggest that a small portion of B cells express Tim-1 which may serves as a marker for germinal centre B cells [14, 15]. Initial studies suggested that Tim-1 on T cells is a positive regulator of T cell activity. Crosslinking of Tim-1 with an agonistic anti-Tim-1 mAb (clone 3B3) or with its ligand, Tim-4, costimulated T cell proliferation [11, 12]. Furthermore, we have shown that this agonistic anti-Tim-1 mAb enhances both CD3 capping and T cell activation [16], suggesting that Tim-1 might be intimately involved in regulating TcR-driven activation. Indeed, it has been reported that human TIM-1 physically associates with the TcR/CD3 complex and upregulates activation signals [17]. This agonistic anti-Tim-1 mAb prevented the development of respiratory tolerance and increased pulmonary inflammation by substantially increasing production of IL-4 and IFN-γ [11]. The same antibody enhanced both pathogenic Th1 and Th17 responses in vivo and worsened experimental autoimmune encephalomyelitis (EAE) in an autoimmune disease setting [16]. Since this anti-Tim-1 mAb increased Th2 responses in vitro [11], but enhanced both Th1 and Th17 responses in vivo [11, 16], this raised the issue of whether Tim-1 might be expressed on other cells besides T cells, which could explain these differences in T cell responses.

Here we report that Tim-1 is constitutively expressed on DC. Using agonistic anti-Tim-1 mAb, we show that Tim-1 signaling promotes the activation of DC, which subsequently enhanced effector T cell responses, but inhibited Foxp3⁺ Treg responses. In an autoimmune disease setting, when given with immunogen, agonistic anti-Tim-1 mAb not only worsened EAE in disease-susceptible mice but also abrogated resistance and induced EAE in genetically resistant mice. Collectively, our findings show that Tim-1 is constitutively expressed on DC and Tim-1 signaling in DC serves to decrease immune regulation by Treg cells and to promote effector T cell responses.

Results

Constitutive expression of Tim-1 on DC.

To test our hypothesis that Tim-1 may be expressed on and affect the function of other cell types than T cells, we examined different populations of immune cells for Tim-1 expression ex vivo. As shown in Figure 1A, Tim-1 expression was low or undetectable on unactivated CD4⁺ or CD8⁺ T cells, B cells (CD19⁺), or macrophages (CD11b⁺CD11c⁻). Consistent with our previous report [11], activated CD4⁺ T cells exhibited high levels of Tim-1 expression, but bone marrow-derived DC cultured with GM-CSF showed little expression of Tim-1 even after LPS stimulation (Figure S1). Interestingly, we found that over 50% of ex vivo purified splenic DC (CD11c⁺) constitutively expressed Tim-1 (Figure 1A). While all DC subsets studied expressed Tim-1, the relative intensity of Tim-1 expression was higher on myeloid (CD11b⁺) DC and lower on plasmacytoid (B220⁺) DC (Figure 1B). Although culturing cells overnight with media alone upregulated Tim-1 expression on DC, activation by TLR signals (LPS/CpG) further increased Tim-1 expression on DC (Figure 1C).

Figure 1. DC constitutively express Tim-1.

(A) Different cell populations in the spleens of SJL mice were isolated and stained with anti-Tim-1 (thick lines) or rIgG2a (thin lines). (B) Purified CD11c⁺ cells were stained for Tim-1 expression and for the indicated surface molecules. The gating strategy of different DC subsets is shown. Right panel shows the MFI (shown as mean ± SEM of four experiments) of DC subsets stained with rIgG2a and anti-Tim-1. * P < 0.01, Student’s t test. (C) Purified CD11c⁺ cells were cultured with LPS/CpG overnight, and then stained for Tim-1 expression. In B and C, numbers in parentheses represent MFI; numbers without parentheses represent percentage. (D) SJL mice were immunized with PLP_139-15/CFA to induce EAE. At the peak of the disease, CNS-infiltrating cells were isolated and stained with Tim-1 (thick lines) or rIgG2a (grey areas). Stained cells were analyzed by flow cytometry. Data represent Tim-1 expression on gated populations and are representative of five experiments for (A), four experiments for (B) and (C), three experiments for (D).

