Tim-1 Is Essential for Induction and Maintenance of IL-10 in Regulatory B Cells and Their Regulation of Tissue Inflammation

This article is featured in In This Issue, p.1387

In addition to producing Abs and acting as positive regulators of immune responses by serving as APCs and producing cytokines for optimal T cell activation, B cells have also been shown to negatively regulate immune responses (1–6). Lack or loss of IL-10–producing regulatory B cells (Bregs) accelerates and exacerbates many autoimmune and inflammatory diseases, including experimental autoimmune encephalomyelitis (EAE), chronic colitis, arthritis, type 1 diabetes, lupus, and delayed type contact hypersensitivity. Alternatively, transfer or increase in the number of Bregs reduces autoimmune and inflammatory diseases (1–4, 6). In many models, IL-10 appears to be critical for the regulatory function of Bregs, although other mechanisms in addition to IL-10 production might also be operational for the regulatory function of Bregs (1–4, 6). Despite their critical role in regulating immune and autoimmune responses, lack of a universal marker for identifying Bregs has hampered our understanding of the critical biologic functions of Bregs. Furthermore, the processes and mechanisms by which Bregs are generated have not been identified.

T cell Ig and mucin domain (Tim)-1, a transmembrane glycoprotein, was identified as a member of the Tim family genes that regulates immune responses (7). In the immune system, Tim-1 was first identified to be expressed on T cells and dendritic cells where it plays an important role in regulating important cellular functions (7–10). More recently, Tim-1 has also been shown to be expressed on B cells (11, 12). The vast majority of Tim-1⁺ B cells produce IL-10, and transfer of Tim-1⁺ Bregs led to long-term acceptance of islet allografts and inhibited allergic airway responses (13). We have also demonstrated that B cell–derived IL-10 is produced mainly by Tim-1⁺ B cells (14). We generated a Tim-1 mutant mouse (Tim-1^∆mucin) and demonstrated that the mouse has a profound defect in B cell–derived IL-10 production. Associated with the loss of IL-10 production in B cells, 10- to 12-mo-old Tim-1^∆mucin mice showed increased effector/memory Th1 responses and autoantibody production without any systemic autoimmunity (14). These data supported the idea that Tim-1 may be critical for Breg function.

In this study, we demonstrate that Tim-1 is required for optimal IL-10 production in Bregs. B cells with Tim-1 deficiency or mutation show a defect in IL-10 production with an increase in proinflammatory cytokine production. In vitro, Tim-1–deficient B cells promote IL-17 and IFN-γ production in T cells and inhibit the generation of Foxp3⁺ regulatory T cells (Tregs) and type 1 regulatory T (Tr1) cells. In in vivo transfer models of EAE, hosts with Tim-1–deficient B cells developed more severe disease associated with increased generation of pathogenic Th1/Th17 cells and decreased Foxp3⁺ Treg frequency and IL-10 production in the CNS. In contrast, transfer of Tim-1⁺ Bregs but not Tim-1⁻ B cells reduced the incidence and severity of EAE. As a phosphatidylserine (PS) receptor, Tim-1 is essential for binding of apoptotic cells (ACs) to Bregs. Coculturing of B cells with ACs increased IL-10 production in wild-type (WT) but not Tim-1–deficient B cells. Furthermore, AC treatment reduces EAE in hosts with WT but not Tim-1–deficient B cells. Tim-1^∆mucin mice that progressively lose IL-10 in Bregs develop severe spontaneous inflammation in multiple organs with massive inflammatory cell infiltration at 16–18+ months of age.

C57BL/6 mice, Rag1^−/−, and IL-10^GFP reporter mice (only heterozygous mice were used; also known as Tiger) were purchased from The Jackson Laboratory. Tim-1^−/− and Tim-1^∆mucin mice were described (11, 14). Tim-1^−/− mice were bred with IL10^GFP reporter mice to obtain Tim-1^−/−IL-10^GFP mice. Mice were maintained and all animal experiments were done according to the animal protocol guidelines of Harvard Medical School. Myelin oligodendrocyte glycoprotein (MOG)_35–55 was synthesized by Quality Controlled Biochemicals. Cytokines and Abs for cell culture, flow cytometry, and cytometric bead array were obtained from BioLegend, eBioscience, BD Biosciences, and R&D Systems. Anti–Tim-1 Ab RMT1-4 (BioLegend) was used for flow cytometry. Anti–Tim-1 Ab 5F12 was described previously (14).

