Regulatory B Cells (B10 Cells) and Regulatory T Cells Have Independent Roles in Controlling Experimental Autoimmune Encephalomyelitis Initiation and Late-Phase Immunopathogenesis

Multiple sclerosis (MS) has been classically viewed as a predominantly T cell-dependent autoimmune disease, a feature shared by rheumatoid arthritis, systemic sclerosis, and type 1 diabetes. This conclusion derives from the finding that the adoptive transfer of T cells from diseased animals can initiate disease symptoms in healthy recipients. By contrast, CD4⁺ regulatory T cells (Tregs) are critically important for limiting T cell activation during MS and other autoimmune diseases (1, 2), in part through their production of IL-10 (3). However, B cells also regulate immune responses and can contribute to disease pathogenesis (4, 5) by functioning as cellular adjuvants for CD4⁺ T cell activation (6) and through the production of cytokines that regulate T cell function and inflammation (7). Furthermore, recent phase I and II clinical trials in MS patients using depleting CD20 mAb (rituximab) suggest that pan-mature B cell depletion has clinical efficacy for the treatment of MS (8, 9), in addition to demonstrated efficacy for other autoimmune disorders (5). Despite these advances, understanding of the complex mechanisms through which B cells influence disease activity in humans and mice remains largely incomplete.

B cell-negative regulation of immune responses has been demonstrated in the mouse experimental autoimmune encephalomyelitis (EAE) model of human MS (10–12), and other mouse models of autoimmunity and inflammation (7, 12–20). A relatively rare spleen B cell subset with IL-10–dependent negative regulatory functions has recently been identified that is predominantly contained within the phenotypically unique CD1d^highCD5⁺CD19^high spleen B cell subpopulation in mice (12, 17–19). A specific subset of the CD1d^highCD5⁺ B cells can be induced to express cytoplasmic IL-10 following 5 h of in vitro stimulation with LPS, PMA, and ionomycin, with monensin included in the cultures to block IL-10 secretion (L + PIM stimulation). Given that multiple regulatory B cell subsets are likely to exist, as is now well recognized for T cells, we have specifically labeled the IL-10–competent CD1d^highCD5⁺ B cells as B10 cells because they appear to only produce IL-10 and they are responsible for most B cell IL-10 production (21). B10 progenitor (B10pro) cells have also been functionally identified in mice (5, 21). Spleen B10pro cells are also found within the CD1d^highCD5⁺ B cell subpopulation, but these cells require 48 h of in vitro stimulation with LPS or through CD40 before they acquire the ability to express cytoplasmic IL-10 after 5-h stimulation with L + PIM (21). Although B10 cells normally represent only 1–2% of spleen B cells, they dramatically inhibit the induction of Ag-specific inflammatory reactions and autoimmunity (12, 17).

Significant roles for B10 and B cells have been reciprocally identified during the initiation and progression of EAE (12). Mature B cell depletion in mice before EAE induction significantly exacerbates disease symptoms, whereas B cell depletion during EAE progression dramatically inhibits disease symptoms. B10 cell depletion from mice before disease initiation accounts for exacerbated disease, which can be ameliorated by the adoptive transfer of spleen CD1d^highCD5⁺ B cells. Similarly, IL-10 deficiency enhances the severity of EAE (22). Thereby, the balance between opposing positive and negative regulatory B cell functions shapes the normal course of EAE immunopathogenesis.

Whether and how the regulatory functions of B10 cells and Tregs overlap or influence EAE immunopathogenesis independently has remained unanswered. To address this question, the regulatory effects of adoptively transferring increasing numbers of naive or EAE-sensitized B10 cells, or IL-10–deficient CD1d^highCD5⁺ B cells into wild-type mice at various stages of disease were evaluated, in addition to depleting Tregs during both disease initiation and progression. Furthermore, we are the first to show in this study that in vivo CD22 mAb treatment preferentially depletes spleen B10 cells, which dramatically exacerbates EAE severity during the initiation phase of disease. This study thereby demonstrates that B10 cells have different regulatory functions when compared with Tregs, as they function at different time points during EAE initiation and disease progression. Moreover, B10 cells directly influenced the production of proinflammatory cytokines by CD4⁺ T cells and suppressed the Ag-presenting function of dendritic cells (DCs). Thereby, independent, but overlapping B10 cell and Treg functions shape the normal course of EAE immunopathogenesis.

Single-cell leukocyte suspensions from spleens and peripheral lymph nodes (paired axillary and inguinal) were generated by gentle dissection. Blood mononuclear cells were isolated from heparinized blood after centrifugation over a discontinuous Lympholyte-Mammal (Cedarlane Laboratories, Hornby, Ontario, Canada) gradient. CNS mononuclear cells were isolated after cardiac perfusion with PBS, as described (23). Briefly, CNS tissues were digested with collagenase D (2.5 mg/ml; Roche Diagnostics, Mannheim, Germany) and DNase I (1 mg/ml; Roche Diagnostics) at 37°C for 45 min. Mononuclear cells were isolated by passing the tissue through 70-mm cell strainers (BD Biosciences, San Diego, CA), followed by Percoll gradient (70/37%) centrifugation. Lymphocytes were collected from the 37:70% interface and washed.