We also analyzed Tim-1 expression on various immune cell populations isolated from the central nervous system (CNS) at the peak of EAE. Interestingly, CD4⁺ and CD11b⁺ cells showed little Tim-1 expression, whereas the majority of CD11c⁺ cells clearly showed Tim-1 expression on the surface (Figure 1D), suggesting that under autoimmune inflammatory conditions, DC are the major Tim-1-expressing population in CNS-infiltrating immune cells.

Tim-1 signaling in DC promotes DC maturation.

To examine whether Tim-1-crosslinking could induce signaling into DC, we measured NF-κB activity in DC after treatment with anti-Tim-1 antibodies. Treatment with agonistic/high-avidity anti-Tim-1 mAb 3B3 increased NF-κB activity in DC in a dose-dependent manner (Fig. 2A). In contrast, treatment with low-avidity anti-Tim-1 mAb RMT1-10 [16] did not change NF-κB activity (Figure 2A), although treatment with RMT1-10 changed T cell responses [16]. As a positive control, treatment with LPS/CpG increased NF-κB activity in DC.

Figure 2. Tim-1 signaling increases NF-κB activity in DC and promotes the expression of costimulatory molecules and production of proinflammatory cytokines.

(A) DC were incubated with the indicated stimuli for 2 h. Cell nuclear extracts were prepared to determine the NF-κB p65 activity (shown as mean ± SEM of triplicates in one experiment) using the TransAM NF-κB p65 Assay Kit. Data are representative of three experiments. (B) DC were treated with the indicated stimuli overnight, stained for surface molecules, and analyzed by flow cytometry. ΔMFI = MFI of indicated surface marker staining on gated CD11c⁺ populations - MFI of isotype staining. (C) The overnight culture supernatants from (B) were collected and assessed by cytometric bead assay (CBA) for IFN-γ, TNF-α, and IL-6. (D) DC were incubated with the indicated stimuli for 6 h. Cells were collected and total RNA was isolated for determining the expression of the indicated cytokine genes. Data in (B-D) are mean ± SEM of four experiments. * P < 0.01, Student’s t test.

Because NF-κB is a key transcription factor responsible for DC activation [18, 19], we next examined whether Tim-1 signaling could induce DC maturation in terms of the expression of surface molecules and the production of cytokines. Compared to the control rIgG2a treatment, treatment with agonistic/high avidity anti-Tim-1 3B3 resulted in marked upregulation of MHC class II, CD80 and CD86 on DC, (Figure 2B). As a positive control, LPS plus anti-CD40 resulted in maximal expression of all molecules on treated DC. Furthermore, Tim-1 signaling into DC enhanced production of proinflammatory cytokines IFN-γ, TNF-α, and IL-6 as determined by both cytometric bead array and real-time PCR (Figure 2C and 2D). Moreover, treatment with 3B3 anti-Tim-1 increased the expression of IL-1β and IL-23 (p19/p40), but did not significantly alter the expression of IL-12 p35, TGF-β or IL-10 (Figure 2D). Since low-avidity anti-Tim-1 mAb RMT1-10 did not trigger Tim-1 signaling in DC, treatment with RMT1-10 neither increased expression of MHC Class II, CD80 or CD86 nor enhanced production of IFN-γ, TNF-α or IL-6 (Figure 2B and 2C). As a positive control, LPS/CpG increased the production of all tested cytokines in DC.

Tim-1 signaling activated DC promotes effector T cell responses and suppress Foxp3⁺ Treg generation.

Since cytokines that promote differentiation of Th1 (e.g. IFN-γ) and Th17 cells (e.g. IL-6, IL-23, and IL-1β) were predominantly increased from DC after treatment with high-avidity anti-Tim-1 3B3, we then examined whether DC activated with 3B3 would affect Th cell differentiation/responses. Because activated CD4⁺ T cells and DC both express Tim-1, we first tested the effect of Tim-1 crosslinking on CD4⁺ T cells in an APC free system. In an APC free culture, activation with anti-CD3/anti-CD28 in the presence of 3B3 anti-Tim-1 increased the frequency of IL-4- and IL-10-producing CD4⁺ T cells, while the treatment did not significantly change IFN-γ⁺ and IL-17⁺ T cells (Figure 3A). However, when naïve CD4⁺ T cells were cultured with syngeneic DC plus antigen together with 3B3, the responding T cells produced more IFN-γ and IL-17, in addition to IL-4 and IL-10 (Figure 3A). Interestingly, in the absence or presence of DC, RMT1-10 increased only Th2 responses (IL-4 and IL-10 production) but had no obvious modification on Th1 (IFN-γ) or Th17 (IL-17) responses, suggesting that the low-avidity anti-Tim-1 RMT1-10 does not modulate DC function (Figure 2). These data suggest that Tim-1 crosslinking with both high-avidity and low-avidity anti-Tim-1 promotes Th2 responses regardless of the presence or absence of DC. However, only the high-avidity anti-Tim-1 enhances Th1 and Th17 responses when DC are present in the cultures.