Splenic CD19⁺ B cells from 2- to 4-mo-old mice were purified using MACS columns following staining with anti-mouse CD19 MACS beads. Cells were cultured in round-bottom 96-well plates in the presence of anti–Tim-1 (clone 5F12), F(ab′)₂ fragment anti-IgM, anti-CD40, IL-21, or their combinations. After 3 d, IL-10 production in culture supernatants was measured by cytokine bead array. MACS-purified CD19⁺ B cells were labeled with PE–anti-Tim-1 (RMT1-4) and then separated into Tim-1⁺ and Tim-1⁻ B cells by FACS for further uses.

CD4⁺CD62L^hiCD25⁻ naive CD4⁺ T cells were purified by FACS after a MACS bead isolation of CD4⁺ cells as previously described (15, 16). Naive CD4⁺ cells were activated with either plate-bound anti-CD3 (2 μg/ml) and anti-CD28 (2 μg/ml) or B cells plus soluble anti-CD3 (1 μg/ml) under Th0 (no cytokine), Th1 (IL-12 plus anti–IL-4), Th2 (IL-4 plus anti–IL-12/anti–IFN-γ), Th17 (TGF-β1 plus IL-6), Tr1 (TGF-β1 plus IL-27), and inducible Treg (iTreg; TGF-β1) conditions. After 96 h, cells were collected for further experiments.

To isolated CNS-infiltrating mononuclear cells, mice were first perfused through the left cardiac ventricle with cold PBS. The forebrain and cerebellum were dissected and spinal cords flushed out with PBS by hydrostatic pressure. CNS tissue was cut into pieces and digested with collagenase D (2.5 mg/ml; Roche Diagnostics) and DNase I (1 mg/ml; Sigma-Aldrich) at 37°C for 30 min. Mononuclear cells were isolated by passing the tissue through a 70-μm cell strainer, followed by a 70/37% Percoll gradient centrifugation. Mononuclear cells were removed from the interphase, washed, and resuspended in culture medium for further analysis.

Thymocytes from C57BL/6 mice were treated with 1 μM dexamethasone (Sigma-Aldrich) for 6 h. After an extensive wash, ACs were used for in vivo injection. For the preparation of labeled ACs for in vitro experiments, thymocytes were first labeled with 0.2 μg/ml CellTracker Green 5-chloromethylfluorescein diacetate (CMFDA; Invitrogen) for 30 min at room temperature and then treated with 1 μM dexamethasone for 6 h. Annexin V and propidium iodide (BD Biosciences) staining was used to confirm apoptosis of thymocytes.

Splenocytes and lymph node cells (pooled superficial cervical, axillary, brachial, and inguinal lymph node cells) from mice were treated with ACK lysis buffer (Lonza) and then cell numbers were determined. Frequencies of immune cell subsets of splenocytes and lymph node cells were determined by flow cytometry using Abs to cell surface molecules. Purified B cells were incubated with CMFDA-labeled ACs for 2 h. After fixation and extensive wash to remove nonassociated ACs, cells were analyzed by flow cytometry and confocal microscopy.

For intracellular cytokine staining, cells were stimulated in culture medium containing PMA (30 ng/ml; Sigma-Aldrich), ionomycin (500 ng/ml; Sigma-Aldrich), and GolgiStop (1 μl/ml; BD Biosciences) in a cell incubator with 10% CO₂ at 37°C for 4 h. After surface markers were stained, cells were fixed and permeabilized with Cytofix/Cytoperm and Perm/Wash buffer (BD Biosciences) according to the manufacturer’s instructions. Then, cells were stained with fluorescence-conjugated cytokine Abs at 25°C for 30 min before analysis. 7-Aminoactinomycin D (BD Biosciences) was also included to gate out the dead cells. All data were collected on a FACSCalibur or an LSR II (BD Biosciences) and analyzed with FlowJo software (Tree Star).

Total CD4⁺ T cells were cotransferred together with CD19⁺ B cells into Rag1^−/− mice. Mice were immunized s.c. in the flanks with an emulsion containing MOG_35–55 (100 μg/mouse) and Mycobacterium tuberculosis H37Ra extract (3 mg/ml; Difco Laboratories) in CFA (100 μl/mouse). Pertussis toxin (100 ng/mouse; List Biological Laboratories) was administered i.p. on days 0 and 2. For AC treatment, ACs were i.v. injected 1 d before immunization. Mice were monitored and assigned grades for clinical signs of EAE as previously described (10, 17).