Mouse CD20-specific mAb MB20-11 was used as described (24). FITC-, PE-, PE Cy5-, PE Cy7-, or allophycocyanin-conjugated CD1d (1B1), CD3 (17A2), CD4 (H129.19), CD5 (53-7.3), CD8 (53-6.7), CD11b (M1/70), CD11c (N418), CD19 (1D3), CD25 (PC61), B220 (RA3-6B2), and Thy1.1 (OX-7) mAbs were from BD Biosciences. PE-conjugated IL-10R mAb (1B1.3a) was from BioLegend (San Diego, CA). Intracellular staining used mAbs reactive with IL-10 (JES5-16E3), IFN-γ (XMG1.2), TNF-α (MP6-XT22), and FoxP3 (FJK-16s) (all from eBioscience, San Diego, CA) and Cytofix/Cytoperm kits (BD Biosciences). For T cell intracellular cytokine staining, lymphocytes were stimulated in vitro with PMA (50 ng/ml; Sigma-Aldrich, St. Louis, MO) and ionomycin (1 μg/ml; Sigma-Aldrich), in the presence of brefeldin A (BFA, 1 μl/ml; eBioscience) for 5 h before staining. Background staining was assessed using nonreactive, isotype-matched control mAbs (Caltag Laboratories, San Francisco, CA). For two- to six-color immunofluorescence analysis, single-cell suspensions (10⁶ cells) were stained at 4°C using predetermined optimal concentrations of mAb for 20 min, as described (25). Blood erythrocytes were lysed after staining using FACS lysing solution (BD Biosciences). Cells with the forward and side light scatter properties of lymphocytes were analyzed using a FACScan flow cytometer (BD Biosciences) or BD FACSCanto II (BD Biosciences).

C57BL/6 and IL-10^−/− (B6.129P2-Il10^tmlCgn/J) mice (26) were from The Jackson Laboratory (Bar Harbor, ME). CD19^−/− and human CD19 transgenic (hCD19Tg; h19-1 line) mice were backcrossed with C57BL/6 mice for 14 and 7 generations, respectively, as described (27, 28). TCR^MOG transgenic mice (Thy1.2⁺; provided by V. Kuchroo, Harvard Medical School, Boston, MA) were crossed to C57BL/6.Thy1.1 mice to generate Thy1.1-expressing T cells. All mice were bred in a specific pathogen-free barrier facility and used at 6–12 wk of age.

To deplete B10 cells in vivo, sterile CD22 (MB22-10, IgG2c) or isotype-matched mAbs (250 μg) were injected in 200 μl PBS through lateral tail veins (29). To deplete CD4⁺CD25⁺FoxP3⁺ Tregs in vivo, denileukin diftitox (5 μg in 500 μl PBS; Ligand Pharmaceuticals, San Diego, CA) or PBS was injected i.p. The Duke University Animal Care and Use Committee approved all studies.

Active EAE was induced in 6- to 8-wk-old female mice by s.c. immunization with 100 μg myelin oligodendrocyte glycoprotein (MOG)₃₅–₅₅ peptide (MEVGWYRSPFSRVVHLYRNGK; NeoMPS, San Diego, CA) emulsified in CFA containing 200 μg heat-killed Mycobacterium tuberculosis H37RA (Difco, Detroit, MI) on day 0. Additionally, mice received 200 ng pertussis toxin (List Biological Laboratories, Campbell, CA) i.p. in 0.5 ml PBS on days 0 and 2. Clinical signs of EAE were assessed daily with a 0- to 6-point scoring system, as follows: 0, normal; 1, flaccid tail; 2, impaired righting reflex and/or gait; 3, partial hind limb paralysis; 4, total hind limb paralysis; 5, hind limb paralysis with partial fore limb paralysis; 6, moribund state (11). Moribund mice were given disease severity scores of 6 and euthanized. Disease scores over the course of the 28-d experiments were totaled for each animal, and the mean for the experimental group was expressed as a cumulative EAE score.

Following an initial perfusion with PBS, animals were perfused transcardially with 4% paraformaldehyde and spinal cords were removed. Tissues were processed and blocked in paraffin wax. Two transverse sections of the thoracic and lumber spinal cord were stained with H&E and Luxol Fast Blue. The number of inflammatory foci that contained at least 20 cells were counted in each H&E-stained section in a blinded fashion. When foci coalesced, estimates were made of the number of foci. Areas of demyelination were assessed for Luxol Fast Blue-stained sections. ImageJ software (National Institutes of Health, Bethesda, MD) was used to manually trace the total cross-sectional area and the demyelinated area of each section. Total demyelination was expressed as a percentage of the total spinal cord area.

Intracellular IL-10 expression was visualized by immunofluorescence staining and analyzed by flow cytometry, as described (17). Briefly, isolated leukocytes or purified cells were resuspended (2 × 10⁶ cells/ml) in complete medium (RPMI 1640 media containing 10% FCS, 200 μg/ml penicillin, 200 U/ml streptomycin, 4 mM l-glutamine, and 5 × 10⁻⁵ M 2-ME [all from Life Technologies, Carlsbad, CA]) with LPS (10 μg/ml, Escherichia coli serotype 0111:B4; Sigma-Aldrich), PMA (50 ng/ml; Sigma-Aldrich), ionomycin (500 ng/ml; Sigma-Aldrich), and monensin (2 μM; eBioscience) for 5 h, in 48-well flat-bottom plates. In some experiments, the cells were incubated for 48 h with an agonistic anti-mouse CD40 mAb (1 μg/ml; HM40-3 mAb; BD Biosciences), as described (21). For IL-10 detection, FcRs were blocked with mouse FcR mAb (2.4G2; BD Biosciences) with dead cells detected using a LIVE/DEAD Fixable Violet Dead Cell Stain Kit (Invitrogen-Molecular Probes, Carlsbad, CA) before cell surface staining. Stained cells were fixed and permeabilized using a Cytofix/Cytoperm kit (BD Biosciences), according to the manufacturer’s instructions, and stained with PE-conjugated mouse anti–IL-10 mAb. Leukocytes from IL-10^−/− mice served as negative controls to demonstrate specificity and to establish background IL-10–staining levels.

