Cutting Edge: Integrin α₄ Is Required for Regulatory B Cell Control of Experimental Autoimmune Encephalomyelitis

Multiple sclerosis (MS) is an immune-mediated disorder of the CNS characterized by multifocal areas of leukocyte infiltration, demyelination, and axonal damage (1). Effector CD4⁺ T cells play an important role in the initiation and progression of the disease. CD4⁺ T cells provide help to B cells for their maturation into high-affinity self-reactive B cells (1). Autoreactive B cells also promote the expansion of autoreactive CD4⁺ T cells through efficient Ag presentation and release of proinflammatory cytokines (2). In addition to disease-promoting activities, emerging evidence supports the notion that B cells can have regulatory functions (3–6). Pathogenic T cells are controlled by different regulatory mechanisms, which include regulatory T cells (Tregs) and B cells (Bregs) (7, 8). Bregs produce regulatory cytokines and express inhibitory molecules that suppress pathogenic T cells and autoreactive B cells (3, 5, 6, 9, 10). Recently, IL-10– and IL-35–producing Bregs have been shown to control the development of MS and its animal model experimental autoimmune encephalomyelitis (EAE) (4–6, 10), and to control the persistence of chronic infections (5). However, little is known about the factors that are necessary for the generation and stability of these Bregs. A humanized mAb against the integrin α₄ (Itga4) subunit of the very late Ag 4 is currently used as an MS-modifying therapy (1). Although studies have focused primarily on its capacity to prevent the migration of lymphocytes into the CNS during the progression of CNS autoimmunity (1), it can also affect the homing of lymphocytes to lymphoid organs (11). Additionally, despite its efficacy and overall safety profile, it has been associated with the development of progressive multifocal leukoencephalopathy (PML), a severe disorder caused by JC virus (JCV) infection of the CNS (12). To improve our understanding of the mechanism of action of natalizumab and the possible risk factors for PML development, we have addressed the role of Itga4 neutralization on Breg functions during the course of EAE.

Mice used in this study were all on the C57BL/6 background. Itga4^fl/fl mice were crossed with CD19^Cre mice obtained from The Jackson Laboratory. Heterozygous CD19^Cre mice were used in this study. Deletion of Itga4 in Itga4^fl/fl CD19^Cre mice was efficient and equivalent in all B cells subsets in the spleen (transitional [TN], follicular [FO], and marginal zone [MZ]) as determined by flow cytometry. All animals were bred and maintained under specific pathogen-free conditions at the Benaroya Research Institute (Seattle, WA). All experiments have been approved and were performed in accordance with the guidelines of the Benaroya Research Institute Animal Care and Use Committee.

EAE was induced by s.c. immunization with an emulsion of 150 μg myelin oligodendrocyte glycoprotein (MOG)_35–55 peptide in CFA supplemented with 4 mg/ml of Mycobacterium tuberculosis extract H37Ra (Difco). Additionally, the animals received 200 ng of pertussis toxin i.p. on days 0 and 2 after immunization. In B cell transfer experiments, mice were sublethally irradiated (400 rad) and injected i.v. one day prior to immunization with 10 × 10⁶ untouched B cells (Stemcell Technologies). Animals were monitored daily for development of EAE with the scoring system as follows: 0, normal; 1, flaccid tail; 2, impaired righting reflex and/or gait; 3, partial hindlimb paralysis; 4, total hindlimb paralysis; 5, hindlimb paralysis with partial forelimb paralysis; 6, moribund state.

CNS mononuclear cells were isolated as previously described (7). Intracellular cytokine staining was performed as described previously (7). Cells were acquired on an LSR II (BD Biosciences), and data were analyzed with FlowJo software. Abs were purchased from eBioscience and BioLegend.

Mice were immunized s.c. with 150 μg MOG_35–55 emulsified in CFA without pertussis toxin. Draining lymph node (dLN) cells were collected 8 d after immunization. Cells were cultured at 5 × 10⁶ cells/ml in RPMI 1640 in the presence of increasing concentrations of MOG_35–55 for 72 h. During the last 16 h, cells were pulsed with 1 μCi [³H]thymidine. [³H]thymidine incorporation was measured using a beta counter.

A two-way ANOVA was used for statistical comparison of clinical EAE scores and cell proliferation. A one-way ANOVA was applied for statistical analysis of Fig. 2B and 2C and Fig. 3C and 3E. Unpaired t test was applied for statistical analysis of all of the other experiments. A p value <0.05 was considered significant.