RNA was extracted with RNeasy Plus kits (Qiagen) and cDNA was made by iScript (Bio-Rad). All of the real-time PCR probes were purchased from Applied Biosystems. Quantitative PCR were performed using a ViiA 7 real-time PCR System (Applied Biosystems).

Tissues and organs from mice were fixed in 10% neutral buffered formalin for 12 h, processed, embedded in paraffin wax, sectioned, and stained with H&E using standard procedures. Evaluations were made in a blinded fashion.

The clinical score and incidence of EAE were analyzed by a Fisher exact test, and comparisons for cytokine bead array and real-time PCR results were analyzed by a Student t test. A p value <0.05 was considered significant.

Tim-1 has been shown to identify most of IL-10–producing Bregs (13, 14). We have previously reported generation of Tim-1^∆mucin mice, which express a loss-of-function form of Tim-1, because of deletion of the mucin domain (14). We demonstrated that the major defect in young (<6 mo old) Tim-1^∆mucin mice is impaired Breg IL-10 production. Associated with the progressive loss of IL-10 production in B cells, 10- to 12-mo-old Tim-1^∆mucin mice showed increased effector/memory Th1 responses and autoantibody production; however, these mice did not develop frank systemic autoimmune disease (14). Interestingly, Tim-1^∆mucin mice at 16–18+ mo of age developed splenomegaly and lymphadenopathy with hyperactivated IFN-γ– and IL-17–producing T cells (Fig. 1A, 1B). Additionally, three of ten 16- to 18+-month old Tim-1^∆mucin mice also showed enlarged livers that were necrotic and hemorrhagic. There were massive mononuclear cell infiltrates in multiple organs composed of macrophages/monocytes and T and B cells, particularly in livers and lungs (Fig. 1A, 1C). Histopathologic analysis demonstrated that WT liver showed few aggregates of mononuclear cells confined to the periportal region, whereas Tim-1^∆mucin liver had massive periportal and diffuse parenchymal mononuclear cell infiltrates. Similarly, in lungs of WT mice there were small aggregates of mononuclear cells confined to the periarterial and peribronchial regions and there was minimal interstitial infiltration, whereas lungs in age-matched Tim-1^∆mucin mice showed massive peribronchial and diffuse interstitial mononuclear cell infiltrates (Fig. 1D). Tim-1^∆mucin mice that develop progressive loss of IL-10 production from Bregs develop severe autoimmune disease with multiorgan/tissue inflammation, which may lead to end-organ damage, especially in liver and lungs. The disease pattern in Tim-1^∆mucin mice is very different from that in the hosts with impaired Foxp3⁺ Tregs, which develop very severe tissue inflammation and die within a few months after birth.

BCR and CD40 signaling has been shown to be required for the generation of IL-10⁺ Bregs (2) and to increase Tim-1 expression (11, 18). We have previously reported that treatment with an anti–Tim-1 mAb promotes IL-10 production in WT but not Tim-1^∆mucin B cells (14). Thus, we studied whether BCR and CD40 signaling–mediated IL-10 production was affected in B cells from Tim-1–deficient (Tim-1^−/−) (11) or Tim-1^∆mucin mice. Indeed, anti-IgM treatment in in vitro cultures increased B cell Tim-1 expression. Both anti-IgM and anti–Tim-1 treatment alone modestly but significantly enhanced IL-10 production from WT B cells (Fig. 2A). Strikingly, treatment with anti-IgM and anti–Tim-1 together strongly promoted IL-10 production in WT B cells, which is much higher than either treatment alone. However, IL-10 production induced by all these treatment conditions was significantly reduced in Tim-1^−/− and Tim-1^∆mucin B cell cultures, when compared with the WT B cells (Fig. 2A). A similar observation was obtained when anti-IgM was replaced with Abs against CD40, which is also required for Breg IL-10 production. Anti-CD40 treatment also increased Tim-1 expression on B cells, and CD40 and Tim-1 signaling together synergistically promoted IL-10 production from WT but not Tim-1^−/− or Tim-1^∆mucin B cells (Supplemental Fig. 1).