Total RNA was extracted from cell sorter-purified B cells using Qiagen RNeasy spin columns (Qiagen, Crawley, United Kingdom). Random hexamer primers (Promega, Madison, WI) and Superscript II RNase H reverse transcriptase (Invitrogen, Carlsbad, CA) were used to generate cDNA. IL-10 transcripts were quantified by real-time PCR analysis using SYBR Green as the detection agent. The PCR was performed with the iCycler iQ system (Bio-Rad, Hercules, CA). All components of the PCR mix were purchased from Bio-Rad and used according to the manufacturer’s instructions. Cycler conditions were one amplification cycle of denaturation at 95°C for 3 min, followed by 40 cycles of 95°C for 10 s, 59°C for 1 min, and 95°C for 1 min. Specificity of the RT-PCR was controlled by the generation of melting curves. IL-10 expression threshold values were normalized to GAPDH expression using standard curves generated for each sample by a series of four consecutive 10-fold dilutions of the cDNA template. For all reactions, each condition was performed in triplicate. Data analysis was performed using iQ Cycler analysis software. The sense IL-10 primer was 5′-GGTTGCCAAGCCTTATCGGA-3′, and the antisense primer was 5′-ACCTGCTCCACTGCCTTGCT-3′. The sense GAPDH primer was 5′-TTCACCACCATGGAGAAGGC-3′, and the antisense primer was 5′-GGCATGGACTGTGGTCATGA-3′.

MACS (Miltenyi Biotec, Auburn, CA) was used to purify lymphocyte populations according to the manufacturer’s instructions. CD19 mAb-coated microbeads and CD4⁺ T cell isolation kits (Miltenyi Biotec) were used to purify B cells and CD4⁺ T cells, respectively. When necessary, the cells were enriched a second time using a fresh MACS column to obtain >95% cell purities, respectively. Spleen DCs were obtained as previously described, with minor modifications (30). Briefly, spleens were minced and incubated with collagenase D (1 mg/ml) and DNase I (0.2 mg/ml) for 30 min at 37°C. Cold EDTA was added to a final concentration of 20 mM, and cell suspensions were incubated for 5 min at room temperature before filtering through nylon mesh to remove tissue and cell aggregates. To enrich for cells with a low buoyant density, cellular suspensions were separated over a 30% BSA gradient and cells were collected from the interface. After enrichment, DCs were isolated using CD11c mAb-coated microbeads (Miltenyi Biotec).

Naive wild-type mice, wild-type mice with EAE, or IL-10^−/− mice with EAE were used as B cell donors. Splenic B cells were first enriched using CD19 mAb-coated microbeads (Miltenyi Biotec). In addition, CD1d^highCD5⁺ and CD1d^lowCD5⁻ B cells were isolated using a FACSVantage SE flow cytometer (BD Biosciences) with purities of 95–98%. After purification, 1 × 10⁶ cells were immediately transferred i.v. into recipient mice.

Splenic CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells from day 28 EAE mice were purified by cell sorting. The sorted cell populations were stimulated with agonistic CD40 mAb for 48 h, with LPS added during the final 5 h of culture. TCR^MOG CD4⁺ T cells were purified by MACS, and CFSE labeled. CFSE-labeled TCR^MOG CD4⁺ T cells (1 × 10⁶/ml) were cultured alone or with CD40/LPS-stimulated CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells (1 × 10⁶/ml) in the presence of MOG_35–55 (25 μg/ml) for 72 h.

In additional experiments, splenic CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells from day 28 EAE mice were purified by cell sorting. The sorted cell populations were stimulated with agonistic CD40 mAb for 48 h, with LPS added during the final 5 h of culture. Splenic DCs from day 10 EAE mice were purified by MACS and cultured (1 × 10⁶/ml) with CD40 mAb/LPS-stimulated CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells (1 × 10⁶/ml) in the presence of MOG_35–55 (25 μg/ml) for 72 h. DCs were purified by cell sorting after 72-h cocultures with B cells, and then cultured (5 × 10⁴/ml) with MACS-purified TCR^MOG CD4⁺ T cells (2 × 10⁵/ml) for 72 h.

All data are shown as means (±SEM). The significance of differences between sample means was determined using the Student t test.

B10 cells and the spleen CD1d^highCD5⁺ B cell subpopulation are significantly expanded in autoimmune prone mice (21). To determine whether B10 cells expand during EAE, B10 cell numbers were quantified after L + PIM stimulation and staining for cytoplasmic IL-10 expression. After MOG_35–55 immunization, spleen CD1d^highCD5⁺ B cell frequencies and numbers were significantly increased on days 7, 21, and 28 in contrast to naive mice (Fig. 1A). B10 cell frequencies and numbers were also significantly increased on days 7, 21, and 28 after MOG_35–55 immunization (Fig. 1B). Increased B cell IL-10 production paralleled B10 cell frequencies (Fig. 1C), whereas immunizations with CFA alone had no effect on B10 cell numbers (Fig. 1D). Thus, there was an initial increase in B10 cell numbers and IL-10 transcripts following MOG_35–55 immunization, which resolved, with a subsequent increase during EAE disease onset and resolution.

Because Tregs negatively regulate EAE symptoms (2), spleen CD25^highFoxP3⁺CD4⁺ Treg numbers were also quantified. Treg frequencies and numbers were only significantly higher by days 21 and 28 after MOG_35–55 immunization in contrast to naive mice (Fig. 1E). Thereby, B10 cell numbers increased during both the initiation and late phases of EAE progression, whereas Treg numbers were only increased during late-phase EAE.