To assess the importance of Itga4 blockage on Breg functions, we used MOG_35–55-induced EAE. In this model, pathogenic Th1 and Th17 cells are generated and Bregs are controlling disease severity (3, 13). We immunized mice with specific deletion of Itga4 on B cells (CD19^Cre Itga4^fl/fl mice) with MOG_35–55 and followed clinical signs of EAE. Most control mice developed a monophasic disease with ascending paralysis between days 12 and 16 after immunization but their clinical signs improved after day 21 (Fig. 1A). In contrast, CD19^Cre Itga4^fl/fl mice developed more severe EAE than did control mice between days 12 and 21 and maintained higher clinical scores than did control mice during recovery (1.94 ± 0.24 for CD19^Cre versus 2.84 ± 0.21 for CD19^Cre Itga4^fl/fl mice at day 26; Fig. 1A).

To determine the immunological mechanism leading to the increased severity of EAE observed in CD19^Cre Itga4^fl/fl mice, we investigated the composition of CNS-infiltrating immune cells. Consistent with the enhanced severity of the disease in CD19^Cre Itga4^fl/fl mice, we observed an increase in the number of CNS-infiltrating CD4⁺ T cells and CD11b⁺ monocytes in those mice compared with control mice (Fig. 1B). The increased number of CD4⁺ T cells reflected an overall increase in all subsets of effector T cells, that is, IL-17⁺, IFN-γ⁺, and IL-17⁺IFN-γ⁺ T cells (Fig. 1C, 1D). The absolute number of CNS-infiltrating Foxp3⁺ Tregs was similar between CD19^Cre Itga4^fl/fl mice and CD19^Cre mice, excluding the possibility that the enhanced disease susceptibility in CD19^Cre Itga4^fl/fl mice was due to a decrease number of Tregs infiltrating the CNS (Fig. 1D). In contrast to an increase in pathogenic CD4⁺ T cell numbers, we observed a significant decrease in the number of CNS-infiltrating B cells in CD19^Cre Itga4^fl/fl mice compared with control mice (Fig. 1B).

This suggests that the presence of B cells in the CNS might be important for the control of pathogenic T cell responses in the target tissue. In this case, pathogenic T cells should be generated at the same frequency in the periphery but their number would be increased in the CNS where Bregs are lacking. Alternatively, the absence of B cells in the CNS might reflect a defect in a subset of B cells that exhibit regulatory functions in the periphery. In this case, we should see an enhanced pathogenic T cell response generated in the periphery. To determine which one of these possibilities was at play, we immunized CD19^Cre and CD19^Cre Itga4^fl/fl mice with MOG_35–55 and evaluated MOG-specific CD4⁺ T cell proliferation and cytokine production 8 d after immunization. We found increased frequencies and absolute numbers of IL-17– and IFN-γ–producing CD4⁺ T cells in the spleen and LNs of CD19^Cre Itga4^fl/fl mice compared with control CD19^Cre mice (Fig. 1E). Furthermore, proliferation of splenic and dLN cells from CD19^Cre Itga4^fl/fl mice in response to MOG_35–55 was significantly increased compared with control CD19^Cre mice (Fig. 1F). Collectively, these data show that the enhanced frequency of pathogenic IL-17⁺ and IFN-γ⁺ T cells found in the CNS of CD19^Cre Itga4^fl/fl mice, and associated with enhanced disease development, originates in the secondary lymphoid organs. It further suggests that Itga4 could interfere with the development and/or function of B cells in secondary lymphoid organs.

Populations of Bregs that have a protective role during EAE were shown to produce either IL-10 or IL-35 (3, 5, 6). Therefore, we determined whether B cells from CD19^Cre Itga4^fl/fl mice could produce these regulatory cytokines. Upon analysis, Itga4-deficient splenic B cells produce significantly less IL-10 and Ebi3 (a subunit of IL-35) than do control B cells (Fig. 2A). In the spleen of naive wild-type (WT) mice, MZ, TN, and FO B cells differ in their ability to produce inflammatory and regulatory cytokines (14). Splenic B cell subsets were isolated from naive control CD19^Cre mice to determine which ones produced immunosuppressive cytokines. TN and MZ B cells in naive mice were the main producers of IL-10 and Ebi3 (Fig. 2B). Additionally, we compared Itga4 expression on B cell subsets in naive mice and observed that MZ B cells express the highest levels of Itga4 compared with FO and TN B cells (Fig. 2C). This differential expression of regulatory cytokine production by splenic B cells and enhanced Itga4 expression by MZ B cells prompted us to evaluate the distribution of B cell subsets in the spleen of CD19^Cre Itga4^fl/fl mice. Comparative analysis of B cell distribution in the spleen revealed that the B cell MZ compartment (B220⁺CD21^highCD23^low/−) was severely compromised in CD19^Cre Itga4^fl/fl mice compared with CD19^Cre control mice (Fig. 2D). This drastic reduction of MZ B cell number was associated with an increased frequency of TN B cells (B220⁺CD21⁻CD23⁻) (Fig. 2D). Previous analysis demonstrated that Bregs and, in particular, IL-10–producing MZ B cells express high levels of CD1d (10). Therefore, we analyzed the proportion of B220⁺CD21⁺CD23⁻CD1d^high B cells in CD19^Cre Itga4^fl/fl mice and their controls. As shown in Fig. 2E, naive CD19^Cre Itga4^fl/fl mice had fewer CD1d^high MZ B cells compared with CD19^Cre control mice. Accordingly, during EAE, splenic B cells expressing high levels of CD1d were significantly decreased in the spleen of CD19^Cre Itga4^fl/fl mice compared with CD19^Cre control mice (Fig. 2F). Furthermore, phenotypical analysis of CNS-infiltrating B cells in control CD19^Cre mice showed that these CD1d^high Bregs were 100-fold less numerous in the CNS than in the spleen of control mice with EAE (Fig. 2F).