Figure 3. Tim-1 crosslinking induces differential T cell responses in the absence or presence of DC.

Naïve CD4⁺ T cells from 5B6 TcR Tg SJL mice were activated with either plate-bound anti-CD3/anti-CD28 or with PLP_139-151 plus purified syngeneic DC in the (A, B) absence or (C, D) presence of TGF-β. (A) Cytokine production from T cells was determined on d6 by intracellular cytokine staining. Number in upper right quadrants indicates the percentage of cytokine-producing cells in CD4⁺ populations. (B) DC purified from SJL mice were cultured with PLP_139-151 plus anti-Tim-1, rIgG2a or LPS for 6 h. Cells were then washed with PBS and subcutaneously transferred into SJL mice on d0 and d7. Draining LN cells were isolated on d14 and restimulated with PLP_139-151. After 4 days, cytokine production was measured by intracellular cytokine staining. (C) Naïve 5B6 CD4⁺ T cells were activated with either plate-bound anti-CD3/CD28 or with PLP_139-151 plus purified syngeneic DC or subsets in the presence of TGF-β. Cells were then measured for CD103 and Foxp3 expression on CD4⁺ T cells by flow cytometry. (D) Supernatant from the cell cultures described in (C) was collected after 48 h and measured for IL-17 production by ELISA. Data are mean ± SEM of triplicates in one experiment. * P <0.01, Student’s t test. All above data are representative of three experiments.

To demonstrate that Tim-1 signaling in DC is responsible for promoting Th1 and Th17 responses in vivo, PLP_139-151-loaded/anti-Tim-1-treated DC were subcutaneously transferred into syngeneic SJL mice. Draining LN cells were then isolated and antigen-specific T cell responses were examined ex vivo. We found that immunization with 3B3-treated DC enhanced the production of IFN-γ and IL-17 as well as IL-4 and IL-10 in PLP_139-151-responding T cells, while immunization with RMT1-10-treated DC seemed not to significantly modulate any of these cytokines (Figure 3B). LPS-treated DC enhanced production of IFN-γ and IL-17 but strongly inhibited IL-4 and IL-10 from T cells (Figure 3B). There was no detectable production of these cytokines in the absence of antigen in any case (data not shown). These data further confirmed that only the high-avidity anti-Tim-1 induces DC activation, and Tim-1 signaling-activated DC promote Th1 and Th17 as well as Th2 responses.

TGF-β acts on naïve T cells to induce Foxp3 expression and these cells attain most of Treg properties. Addition of 3B3 anti-Tim-1 in the presence of either CD11b⁺ or CD11b⁻ DC to cultures where TGF-β was used to induce Foxp3⁺ Treg cells, led to the inhibition of Foxp3⁺ Treg generation. The frequency of Foxp3⁺ Treg cells upon 3B3-treatment of CD11b⁻ DC was only about 4% compared to about 40% induction under control conditions (Figure 3C). However, addition of 3B3 in APC-free cultures did not significantly change Foxp3⁺ Treg generation, with about 70% of Foxp3⁺ cells regardless of whether anti-Tim-1 was used. However, 3B3 treatment increased CD103 expression on both Foxp3⁺ and Foxp3⁻ T cells (Figure 3C). Furthermore, treatment with 3B3 increased the production of IL-17 from T cells in the presence of DC (Figure 3D). These data suggest that Tim-1 signaling in DC, but not in T cells, leads to the inhibition of Foxp3⁺ Treg generation in association with increased Th17 responses. The observation that 3B3-activated DC produced IL-6 and IL-23 (Figure 2C an 2D) at least partly explains the inhibition of Foxp3 induction, as blocking IL-6 and IL-23 in the Treg cultures restored Foxp3 expression and inhibited IL-17 production (Figure S2).