To determine whether quantitative differences in spleen B10 cell numbers influenced EAE, disease initiation and progression were compared in wild-type, CD19-deficient (CD19^−/−), and hCD19Tg mice that overexpress CD19. Spleen CD1d^highCD5⁺ B cells were present at similar frequencies and numbers in wild-type and IL-10^−/− mice (Fig. 2A), as described (17). In hCD19Tg mice, CD1d^highCD5⁺ B cell frequencies and numbers were 4.4- and 1.6-fold higher than in wild-type littermates, respectively. B10 cell frequencies and numbers were also 5.6- and 1.8-fold higher in hCD19Tg mice, respectively (Fig. 2B). By contrast, CD1d^highCD5⁺ B cell frequencies (93% decrease, p < 0.01) and numbers (92% decrease, p < 0.01) were reduced in CD19^−/− mice when compared with wild-type mice, whereas B10 cell frequencies and numbers were 62 and 76% lower, respectively (p < 0.01). Thus, hCD19Tg and CD19^−/− mice provided an optimal model system for assessing the importance of B10 cell numbers during EAE.

EAE responses were assessed in MOG_35–55-immunized CD19^−/−, hCD19Tg, and wild-type mice. EAE symptoms first appeared on about day 11 after immunization of wild-type mice and peaked on about day 18, then declined gradually (Fig. 2C). By contrast, EAE severity was significantly diminished in hCD19Tg mice (cumulative EAE score, 17.5 ± 4.7) when compared with wild-type mice (38.1 ± 5.0, p < 0.05). By contrast, EAE severity was significantly enhanced in CD19^−/− mice (62.5 ± 3.3, p < 0.005), as reported (31). Thus, enhanced or reduced B10 cell numbers in mice inversely paralleled their disease symptoms.

To confirm that B10 cells regulated EAE responses in CD19^−/− mice, spleen CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells were purified from naive wild-type mice and adoptively transferred (10⁶ per recipient) into CD19^−/− mice 24 h before immunization with MOG_35–55. CD1d^highCD5⁺ B cells were used for the adoptive transfer experiments because this small subpopulation contains both the B10 and B10pro cell subsets and can be identified without in vitro stimulation (17). Transferring CD1d^highCD5⁺ B cells into CD19^−/− mice significantly reduced EAE severity (cumulative EAE score, 40.6 ± 5.0) to levels observed in wild-type mice (37.2 ± 5.6; Fig. 2D, left panel) when compared with CD19^−/− mice (62.2 ± 3.9, p < 0.01). By contrast, the adoptive transfer of CD1d^lowCD5⁻ B cells into CD19^−/− mice before EAE induction did not affect EAE severity (61.8 ± 5.1; Fig. 2D, right panel). Furthermore, the adoptive transfer of CD1d^highCD5⁺ B cells purified from IL-10^−/− mice did not affect EAE severity (Fig. 2D). Thus, B10 cells negatively regulated EAE development through the production of IL-10, with increased B10 cell numbers significantly reducing disease severity.

To determine whether endogenous B10 cells regulate EAE, methods were developed to preferentially deplete B10 cells. The in vivo treatment of mice with mAbs that bind CD22 ligand binding domains preferentially depletes splenic B cells with a marginal zone phenotype, while leaving follicular B cells largely intact (29). Because B10 and marginal zone B cells share some overlapping cell surface markers (17), the ability of CD22 mAb (MB22-10) to deplete B10 cells in vivo was quantified. Remarkably, CD22 mAb treatment reduced both the frequency (86%) and number (90%) of CD1d^highCD5⁺ spleen B cells by day 7 (Fig. 3A). CD22 mAb treatment also reduced both the frequency (48%) and number (54%) of spleen B10 cells by day 7, whereas control mAb treatment was without effect. Thereby, CD1d^highCD5⁺ B cells and B10 cells could be preferentially removed without eliminating the majority of spleen B cells.

The functional contributions of B10 cells to EAE initiation and pathogenesis were measured after depleting B10 cells using CD22 mAb. First, mice were given CD22 mAb 7 d before and 0, 7, 14, and 21 d after EAE induction to deplete B10 cells before and during EAE onset. B10 cell depletion did not accelerate disease onset, but made disease severity significantly worse (cumulative EAE score, 52.3 ± 4.4) in comparison with control mAb-treated littermates (32.0 ± 4.3; p < 0.01; Fig. 3B). In fact, 20% of the CD22 mAb-treated mice became moribund and were euthanized. Mice were also given CD22 mAb on days 7, 14, and 21 after MOG_35–55 immunization, which increased disease severity (44.8 ± 2.08) in comparison with control mAb-treated littermates (35.1 ± 2.6). B10 cell depletion in mice given CD22 mAb on days 14 and 21 after MOG_35–55 immunization did not alter disease severity (35.1 ± 2.3) in comparison with control mAb-treated littermates (30.9 ± 3.3). B10 cell depletion in mice given CD22 mAb on only day 21 did not alter disease severity. These findings argue that B10 cell regulatory function is critical during disease initiation, but not after disease onset.

Treg depletion before MOG_35–55 immunization can increase the severity of EAE, whereas Tregs accumulate in the CNS during the recovery phase of disease (32–35). Therefore, the functional contributions of Tregs to EAE initiation and pathogenesis were measured after their in vivo depletion. CD25 mAb was not used for depleting Tregs because persistent mAb can deplete or inhibit the expansion of activated T cells expressing CD25 and thereby influence EAE pathogenesis independent of Treg depletion. Rather, denileukin diftitox was used, a fusion protein of IL-2, and diptheria toxin that binds to Tregs expressing high-affinity CD25 (36–38). Denileukin diftitox has limited effects on subsequently activated effector T cells due to its short t_1/2 in vivo. Spleen CD25^highFoxP3⁺CD4⁺ Tregs were maximally reduced 3 d after denileukin diftitox injection, as previously reported (38), but spleen CD25^highFoxP3⁺CD4⁺ Tregs had returned by day 6 after the initial injection. Therefore, multiple denileukin diftitox injections were given. Treg numbers were significantly reduced (60%) in mice given denileukin diftitox 1, 4, and 7 d before Treg numbers were quantified (Fig. 3C). For Treg depletion before EAE onset, mice were given denileukin diftitox 1, 4, 7, 10, and 13 d after MOG_35–55 immunization. This Treg depletion strategy delayed EAE onset by 2 d, but the severity of disease symptoms was not altered (cumulative EAE score, 29.6 ± 5.5) in comparison with PBS-treated littermates (33.6 ± 5.7; Fig. 3D). For Treg depletion after EAE onset, mice were given denileukin diftitox 14, 17, 20, 23, and 26 d after EAE induction. Late-phase Treg depletion worsened disease (54.0 ± 4.4) significantly with 20% of the treated mice becoming moribund in comparison with PBS-treated littermates (35.6 ± 5.0, p < 0.05). Thus, B10 cell function was important for regulating EAE induction, whereas Treg function was important for regulating late-phase disease.