IL-21 has recently been shown to be required for IL-10 production not only in T cells but is also critical for Breg development and expansion (19). Indeed, IL-21 treatment alone or together with anti-IgM or anti-CD40 increased IL-10 production in WT B cell cultures (Fig. 2B and data not shown). IL-21 treatment also significantly increased the frequency of Tim-1⁺ B cells (Fig. 2C). Interestingly, IL-21 and anti–Tim-1 together dramatically promoted IL-10 production in WT B cell cultures with or without addition of anti-IgM or anti-CD40. In contrast, IL-21–induced IL-10 production was dramatically reduced in Tim-1^−/− and Tim-1^∆mucin B cells under all these conditions (Fig. 2B and data not shown).

Collectively, these data suggest that Tim-1 expression and signaling are essential for the maintenance and promotion of IL-10 production in Bregs. A defect in Tim-1 expression/signaling severely impairs Breg-derived IL-10 production, which cannot be rescued by BCR, CD40, or IL-21 signaling. These data also confirm that Tim-1^∆mucin is a loss of function form of Tim-1 mutant, because Tim-1^∆mucin can be normally expressed on cell surface in the mutant mice but does not act normally to maintain/induce IL-10 production from Bregs (14). Tim-1^∆mucin mice, therefore, provide a valuable tool for studying the effect of loss of Tim-1 signaling on Breg function and also provide a tool by which Bregs can be isolated from Tim-1^∆mucin+ cells.

Because Tim-1 defects in Bregs impair their IL-10 production, we next studied whether Tim-1 defects would alter proinflammatory cytokine expression in B cells. WT or Tim-1^−/− splenic B cells were stimulated with BCR ligation, and expression of Tim-1, IL-10, IL-12, IL-6, IL-23, and IL-1b mRNA was measured by real-time PCR analysis. The results showed that there was no detectable Tim-1 mRNA expression in Tim-1^−/− B cells due to Tim-1 deficiency (Fig. 3A and data not shown). Compared to WT B cells, Tim-1^−/− B cells had a >50% reduction in IL-10 mRNA expression, consistent with reduced IL-10 cytokine production (Fig. 2). Interestingly, expression of IL-12, IL-6, and IL-1b mRNA in Tim-1^−/− B cells was increased, whereas IL-23 mRNA was not detected in either WT or Tim-1^−/− B cells (Fig. 3A). These data suggest that Tim-1 deficiency in B cells alters the balance between regulatory and proinflammatory cytokines toward a proinflammatory response.

Because Tim-1^−/− B cells produce less IL-10 but more IL-6, IL-1β, and IL-12 than do WT B cells, we then analyzed whether Tim-1⁺ and Tim-1⁻ B cells differentially express these proinflammatory factors and, if so, how Tim-1 mutation in B cells affects Tim-1⁺ and Tim-1⁻ B cell responses. For this purpose, we chose an in vivo setting by cotransferring WT T cells together with WT or Tim-1^∆mucin B cells into Rag1^−/− mice that were then immunized for the induction of EAE. At the peak of disease, we examined expression of these proinflammatory cytokines in Tim-1⁺ and Tim-1⁻ B cells between WT and Tim-1^∆mucin groups. The results showed that Tim-1⁻ B cells from both WT and Tim-1^∆mucin groups had no detectable Tim-1 and little IL-10 mRNA whereas Tim-1⁺ B cells from both groups expressed Tim-1 mRNA. However, WT Tim-1⁺ B cells had much higher IL-10 mRNA than did Tim-1^∆mucin Tim-1⁺ B cells (Fig. 3B). These data are consistent with the notion that Tim-1 identifies IL-10⁺ Bregs and a Tim-1 defect impairs Breg-derived IL-10 production. Interestingly, Tim-1⁻ B cells from both groups had much higher IL-6, IL-1b, and IL-12 mRNA than did Tim-1⁺ B cells. More interestingly, both Tim-1⁺ and Tim-1⁻ B cells from Tim-1^∆mucin mice had much higher IL-6, IL-1b, and IL-12 mRNA than did Tim-1⁺ and Tim-1⁻ B cells from WT mice (Fig. 3B). Because only ∼10% of B cells are Tim-1⁺, these data indicate that these proinflammatory cytokines are largely produced by Tim-1⁻ cells, which are proinflammatory. These data further support a critical and essential role of Tim-1⁺ Bregs in limiting inflammatory responses of effector B cells; a Tim-1 defect in Bregs alters the balance between regulatory and proinflammatory activities in B cells toward a proinflammatory response.