High-avidity anti-Tim-1 antibody breaks tolerance and induces EAE in genetically resistant mice.

We have reported that i.p. injection of 3B3 worsened EAE in SJL mice immunized with PLP_139-151/CFA emulsion [16]. However, the systemic administration would allow the antibody access to many types of cells that express Tim-1 and thus could affect their function and the disease. Therefore, to directly assess a role for Tim-1 signaling on DC function, we immunized mice with PLP_139-151/CFA emulsion containing anti-Tim-1. We reasoned that DC, at the frontline of pathogen recognition, would most likely be the first major population affected by anti-Tim-1 in the emulsion. In this approach, anti-Tim-1 was not detectable in the sera from the mice (data not shown), indicating antibodies remained at the local administration sites. Interestingly, draining LN cells from mice treated with high-avidity anti-Tim-1 3B3 in emulsion showed both higher basal and Ag-dependent proliferation in the responding T cells (Figure 4A) and an increased frequency of IFN-γ- and IL-17-producing CD4⁺ T cells (Figure 4B). The treatment consistently resulted in more severe and accelerated EAE compared to the control group (Figure 4C and Table I), while inclusion of low-avidity anti-Tim-1 RMT1-10 did not change the course of EAE (Figure S3). These data suggest that the high-avidity anti-Tim-1 in the emulsion during the induction of EAE enhances the immunogenic function of DC, which then increase pathogenic Th1 and Th17 responses resulting in worsened disease in SJL mice.

Figure 4. High-avidity anti-Tim-1 as a co-adjuvant enhances antigen-specific Th1 and Th17 responses and worsens EAE in SJL mice.

(A-C) SJL mice were immunized with PLP_139-151/CFA/3B3 emulsion to induce EAE (25 μg of PLP_139-151 per mouse). Draining LN cells were isolated from one group of mice on d10 and restimulated with PLP_139-151. (A) Proliferation and (B) cytokine production were measured by ³[H]-thymidine incorporation and intracellular cytokine staining, respectively. Data are mean ± SEM (A) and representative (B) of six mice in each group. (C) Clinical EAE score in another group of six mice was evaluated daily. * P < 0.01, Fisher’s exact test.

Table I.

EAE in 3B3 anti-Tim-1- and control rIgG-treated SJL mice

Treatment	Incidence	Mean day of onset (mean ± SEM)	Mean Maximum score (mean ± SEM)
rIgG	8 of 8 (100%)	12.8 ± 2.76	2.25 ± 0.46
Anti-Tim-1	8 of 8 (100%)	9.75 ± 0.89a	3.00 ± 0.65a

^aP<0.05, anti-Tim-1-treated group compared with rIgG-treated group

B10.S mice are congenic with SJL mice at MHC; however, in contrast to SJL mice, B10.S mice are resistant to EAE. Previous studies have suggested that EAE resistance in B10.S mice is in part due to a lower APC capacity to stimulate proinflammatory T cell responses against myelin self-antigens [20]. Furthermore, B10.S mice express relatively high levels of myelin-specific Foxp3⁺ Treg cells in their peripheral repertoire [21]. Since inclusion of 3B3 anti-Tim-1 in CFA enhanced pathogenic Th1/Th17 responses and exacerbated EAE in disease susceptible SJL mice, we asked whether the treatment would break tolerance and induce EAE in B10.S mice. In addition to having lower expression of MHC and costimulatory molecules [20], we observed that B10.S-derived DC produced much less proinflammatory cytokines, such as IL-6, upon LPS treatment than SJL-derived DC did. However, treatment with 3B3 anti-Tim-1 alone or together with LPS restored IL-6 production from B10.S-derived DC to the level from SJL-derived DC treated with LPS (Figure S4).

Next, B10.S mice were immunized with PLP_139-151/CFA emulsion containing 3B3 anti-Tim-1, and the development of pathogenic effector T cell responses was analyzed. In addition to higher basal proliferation, draining LN cells from B10.S mice immunized with 3B3/PLP_139-151/CFA showed much higher proliferation upon antigen restimulation (Figure 5A). The treatment dramatically enhanced both IFN-γ- and IL-17-producing CD4⁺ T cells, while the treatment did not increase IL-4/IL10-producing T cells (Figure 5B). Consistently, the 3B3-treated mice became susceptible to the development of EAE, with over 70% of B10.S mice developing EAE (Figure 5C and Table II).