Because B10 cells regulated EAE induction, whereas Treg function regulated late-phase disease, the relative frequencies of B10 and Tregs within the CNS, inguinal and axillary lymph nodes draining the site of MOG immunization, and blood were compared. Within the CNS, B10 cell frequencies relative to other B cells did not change significantly during EAE development, although B10 cell numbers were significantly increased on days 21 and 28 after MOG_35–55 immunization (Fig. 4A, left panels). By contrast, CNS-infiltrating Treg frequencies relative to CD4⁺ T cells and numbers were dramatically increased on days 7, 21, and 28 after MOG_35–55 immunization relative to naive mice (Fig. 4A, right panels). Lymph node B10 cell frequencies were slightly increased by 28 d after MOG_35–55 immunization in contrast to naive mice, whereas B10 cell numbers were significantly increased on days 21 and 28 after MOG_35–55 immunization (Fig. 4B, left panels). Treg frequencies within lymph nodes were significantly increased on days 21 and 28 after MOG_35–55 immunization in contrast to naive mice, whereas Treg numbers were significantly increased during all stages of EAE (Fig. 4B, right panels). In blood, B10 cell frequencies and numbers were significantly increased on 28 d after MOG_35–55 immunization (Fig. 4C, left panels). However, circulating Treg frequencies and numbers were not changed during the course of EAE (Fig. 4C, right panel). As a consequence, Tregs far outnumbered B10 cells within CNS tissues after EAE development, although the relative numbers of B10 and Tregs within the spleen and lymph nodes were essentially unchanged during the course of EAE (Fig. 4D). By contrast, blood B10 cell numbers increased gradually relative to Tregs during EAE progression. Thus, B10 cells were far more prevalent than Tregs within the CNS before disease initiation, with Tregs dominating during late-phase EAE.

The effect of increasing B10 cell numbers in wild-type mice on EAE responses was assessed using adoptive transfer experiments. Spleen CD1d^highCD5⁺ B cells and CD1d^lowCD5⁻ B cells were purified from naive mice (Fig. 5A, left panels), and transferred into mice that were immunized with MOG_35–55 24 h after the transfer. In these adoptive transfer experiments, a total of 1.23 × 10⁵ of the transferred CD1d^highCD5⁺ B cells or 0.05 × 10⁵ of the transferred CD1d^lowCD5⁻ B cells expressed cytoplasmic IL-10 after 5 h in vitro stimulation with L + PIM in two independent experiments. In one of two experiments with identical results, the adoptive transfer of CD1d^highCD5⁺ B cells from naive mice reduced disease severity (cumulative EAE score, 30.8 ± 2.5) in comparison with mice given CD1d^lowCD5⁻ B cells (42.6 ± 4.5), but this effect was not statistically different from EAE induction or progression in PBS-treated littermates (36.6 ± 3.8; Fig. 5C, upper-left panel). Therefore, spleen CD1d^highCD5⁺ B cells and CD1d^lowCD5⁻ B cells were purified from mice with EAE on day 28 after MOG_35–55 immunization. The frequency of B10 cells within the CD1d^highCD5⁺ B cell subsets was comparable when the donors were either naive mice (Fig. 5A) or mice with EAE (day 28; Fig. 5B, middle panels). Thereby, a total of 1.27 × 10⁵ of the transferred CD1d^highCD5⁺ B cells or 0.05 × 10⁵ of the transferred CD1d^lowCD5⁻ B cells expressed cytoplasmic IL-10 after 5-h stimulation with L + PIM in two independent experiments. In one of two experiments with identical results, transferring CD1d^highCD5⁺ B cells purified from mice with EAE significantly reduced EAE severity (31.6 ± 4.1; days 13–18, p < 0.05) in recipients when compared with littermates given CD1d^lowCD5⁻ B cells (51.4 ± 4.5) or PBS-treated littermates (43.4 ± 4.7; Fig. 5C, upper-second panel). Thus, providing naive mice with B10 cells from Ag-primed mice significantly reduced EAE symptoms, whereas more modest effects were obtained when the B10 cells were isolated from naive mice.

The spleen CD1d^highCD5⁺ B cell subset contains B10pro cells that become competent to express IL-10 after in vitro CD40 engagement for 48 h to induce their maturation into IL-10–competent B10 cells (21). The CD1d^highCD5⁺ B cell subset from naive mice or mice with EAE (day 28) normally contains ≈12% IL-10⁺ B10 cells after 5 h of L + PIM stimulation (Fig. 5A, 5B). However, after 48 h of agonistic CD40 mAb stimulation, ≈40% of the purified CD1d^highCD5⁺ B cells expressed cytoplasmic IL-10, whereas <2% of purified CD1d^lowCD5⁻ B cells produced IL-10 (Fig. 5B, right panels). Thereby, the frequencies of IL-10–competent B10 cells among CD1d^highCD5⁺ B cells can be enhanced significantly by B10pro cell maturation in vitro.