It has been well demonstrated that IL-12 is essential for the development of IFN-γ–producing Th1 responses and that IL-6 and IL-1β are critical in the development of IL-17–producing Th17 responses (20). IL-6 also inhibits natural Treg function and iTreg generation (20). Because Tim-1^−/− B cells produced less IL-10 but more IL-12, IL-6, and IL-1β, we next studied whether Tim-1^−/− B cells would affect T cell differentiation. We cocultured WT naive T cells with either WT or Tim-1^−/− B cells in the presence of anti-CD3 under various T cell–polarizing conditions. Interestingly, compared with WT B cells, Tim-1^−/− B cells enhanced IFN-γ production under an unbiased neutral setting (Th0), which is most likely due to increased IL-12 in Tim-1^−/− B cells. The increased IFN-γ in neutral cultures with Tim-1^−/− B cells was not observed in Th1 cultures because a large amount of exogenous IL-12 was added (Fig. 3C). Tim-1^−/− B cells also promoted IL-17 production in Th17 cultures and inhibited induction of Foxp3⁺ in the presence of TGF-β1. More interestingly, Tim-1^−/− B cells also have reduced differentiation of IL-10–producing Tr1 cells. Tim-1^−/− B cells did not affect IL-4 production in Th2 cultures, however (Fig. 3C).

We also measured IL-10 production from B cells in these T/B cell cocultures. Interestingly, in all the T cell–polarizing cultures, compared with WT B cells, Tim-1^−/− B cells produced much less IL-10 (Fig. 3C), further indicating that Tim-1 is critical and essential for Breg IL-10 production.

We also compared Tim-1⁺ Bregs and Tim-1⁻ B cells isolated from WT and Tim-1^∆mucin mice for their ability to induce differentiation of Th17 cells, Foxp3⁺ iTregs, and Tr1 cells. Compared to Tim-1⁻ B cells, WT Tim-1⁺ Bregs dramatically inhibited Th17 differentiation but promoted Foxp3⁺ Treg and Tr1 generation. In contrast, these differences in T cell differentiation were largely lost when using Tim-1⁺ B cells from Tim-1^∆mucin mice (Fig. 3D).

These data suggest that B cells with defects in Tim-1 differentially regulate the generation of regulatory and proinflammatory T cells at least partly because of the difference in their regulatory and proinflammatory cytokine production.

EAE is an animal model of multiple sclerosis and is considered to be a T cell–mediated autoimmune disease in the CNS. Th1 and Th17 cells are pathogenic whereas IL-10 and Foxp3⁺ Tregs are beneficial in the disease (21). Our data thus far showed that Tim-1 is required for optimal Breg IL-10 production. Furthermore, Tim-1 defects in B cells alter the balance between regulatory and proinflammatory cytokines in B cells, under both in vitro and in vivo settings. We then asked whether Tim-1 defects in B cells would alter the incidence and severity of EAE by enhancing Th1/Th17 responses and inhibiting Foxp3⁺ Treg and Tr1 cells. Thus, WT T cells together with WT or Tim-1^−/− B cells were cotransferred into Rag1^−/− mice. After immunization with MOG_35–55/CFA to induce EAE, Rag1^−/− hosts cotransferred with WT T cells and Tim-1^−/− B cells developed more severe clinical disease than did the hosts cotransferred with WT T cells and WT B cells (Fig. 4A). The recipients that received Tim-1^−/− B cells showed increased pathogenic Th1/Th17 responses but decreased Foxp3⁺ Treg frequency and IL-10 expression in T cells obtained from the CNS (Fig. 4A). We then studied the effect of transfer of Tim-1⁺ B cells on EAE development. Our data showed that transfer of Tim-1⁺ B cells not only reduced EAE severity in WT mice (Supplemental Fig. 2) but they also decreased the severity of EAE in a Tim-1^−/− B cell–mediated transfer model (Fig. 4B). The data further emphasize that Tim-1 indeed identifies Bregs and is functionally critical for Bregs in modulating EAE severity by regulating the balance between pathogenic and regulatory T cells.