Figure 5. High-avidity anti-Tim-1 as a co-adjuvant breaks tolerance and induces EAE in disease-resistant B10.S mice.

(A-C) B10.S mice were immunized with PLP_139-151/CFA/3B3 emulsion to induce EAE (80 μg of PLP_139-151 per mouse). Draining LN cells were isolated on d10 in one group of six mice and restimulated with PLP_139-151. (A) Proliferation (Mean ± SEM) and (B) cytokine production were measured by ³[H]-thymidine incorporation and intracellular cytokine staining, respectively. * P < 0.01, Student’s t test. (C) Clinical EAE score (shown as mean ± SEM) in another group of six mice were evaluated daily. # P < 0.05; * P < 0.01, Fisher’s exact test.

Table II.

EAE in 3B3 anti-Tim-1- and control rIgG-treated B10.S mice

Treatment	Incidence	Mean day of onset (mean ± SEM)	Mean Maximum score (mean ± SEM)
rIgG	0 of 6 (0%)	-	-
Anti-Tim-1	6 of 8 (75%)	19.2 ± 2.48a	1.83 ± 0.41a

^aRepresents the mice that showed clinical signs of EAE (6 of 8 mice)

Effect of high-avidity anti-Tim-1 on DC and effector and regulatory T cells during EAE.

To further examine the effect of high-avidity anti-Tim-1 as a co-adjuvant on DC and effector and regulatory T cells, we generated B10.S Foxp3/GFP “knock-in” mice. The “knock-in” mice were immunized with 3B3 or control rIgG in immunogenic emulsion. DC, Foxp3⁻CD4⁺ effector T cells (Teff) and Foxp3⁺CD4⁺ Treg were isolated from spleen and lymph nodes of the mice and analyzed in Criss-cross proliferation assays (Figure 6A). Teff from 3B3-treated mice showed stronger proliferation and produced higher levels of IFN-γ and IL-17 upon antigen restimulation than Teff from rIgG-treated mice. More interestingly, DC from 3B3-treated mice induced higher Teff proliferation and IFN-γ and IL-17 production than DC from rIgG-treated mice (Figure 6A). The frequency of Foxp3⁺ Treg in spleens, lymph nodes, or the CNS was not significantly affected by 3B3 treatment (Figure 6D and data not shown). However, Foxp3⁺ Treg from 3B3-treated mice were less efficient in suppressing Teff proliferation in the cultures where Foxp3⁻ Teff and DC were obtained from rIgG-treated B10.S mice (Figure 6B). Phenotypically, 3B3 in PLP_139-151/CFA emulsion promoted DC activation as the treatment significantly upregulated the intensity of costimulatory molecules CD80 and CD86 and MHC class II (Figure 6C).

Figure 6. Effect of high-avidity anti-Tim-1 as a co-adjuvant on DC, Teff, and Treg in B10.S mice during EAE.

B10.S Foxp3/GFP “knockin” mice were immunized with PLP_139-151/CFA/3B3 emulsion. At the peak of the disease, Teff (CD4⁺Foxp3/GFP⁻), Treg (CD4⁺Foxp3/GFP⁺) and DC were purified from spleens and lymph nodes of mice treated with 3B3 anti-Tim-1 or control rIgG, and CNS-infiltrating mononuclear cells were isolated from brains and spinal cords of the mice. (A) Different mixtures of Teff and DC as indicated were cultured in the presence of antigen, and cell proliferation and cytokine production in the culture supernatant were measured by ³[H]-thymidine incorporation and ELISA, respectively. (B) Teff and DC from rIgG-treated mice were co-cultured with different ratios of Treg (as indicated) in the presence of antigen, and cell proliferation was measured by ³[H]-thymidine incorporation. Data in (A) and (B) are mean ± SEM of triplicates for one mouse and representative of four mice in each group. * P < 0.05, Student’s t test. (C) DC were stained and analyzed for CD80, CD86, and MHC class II expression. Numbers in Italic font represent MFI of rIgG-treated samples; Numbers in the normal font represent MFI of 3B3-treated samples; (D) CNS-infiltrating cells were stained to determine the percentage of DC, CD4⁺ T cells, and Foxp3⁺ Treg by flow cytometry. Cytokine production of CNS-infiltrating Foxp3⁺ and Foxp3⁻ CD4⁺ T cells were determined by intracellular staining.