For subsequent adoptive transfer experiments, CD1d^highCD5⁺ and CD1d^lowCD5⁻ B cells were purified from mice with EAE and stimulated with agonistic CD40 mAb for 48 h, with LPS added during the final 5 h of culture. In these adoptive transfer experiments, a total of 3.95 × 10⁵ of the transferred CD1d^highCD5⁺ B cells or 0.19 × 10⁵ of the transferred CD1d^lowCD5⁻ B cells expressed cytoplasmic IL-10 in vitro after 5-h stimulation with L + PIM in two independent experiments. In one of two experiments with identical results, transferring CD40/LPS-stimulated CD1d^highCD5⁺ B cells dramatically inhibited EAE progression (cumulative EAE score, 15.6 ± 6.9; days 12–22, p < 0.05) in comparison with PBS-treated littermates (42 ± 4, p < 0.05), whereas CD1d^lowCD5⁻ B cells did not (42 ± 6; Fig. 5C, lower-left panel). However, inhibition of disease was only observed when the transferred cells were given before MOG_35–55 immunization of recipients, but not on days 7 or 14 after immunization (Fig. 5C, lower-second and lower-third panels). Furthermore, CD40/LPS-stimulated CD1d^highCD5⁺ B cells purified from IL-10^−/− mice with EAE (day 28) did not affect EAE severity in wild-type recipients (Fig. 5C, lower-right panel). Thus, CD40/LPS-stimulated CD1d^highCD5⁺ B cells optimally inhibited EAE initiation in an IL-10–dependent manner.

B cell depletion before MOG_35–55 immunization exacerbates EAE and results in higher numbers of CD4⁺ T cells within the CNS, whereas B cell depletion following disease initiation reduces CD4⁺ T cell infiltration into the CNS (12). Whether the adoptive transfer of B10 cells from EAE mice inhibited T cell infiltration into the CNS of MOG_35–55-immunized mice was therefore assessed. CNS tissues were collected on day 18 from groups of mice that had been given PBS, or CD40/LPS-stimulated CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells from EAE mice (day 28), and were examined by immunofluorescence staining of lymphocytes or quantitative microscopy. Remarkably, CD40/LPS-stimulated CD1d^highCD5⁺ B cell transfers significantly reduced both Treg and CD4⁺ T cell numbers within the CNS, whereas CD1d^lowCD5⁻ B cells were without effect (Fig. 6A). Spleen Treg and CD4⁺ T cell numbers were not changed among these groups. As quantified by microscopy, CD40/LPS-stimulated CD1d^highCD5⁺ B cell transfers reduced leukocyte infiltration (84% decrease thoracic, 84% lumbar; Fig. 6B) and significantly reduced demyelination (93% decrease thoracic, 88% lumbar; Fig. 6C) within CNS tissues when compared with mice given PBS (p < 0.01; Fig. 6D). Thus, adoptively transferred CD1d^highCD5⁺ B cells had profound effects on Treg and leukocyte infiltration into the CNS.

In vitro T cell–B cell coculture systems were developed to determine how B10 cells could regulate T cell-mediated autoimmune disease in vivo. First, purified spleen CD1d^highCD5⁺ B cells or CD1d^lowCD5⁻ B cells were stimulated with agonistic CD40 mAb (48 h) and LPS (last 5 h of culture), washed extensively, and added to cultures containing MOG_35–55 and CFSE-labeled CD4⁺ T cells from TCR^MOG transgenic mice whose CD4⁺ T cells respond to MOG_35–55 peptide (39). CFSE dilution was assessed 72 h later as a marker for T cell proliferation. TCR^MOG CD4⁺ T cells cultured without B cells or without MOG_35–55 added to the cultures did not proliferate (Fig. 7A, data not shown). However, TCR^MOG CD4⁺ T cells cultured with either CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells proliferated equally well in response to MOG_35–55. Thus, B10 cells did not regulate T cell proliferation in these in vitro assays. However, when the TCR^MOG CD4⁺ T cells were cocultured with B cells in the presence of MOG_35–55, there were significant differences in T cell cytokine induction observed following PMA, ionomycin, and BFA stimulation during the final 5 h of culture. TCR^MOG CD4⁺ T cells cultured with CD1d^highCD5⁺ B cells had dramatically reduced IFN-γ and TNF-α production when compared with TCR^MOG CD4⁺ T cells that were cultured with CD1d^lowCD5⁻ B cells (Fig. 7B). These changes depended on B cell IL-10 production, because CD1d^highCD5⁺ B cells from IL-10^−/− mice did not affect IFN-γ or TNF-α production by TCR^MOG CD4⁺ T cells. IL-10 production by TCR^MOG CD4⁺ T cells was not significantly changed in these culture systems. Thus, B10 cell IL-10 can regulate Ag-specific T cell cytokine production.

IL-10R expression is heterogeneous among cells of the immune system (40). Therefore, IL-10R expression by splenic DCs, macrophages, B cells, CD4⁺ T cells, and CD8⁺ T cells from wild-type mice was assessed. IL-10R expression was highest on DCs and macrophages, with modest expression by CD4⁺ T cells, CD8⁺ T cells, and B cells (Fig. 7C). IL-10R expression was not increased on any of these leukocyte subpopulations during the course of EAE (day 7).

Because high IL-10R expression by DCs may render them more sensitive to the regulatory effects of IL-10 than CD4⁺ T cells, a role for B10 cells in regulating DC activation of CD4⁺ T cells was assessed. Briefly, purified splenic CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells from mice with EAE (day 28) were stimulated with agonistic CD40 mAb for 48 h, with LPS added during the final 5 h of culture. Purified splenic DCs from mice with EAE (day 10) were then cocultured with the CD40/LPS-stimulated CD1d^highCD5⁺ or CD1d^lowCD5⁻ B cells in the presence of MOG_35–55. After 72 h of culture, the DCs were purified by cell sorting and cultured with TCR^MOG CD4⁺ T cells for 72 h. The TCR^MOG CD4⁺ T cells were then stained for CD4 and Thy1.1 expression and analyzed for CFSE dilution. In cultures in which the TCR^MOG CD4⁺ T cells were cultured with DCs cocultured with CD1d^highCD5⁺ B cells, the intensity of CFSE staining was significantly higher and the percentage of dividing TCR^MOG CD4⁺ T cells was significantly reduced when compared with DCs cultured with CD1d^lowCD5⁻ B cells (p < 0.05; Fig. 7D). By contrast, culturing DCs with CD1d^highCD5⁺ B cells from IL-10^−/− mice did not affect their Ag-presenting ability in this T cell activation assay. Thus, IL-10–competent CD1d^highCD5⁺ B10 cells were able to regulate the Ag-presenting capability of DCs.