Tim-1 is a PS receptor for binding ACs (22–24). ACs have previously been shown to promote IL-10 production from Bregs (25, 26). Thus, we determined whether ACs would bind to Tim-1⁺ Bregs and promote IL-10 production. Indeed, ACs bound to Tim-1⁺ B cells at a much higher level than did Tim-1⁻ B cells from WT mice, and this binding of Tim-1⁺ B cells was lost in Tim-1^∆mucin mice (Fig. 5A). Interestingly however, unlike Tim-1⁺ epithelial cells (14, 24), Tim-1⁺ B cells did not phagocytize ACs (data not shown). Furthermore, ACs binding to Tim-1 promoted IL-10 in WT but not Tim-1^−/− B cell cultures (Fig. 5B). These data suggest that both AC binding to Tim-1⁺ Bregs and AC-mediated induction of IL-10 production in Bregs depend on Tim-1 expression on Bregs.

Administration of ACs has been reported to reduce EAE severity through a Breg-dependent manner (26). Therefore, we next asked whether administration of ACs would alter the development of EAE in hosts with Tim-1^−/− B cells. WT T cells together with WT or Tim-1^−/− B cells were cotransferred into Rag1^−/− mice. ACs were administrated 1 d before immunization with MOG_35–55/CFA for EAE induction. As shown in Fig. 4A, Rag1^−/− hosts cotransferred with WT T cells and Tim-1^−/− B cells developed more severe clinical disease than did the hosts cotransferred with WT T cells and WT B cells. AC treatment dramatically reduced EAE severity in hosts with WT B cells but not in hosts with Tim-1^−/− B cells (Fig. 5C). These data indicate that Breg-expressing Tim-1 is almost completely required for AC-mediated Breg-dependent inhibition of EAE.

In the present study, we determined the role of Tim-1 in Bregs and their effect on T cell responses and development of autoimmune diseases. Our data indicate that Tim-1 not only identifies IL-10⁺ Bregs, but also that it is required for Breg regulatory function in inhibition of the development of autoimmune diseases.

Our data in the present study further support the notion that Tim-1 identifies IL-10⁺ Bregs, as IL-10 is detected predominantly in Tim-1⁺ but not Tim-1⁻ B cells (Fig. 3B). In addition to serving as a Breg marker, Tim-1 is functionally required for Breg-derived IL-10 production, as both Tim-1^−/− and Tim-1^∆mucin B cells show impairment in IL-10 production. Further support for the role of Tim-1 in regulating Breg functions comes from the observation that treatment with anti–Tim-1 mAb promotes IL-10 only in WT but not Tim-1^−/− or Tim-1^∆mucin B cells. These data also emphasize the importance of the Tim-1 mucin domain for Tim-1–mediated signaling and function and indicate that Tim-1^∆mucin is a loss-of-function form of Tim-1 mutant, at least in terms of Breg IL-10 production. Because Tim-1^∆mucin is still expressed on cell surfaces and can be identified by anti–Tim-1 staining, Tim-1^∆mucin mice provide a valuable tool for studying the effect of loss of Tim-1 signaling in Bregs.

Many studies have shown that the BCR and CD40 signaling pathways are required for IL-10–producing Breg development and induction; IL-21 also promotes IL-10⁺ Bregs (19). Because Tim-1 identified IL-10⁺ Bregs, it was reassuring to find that Tim-1⁺ B cells increased when B cells were stimulated via BCR, CD40, and IL-21 signaling pathways. However, in all the in vitro and in vivo conditions (Figs. 2, 3B, Supplemental Fig. 1), as well as in various T/B cell cocultures (Fig. 3C), Bregs with Tim-1 defects (Tim-1^−/− or Tim-1^∆mucin) consistently showed ∼50% loss in IL-10 production. This suggests that there are also Tim-1–independent mechanisms by which Bregs produce IL-10. Nevertheless, Tim-1 ligation with anti–Tim-1 Ab synergizes with BCR, CD40, and/or IL-21 signaling pathways to promote Breg IL-10 production. All of these data strongly suggest that in addition to serving as a Breg marker, Tim-1 is required for optimal Breg-derived IL-10 production.

In addition for optimal Breg IL-10 production (and also possibly expression of other factors responsible for Breg suppressive activity), Tim-1 signaling is also required for suppressing proinflammatory cytokine production in Bregs. Tim-1⁺ Bregs mainly produce regulatory cytokines (e.g., IL-10) with low levels of proinflammatory cytokines, whereas Tim-1⁻ B cells presumably are a part of effector B cells and mainly produce proinflammatory cytokines with little IL-10. Thus, in contrast to Tim-1⁻ “effector” B cells, Tim-1⁺ Bregs regulate the balance between proinflammatory Th1/Th17 cells and regulatory Foxp3⁺ Tregs and Tr1 cells toward a regulatory response. In addition to regulating T cell responses directly, Tim-1⁺ Bregs can also regulate the balance between the proinflammatory and regulatory T cell responses indirectly by affecting function and cytokine profile of other immune populations such as Tim-1⁻ effector B cells. Therefore, Tim-1 defects in Bregs affect both Bregs and effector B cells to regulate the balance between proinflammatory and regulatory responses, pushing them toward a dominant proinflammatory response.