In the CNS, treatment with the high-avidity anti-Tim-1 resulted in more mononuclear cell infiltration, containing high frequencies/numbers of CD11c⁺ DC and CD4⁺ T cells (Figure 6D and data not shown). Although the frequency of CD4⁺Foxp3⁺ Treg in 3B3-treated mice was not dramatically decreased, significantly more Foxp3⁺ Treg in the CNS of 3B3-treated mice produced proinflammatory cytokine IL-17 (7.85 ± 2.36% from 3B3-treated mice vs.1.85 ± 0.96% from rIgG-treated mice, n=3; P < 0.05). In addition, the frequency of CNS-infiltrating CD4⁺Foxp3⁻ Teff cells producing IFN-γ and/or IL-17 was also increased in 3B3-treated mice (Figure 6D). Moreover, similar to the observation in Figure 5B, control rIgG-treated B10.S mice showed a very low percentage of IL-17-producing Teff cells in the CNS, which was dramatically increased by the high-avidity anti-Tim-1 treatment (Figure 6D).

Discussion

DC are professional APC with a remarkable capacity to activate naive T cells and prime T cell responses, therefore providing a link between innate and adaptive immunity. Here we show that Tim-1, a molecule initially identified on CD4⁺ T cells and regulating Th2 responses, is constitutively expressed on DC and regulate their function. We confirmed that Tim-1 signaling in T cells mainly serves as a Th2 regulator with no noticeable effect on Th1 or Th17 response. However, under Th1 or Th17 polarization conditions, the high-avidity anti-Tim-1 does not enhance Th2 responses regardless of the presence of DC, while under Th2 conditions, the treatment further increases Th2 cytokine production (Figure S5), suggesting that the positive effects on Th2 responses downstream of Tim-1 signaling in T cells can be inhibited in environments favoring Th1/Th17 development.

The high-avidity, but not low-avidity, anti-Tim-1 induced NF-κB activity in DC, suggesting that Tim-1 binding avidity could be responsible for triggering Tim-1 signaling in DC. Because NF-κB is a key transcription factor responsible for DC activation and production of many DC-derived cytokines [18, 19], this suggests that Tim-1 signaling drives DC maturation at least in part by inducing NF-κB activity. A study suggests that Tim-1 signaling in T cells induces Th2 responses by increasing the activity of NFAT/AP-1 but not NF-κB [22]. This indicates that Tim-1 signaling induces distinct events in innate and adaptive immune cells.

Tim-1 signaling activated DC enhance both innate and adaptive immunity by producing innate cytokines and upregulating costimulatory molecules and antigen-presenting capability. Specifically, due to their production of the proinflammatory cytokines IL-6, IL-23 and IL-1, Tim-1-activated DC enhance Th17 responses and inhibit Foxp3⁺ Treg generation. These cytokines have all been shown to promote Th17 responses [23, 24].

Treg cells play an important role in immune suppression and tolerance [25]. Tim-1-activated DC inhibited TGF-β-mediated Foxp3⁺ Treg generation accompanied by an increased Th17 response. This is at least partly due to proinflammatory cytokines produced by Tim-1-activated DC, such as IL-6 and IL-23 (Figure S2), which have been reported to inhibit the development and function of Treg and promote Th17 responses [26, 27]. It has been reported that 3B3 anti-Tim-1 reduced Foxp3 expression and suppressive function when Foxp3⁺ Treg cells were activated with allogeneic DC [28], but at the time, it was assumed that the observed effects were directly on T cells. We now provide evidence that these effects are due to Tim-1 signaling in DC. While Tim-1 signaling in DC affects the generation and function of Foxp3⁺ Treg cells, Tim-1 signaling in T cells has discernable effects on Treg cells (Figure 3). Although Tim-1 signaling in T cells does not directly affect Foxp3⁺ Treg generation, it alters T cell expression of CD103, a molecule mainly involved in cell migration [29], indicating that Tim-1 signaling in T cells may affect T cell trafficking in addition to T cell differentiation.