This study reveals that B10 cells predominantly reduce disease severity during EAE initiation through the production of IL-10, whereas Tregs reciprocally inhibit late-stage EAE immunopathogenesis (Fig. 3). Remarkably, the early expansion in B10 cell numbers and IL-10 production following MOG_35–55 immunization parallels B10 cell regulation of disease initiation (Fig. 1A–C), whereas Treg expansion during disease progression parallels their negative regulation of late-stage disease (Fig. 1E). The current study also demonstrates that numbers of endogenous or adoptively transferred B10 cells directly influence the outcome of EAE pathogenesis. Specifically, mice with decreased B10 cell numbers exhibited enhanced disease severity relative to wild-type mice, whereas mice with enhanced B10 cell numbers had reduced EAE severity (Fig. 2). Likewise, the preferential depletion of B10 cells in vivo by CD22 mAb treatment enhanced EAE pathogenesis (Fig. 3A, 3B). By contrast, the adoptive transfer of Ag-sensitized B10 cells into naive recipients before MOG_35–55 immunizations inhibited EAE pathogenesis through the production of IL-10, whereas increasing B10 cell numbers in mice exhibiting disease symptoms were without significant effect (Fig. 5). Thus, regulatory B10 cells and Tregs have independent roles in controlling EAE initiation and late-phase immunopathogenesis.

The timing of differential B10 cell and Treg expansion within tissues parallels their importance during disease initiation and late-phase EAE pathogenesis. B10 cell and CD1d^highCD5⁺ B cell numbers expanded significantly (≈70% increase) by 7 d after MOG_35–55 immunization, decreased, and expanded to even higher levels during the course of disease progression, with maximum numbers (≈130% increase) accumulating as disease resolved (Fig. 1A, 1B). IL-10 production by spleen B cells and blood B10 cell numbers followed a similar course of expansion, contraction, and expansion (Figs. 1C, 4C), whereas lymph node B10 cell numbers expanded most significantly during disease resolution (Fig. 4B). By contrast, Treg numbers did not increase following MOG_35–55 immunization, but increased gradually with EAE progression (Fig. 2C). On a relative frequency basis, B10 cells were 4 times more prevalent in the CNS than Tregs before MOG_35–55 immunization, and their numbers remained constant except during disease resolution (Fig. 4A). By contrast, Tregs were 19-fold more prevalent in the CNS on day 28 than B10 cells, whereas spleen B10 cell and Treg frequencies were equal throughout the course of EAE (Fig. 4D). The absence of Treg expansion during EAE initiation provides the likely mechanistic explanation for why B cell depletion by CD20 mAb or B10 cell depletion by CD22 mAb before MOG_35–55 immunization exacerbated EAE (Fig. 3B) (12). Reciprocally, the dramatic expansion of Tregs within the CNS during EAE progression explains why the removal of B cells or B10 cells at this time does not enhance disease pathogenesis. This was confirmed when denileukin diftitox-induced Treg depletion had no effect on disease initiation, but only exacerbated late-phase disease (Fig. 3D). Denileukin diftitox treatment also suppresses active EAE in rats when given early, but results in lethal disease if given later (41). In contrast to the current study in wild-type mice, it has been previously reported that IL-10–producing B cells contribute to EAE recovery in mice that are genetically B cell deficient (11). However, this is explained by the finding that B cell deficiency delays the emergence of Tregs and IL-10 in the CNS during EAE (42). Thereby, the relative balance between B10 cell and Treg numbers in wild-type mice during the course of EAE has dramatic effects on disease outcome.

Ag-specific B10 cell expansion is required to elicit B10 cell regulatory functions, and a diverse repertoire of B cell Ag receptors is required for B10 cell development (12, 17, 21). Thereby, the ability of B10 cells to rapidity expand during EAE initiation and to quickly inhibit disease severity suggests that a sufficient pool of Ag-specific B10 cells exists naturally that are rapidly mobilized to inhibit inflammation. This is supported by adoptive transfer experiments in which MOG_35–55-primed CD1d^highCD5⁺ B cells, but not naive CD1d^highCD5⁺ B cells, inhibited EAE development (Fig. 5). The spleen of adult C57BL/6 mice normally contains 0.5–1.0 × 10⁶ B10 cells and 3.5–4.5 × 10⁶ B10 + B10pro cells (43). Remarkably, the adoptive transfer of 1.0 × 10⁶ CD1d^highCD5⁺ spleen B cells containing ≈1.2 × 10⁵ B10 cells from naive mice reduced EAE severity, whereas 1.0 × 10⁶ CD1d^highCD5⁺ spleen B cells containing only ≈1.3 × 10⁵ B10 cells from mice with EAE significantly reduced EAE severity (Fig. 5C). The transfer of 1.0 × 10⁶ CD1d^highCD5⁺ spleen B cells from mice with EAE containing ≈3.9 × 10⁵ in vitro matured B10 + B10pro cells was able to dramatically reduce EAE severity. Thereby, the identification of Ag-specific B10 cells may significantly reduce the number of adoptively transferred B10 cells needed for inhibiting EAE in vivo. Importantly, the adoptive transfer of IL-10^−/− CD1d^highCD5⁺ B cells did not affect EAE responses under any conditions tested. Thus, in addition to their preprogrammed ability to rapidly proliferate in response to external stimuli (21) and produce IL-10 (17), the size of the endogenous B10 and B10pro cell pool is a critical factor for regulating the magnitude of acute inflammation and the induction of autoimmunity.