We have previously shown that Breg IL-10 production in young (e.g., <6 mo old) Tim-1^∆mucin mice is not as profoundly impaired as in old (10- to 12+-mo-old) Tim-1^∆mucin mice (14). Tim-1^∆mucin mice are overall normal at a young age and develop spontaneous systemic autoimmune disease only as they get old (16- to 18+-month old), which correlates with progressive loss of regulatory function (e.g., IL-10) of Bregs in the mice as they age. However, the impairment in Bregs in young (i.e., 2- to 3-mo-old) mice is severe enough to alter the phenotype and enhance the severity of EAE. Th1 and Th17 cells are pathogenic whereas IL-10 and Foxp3⁺ Tregs are beneficial in the disease (21). Because Tim-1⁺ Bregs are involved in regulating the balance between Th1/Th17 cells and Foxp3⁺ Tregs and Tr1 cells, this begins to explain why Tim-1⁺ Bregs inhibit EAE whereas B cells with Tim-1 defects promote EAE.

The progressive loss of Breg IL-10 in mice with Tim-1 defects with age is apparently not due to a decrease in the Breg population but rather is due to impaired Breg function resulting from Tim-1 defects, as the percentage of Tim-1⁺ Bregs in Tim-1^∆mucin mice is not decreased but rather increased as the mice age (Supplemental Fig. 3). However, these Bregs do not make appropriate levels of IL-10, when compared with the WT Tim-1⁺ Bregs. This further supports the conclusion that Tim-1 expression and signaling are required for maintaining Breg function and their optimal IL-10 production to promote induction of tolerance. The question that still remains is how Tim-1 signaling is triggered and maintained in Bregs for their optimal regulatory function under physiological conditions. Tim-1 has been shown to be a receptor for Tim-4 and PS exposed on ACs (22–24, 27). However, we found that treatment with Tim-4–Ig does not significantly alter IL-10 production in B cells from WT, Tim-1^−/−, or Tim-1^∆mucin B cells (data not shown), indicating that Tim-4 may not be the endogenous Tim-1 ligand for maintaining optimal function of Tim-1⁺ Bregs. ACs have been shown to play a critical role in immunological tolerance and to suppress autoimmune disease via promoting an anti-inflammatory response in terms of IL-10 production (25, 26, 28). Interestingly, we demonstrate that as a PS receptor, crosslinking of Tim-1 by PS exposed on the surface of ACs is required for Breg function. Thus, maintenance of optimal Breg function in the hosts apparently depends on the interaction of Tim-1 with ACs, which mediates persistent Tim-1 signaling to maintain and/or induce Breg function (e.g., IL-10 production). Owing to loss of AC sensing, Bregs from Tim-1 mutant mice have defects in regulatory functions, which shifts the immune balance toward a proinflammatory T cell response. This partly explains why Tim-1^∆mucin mice develop spontaneous multiorgan autoimmunity with age. The spontaneous multiorgan/tissue inflammation is not unique to Tim-1^∆mucin mice, because we have also observed that Tim-1^−/− mice at 12+ mo of age start to develop inflammation with increased infiltration of mononuclear cells in livers (Supplemental Fig. 4). Further investigation is needed to determine whether Tim-1^−/− mice will finally develop spontaneous multiorgan inflammation in multiple organs as seen in 16- to 18+-mo-old Tim-1^∆mucin mice.

In summary, we demonstrate that in addition to serving as a Breg marker, Tim-1 as a PS receptor is critical and essential for optimal Breg regulatory function in maintaining immune tolerance by sensing apoptotic cells. Thus, Tim-1 may be a valuable target for B cell–targeted therapies of autoimmune inflammatory diseases in which Bregs play a critical regulatory role.

We thank Deneen Kozoriz for cell sorting and Lila Fakharzadeh and Saranya Sridaran for technical support.

The authors have no financial conflicts of interest.