EAE is a Th1/Th17 cell-mediated autoimmune inflammatory disease that affects the CNS [30]. DC are essential antigen-presenting cells for inducing proinflammatory Th1 and Th17 responses and are also a prominent component of CNS-infiltrating cells during the development of EAE [31-33]. Interestingly, at the peak of EAE severity, DC in the CNS, but not CD4⁺ T cells, express Tim-1 (Figure 1D). When the CNS-infiltrating mononuclear cells were restimulated with antigen, the addition of high-avidity anti-Tim-1 to the cultures strongly enhanced IL-17 production with a more moderate increase in IFN-γ production (Figure S6). Since only CNS-infiltrating DC express Tim-1 at this stage, it suggests that DC activated via Tim-1 during the autoimmune reaction enhance proinflammatory Th1/Th17 responses. Indeed, inclusion of high-avidity, but not low-avidity, anti-Tim-1 as co-adjuvant in immunogen enhanced antigen-specific Th1/Th17 responses and worsened EAE in disease-susceptible SJL mice (Figure 4 and Figure S4). Strikingly, high-avidity anti-Tim-1 as co-adjuvant also broke tolerance and induced EAE in B10.S mice. B10.S mice are resistant to the induction of EAE associated with defect in APC function [20], high frequency of PLP_139-151-specific Treg [21], and impaired Th17 responses (Figure 5 and 6). Tim-1 signaling in DC appears to rescue these defects in B10.S mice and make these mice susceptible to EAE.

Our data help to explain why administration of an agonistic/high-avidity anti-Tim-1 increased both Th2 and Th1 responses in an animal model of asthma [11]. In addition to the direct effect of Tim-1 signaling in T cells which could have upregulated Th2 responses, Tim-1 signaling in DC could have induced factors (e.g. proinflammatory cytokines) that decreased the suppressive function of Treg cells and promoted Th1 and Th17 as well as Th2 responses in the animal model of asthma.

Although Tim-1 signaling activated DC promote Th1/Th17 responses and inhibited Foxp3⁺ Treg generation, they also promote Th2 responses. Since Th2 responses prevent EAE [34], immunization with PLP_139-151-loaded DC activated with high-avidity anti-Tim-1 3B3 or inclusion of 3B3 in PLP_139-151/IFA emulsion did not induce EAE in SJL mice (data not shown). However, mycobacteria products contain many TLR ligands (e.g. LPS for TLR4) and are the components of CFA for the activation of innate immune cells [18], and LPS-treated DC induced Th1 and Th17 responses but strongly inhibited Th2 responses (Figure 3B). Therefore, when the high-avidity anti-Tim-1 is included in PLP_139-151/CFA emulsion to induce EAE, Tim-1 signaling and TLR signaling together synergistically increases the immunogenic functions of DC (e.g. upregulating expression of MHC and costimulatory molecules and production of proinflammatory cytokines), which subsequently decrease Treg suppression, inhibit Th2 responses, and induce potent pathogenic Th1 and Th17 responses and thus drive EAE in B10.S mice and enhance EAE in susceptible SJL mice.

Tim-1 has recently been shown to be involved in the clearance of apoptotic cells by binding to phosphatidylserine (PS) [35, 36]. Since apoptotic cells can induce inhibitory cytokines and promote tolerance induction [37], it is possible that the promotion of inflammation by the high-avidity anti-Tim-1 could be due to the inhibition of the Tim-1/PS interaction and thus prevent the induction of anti-inflammatory cytokines induced by apoptotic cells. We also reported that Tim-4 could bind to Tim-1 and regulate T cell responses [12]. Interestingly, treatment with Tim-4-hFc fusion proteins did not change DC function in terms of the expression of CD80, CD86, and MHC class II molecules (Figure S7). However, Tim-4 also binds to PS [35, 36] and potentially another unknown receptor [38]. Thus, without knowing whether DC express other Tim-4-binding protein(s) in addition to Tim-1, it is difficult to understand whether the effect of Tim-4-hFc on DC is through Tim-1 and/or other pathway(s). These issues will only be clearly addressed using Tim-1 deficient mice, which just became available most recently [15].

In summary, we show that Tim-1 plays different roles in the innate and adaptive immune responses. Since Tim-1 is constitutively expressed on DC in the steady state, Tim-1 is readily available for crosslinking on DC before it is even expressed on adaptive immune cells. The present study highlights the role of Tim-1 expressed on DC in regulating the balance between effector and regulatory T cells and thus regulating immune responses. A better understanding of the mechanism by which Tim-1 regulates DC and T cell responses will provide a target by which DC/T cell functions can be regulated so as to treat inflammatory diseases including autoimmune diseases, and to improve vaccination and tumor immunotherapy.