B10 cells regulate T cell-mediated inflammatory responses and EAE through IL-10–dependent mechanisms (12, 17). The current study demonstrated that B10 cells did not directly regulate T cell proliferation (Fig. 7A), but significantly reduced CD4⁺ T cell IFN-γ and TNF-α production during in vitro assays (Fig. 7B). By contrast, non-CD1d^highCD5⁺ B cells were not able to influence CD4⁺ T cell IFN-γ and TNF-α production. Lampropoulou et al. (44) have shown that tissue culture supernatant fluid from LPS-stimulated spleen B cells does not suppress CD4⁺ T cell proliferation in response to CD3 mAb stimulation in vitro, but is able to suppress IFN-γ secretion by T cells stimulated with CpG-stimulated DC. Tregs isolated from the CNS are also able to suppress T cell IFN-γ production in response to MOG_35–55 (45). Furthermore, B10 cells may also downregulate the ability of DCs to act as APCs and thereby indirectly modulate T cell proliferation (Fig. 7). Consistent with these findings, others have found that IL-10 suppresses the proliferation of Ag-specific CD4⁺ T cells by inhibiting the Ag-presenting capacity of monocytes and DCs (46) as well as inhibiting proinflammatory cytokine production by monocytes and macrophages (47). Lampropoulou et al. (44) have also shown that tissue culture supernatant fluid from LPS-stimulated spleen B cells is able to suppress T cell activation by CpG-stimulated DCs through IL-10–dependent pathways. IL-10 was initially associated with Th2 cells and was described to inhibit Th1 cytokine production (48–50). However, IL-10 is not only involved in the inhibition of Th1 polarization, but also prevents Th2 responses and exerts anti-inflammatory and suppressive effects on most hematopoietic cells. IL-10 produced by monocytes and cells other than T cells is required to maintain Treg-suppressive function and to maintain expression of the FoxP3 transcription factor in mice with colitis (51). However, the adoptive transfer of in vitro matured spleen B10 + B10pro cells from mice with EAE significantly reduced the number of Tregs within the CNS, in addition to reducing the number of inflammatory foci (Fig. 6). It is thereby unlikely that IL-10 produced by B10 cells contributes significantly to Treg maintenance during inflammation in vivo. However, future studies will be needed to determine whether B10 cells regulate cytokine production by the wide variety of additional T cell subsets that are known to critically influence EAE pathogenesis.

CD22 mAb treatment preferentially depleted spleen B10 cells and exacerbated EAE disease severity (Fig. 3A, 3B). Because the CD22 mAb used in this study blocks CD22 ligand binding, it is possible that CD22 engagement is particularly essential for the survival of B10 cells within lymphoid tissues (29, 52). How these mouse findings relate to the therapeutic benefits of a CD22 mAb, epratuzumab, currently in clinical trials is unknown. However, CD22 mAb treatment only exacerbated EAE severity when given to mice before disease induction and did not exacerbate symptoms in mice with ongoing disease. Thereby, CD22 mAb treatment may not worsen disease in patients diagnosed with autoimmunity. However, there may be instances in which B10 cell depletion would be advantageous, such as in the treatment of cancers, immunosuppression, or vaccination. For example, Treg depletion using denileukin diftitox has demonstrated efficacy in the treatment of hematologic malignancies and solid tumors that do not express CD25, in autoimmune disease, and in enhancing vaccine-mediated T cell immunity (37, 38, 53, 54). Although therapeutic B cell depletion has shown clinical efficacy in treating MS patients (8, 9), B cell depletion may also remove B10 cells and exacerbate MS severity, induce disease in some undiagnosed cases, or promote relapses in some rare cases (55). As examples, B cell depletion was recently suggested to exacerbate ulcerative colitis and trigger psoriasis, diseases that are thought to be predominately T cell dependent (56, 57). In addition, one case report suggests that B cell depletion might have induced relapses in a patient with an 18-y history of MS who developed antimyelin-associated glycoprotein polyneuropathy that was treated with rituximab (58). Thereby, the therapeutic benefits of B10 cell and B cell depletion in humans may also depend on the relative contributions and timing of these opposing B cell functions during immune responses.

This study demonstrates that B10 cells expand during autoantigen-specific adaptive immune responses and that adoptively transferred B10 cells are sufficient to blunt EAE induction. Thus, the development of B10 cell-based therapies may be ideal for treating some autoimmune diseases. This could include the isolation, expansion, and return of expanded B10 cells isolated from Ag-sensitized individuals. Other studies have also shown that the adoptive transfer of B cells can have a therapeutic benefit in the treatment of mice with EAE, type 1 diabetes, and collagen-induced arthritis (11, 12, 59, 60). It may also be possible to identify pathways that regulate B10 cell activation, expansion, and function, which will allow this potent B cell subset to be manipulated for therapeutic benefit. In support of this, B cell IL-10 production in MS patients is significantly lower than in healthy controls and is upregulated following therapy (61). In addition, helminth infections induce regulatory B cells in MS patients and suppress disease activity (62), which may explain environment-related suppression of MS in areas with low disease prevalence. Whereas this study further reveals the regulatory complexities of the immune system, it also opens the door for the identification of B10 cell-directed therapies that may be able to reshape the course of autoimmune disease.

We thank Dr. Karen Haas for helpful suggestions and Dr. Vijay Kuchroo for providing mice for this study.

Disclosures T.F.T. is a paid consultant for MedImmune and Angelica Therapeutics. The authors have no other financial conflicts of interest.