Materials and Methods

Mice and reagents.

SJL mice were purchased from The Jackson Laboratory. B10.S mice and 5B6 SJL mice transgenic for the PLP_139-151-specific TCR 5B6 have been described previously [20]. Foxp3/GFP “knock-in” mice originally generated on the C57BL/6 background [26] were back-crossed for >10 generations onto the B10.S background. The mice were maintained, and all animal experiments were performed according to the animal protocol guidelines of Harvard Medical School. PLP_139-151 and OVA_323-339 peptides were synthesized by Quality Controlled Biochemicals. Anti-Tim-1 antibodies 3B3 and RMT1-10 have been described previously [11, 16]. Cytokines and antibodies for FACS and ELISA were obtained from eBioscience, BD Biosciences, and R&D Systems.

Cell purification.

Different populations of immune cells were purified with MACS beads (Miltenyi Biotec). Naïve CD4⁺ T cells (CD4⁺CD62L^hiCD25⁻) and DC (CD11c⁺CD3⁻CD19⁻) were purified using a FACSAria cell sorter following MACS bead-isolation of CD4⁺ and CD11c⁺ cells, respectively. CNS-infiltrating mononuclear cells were isolated from mice with EAE as previously described [26, 27].

T Cell stimulation and cytokine measurement.

Naïve CD4⁺ cells (1×10⁶/well) were activated with either plate-bound anti-CD3/CD28 (1 μg/ml for both) or with PLP_139-151 (25 μg/ml) plus syngeneic DC (2×10⁵/well) in the presence or absence of anti-Tim-1 (10 μg/ml). Cytokine production was determined by ELISA or intracellular staining assays [27, 39].

Preparation of nuclear extract and colorimetric NF-κB assay.

DC were stimulated with different doses of anti-Tim-1 or rIgG2a, or LPS (200 ng/ml) plus CpG (500 nM), and nuclear extracts were prepared using a nuclear extract kit (Active Motif, Carlsbad, CA). NF-κB/DNA binding activity was detected using a TransAM NF-κB p65 transcription factor assay kit (Active Motif) according to the manufacturer’s protocol.

DC activation and cytokine measurement.

DC were treated with anti-Tim-1 (10 μg/ml), rIgG2a (10 μg/ml), or LPS/CpG. After overnight culture, supernatants were collected and total RNA was extracted from DC using RNeasy Plus Mini kit (Qiagen) according to the manufacturer’s instructions. Production of cytokines in the supernatants was measured by cytometric bead array (CBA, BD Biosciences) according to the manufacturer’s instructions. Levels of cytokine mRNA expression in DC were determined by real-time PCR as previously described [27]. The data were expressed as expression relative to β-actin. Primers and probes for TNF-α, IFN-γ, TGF-β, IL-1β, IL-10 and β-actin were purchased from Applied Biosystems. Primers for IL-23p19, IL-12p35 and p40 have been described previously [40].

Immunization and proliferation assay.

Female SJL and B10.S mice (8-12-wk old) were immunized subcutaneously in the flanks with an emulsion containing PLP_139-151 and anti-Tim-1 or control rIgG (200 μg/mouse) in CFA. Pertussis toxin (100 ng/mouse, List Biological Laboratories) was administered intraperitoneally on days 0 and 2. EAE was evaluated as previously described [16].

For recall assay, draining lymph node cells were isolated from treated mice and plated in round-bottomed 96-well plates (BD Biosciences) in culture medium with various concentrations of antigen. For criss-cross proliferation assays and suppression assays, 50,000 Teff were cultured with the indicated number of Treg and 15,000 DC per well in the presence of PLP_139-151 (10 μg/ml). After 48 h, plates were pulsed for 16 h with 1 μCi [³H]-thymidine per well. Proliferation was measured as counts per minute by using a Wallac Liquid Scintillation Counter (Perkin Elmer).

Statistics

The clinical score and incidence of EAE were analyzed by Fisher’s exact test, and other comparisons were analyzed by Student’s t test. P < 0.05 was considered significant.