Therefore, our data suggest that a significant portion of Breg activity might occur in the periphery and might be compromised in CD19^Cre Itga4^fl/fl mice. To test this hypothesis, we isolated B cells from the spleen of MOG_35–55-immunized CD19^Cre Itga4^fl/fl or control mice and cultured them with naive MOG-specific CD4⁺ T cells (2D2) in vitro. We found that Itga4-deficient splenic B cells, which did not contain MZ B cells, induced a stronger proliferation of MOG-specific CD4⁺ T cells than did Itga4-competent B cells, which had MZ B cells (Fig. 3A). In the LN, plasmablasts have been shown to regulate EAE severity through the control of T cell responses (4). Thus, we have determined how B cells isolated from the dLN influence CD4 T cell responses. We observed a similar increased proliferation of 2D2 CD4⁺ T cells after culture with Itga4-deficient B cells isolated from dLNs of immunized CD19^Cre Itga4^fl/fl mice compared with CD19^Cre mice (Fig. 3A). This was associated with a decrease in frequency and absolute number of CD138⁺ plasmablasts in the LN of CD19^Cre Itga4^fl/fl mice compared with control (data not shown). To address whether the modulation in B cell populations in CD19^Cre Itga4^fl/fl mice could directly control EAE, we adoptively transferred WT splenic B cells into CD19^Cre Itga4^fl/fl mice prior to EAE induction. Whereas CD19^Cre Itga4^fl/fl mice develop exacerbated EAE compared with control CD19^Cre mice, the transfer of MZ-containing WT B cells into CD19^Cre Itga4^fl/fl mice was able to reduce EAE severity to the level observed in CD19^Cre mice that did not receive B cells (Fig. 3B, 3C). To confirm that MZ B cells were responsible for this beneficial effect, we transferred total WT splenic B cells or MZ depleted B cells into CD19^Cre Itga4^fl/fl mice immunized for the development of EAE. Whereas the transfer of splenic B cells reduced EAE severity (Fig. 3D), as seen earlier (Fig. 3B, 3C), the transfer of MZ depleted B cells did not control EAE progression and severity (Fig. 3D, 3E). Therefore, these data highlight the crucial role of Itga4 expression for the regulatory function initiated by MZ B cells during CNS autoimmunity.

A dual role for B cells in EAE and MS is now emerging (2, 8). On the one hand, B cells contribute to disease pathogenesis through anti-myelin Abs production that can induce demyelination, presentation of myelin Ag, and production of proinflammatory cytokines (2). On the other hand, B cells can control disease progression and severity through the production of immunoregulatory cytokines and/or the induction of other regulatory cells (3–6, 8–10, 13, 15). Because the role of B cells in CNS autoimmunity is complex and not all B cell subsets fulfill similar functions, it becomes important to address how current MS-modifying therapies affect B cell populations.

Natalizumab, a humanized mAb against the integrin α₄ subunit of the very late Ag 4, is currently used as a disease-modifying therapy for MS (1). Most studies have focused on the importance of Itga4 for the trafficking and migration of lymphocytes to the CNS (7, 16). In this study, we have addressed the role of Itga4 neutralization on the generation and function of Bregs. The balance between pathogenic B cells and Bregs in EAE can be modulated by the immunization regimen. Indeed, immunization of C57BL/6 mice with recombinant human (rh)MOG protein induces B cell activation and production of anti-MOG Abs, which are pathogenic in association with T cell–mediated CNS inflammation (17, 18). Consistent with a pathogenic role of B cells after rhMOG immunization and the capacity of Itga4 neutralization to block the entry of lymphocytes in the CNS, a recent study showed that CD19^Cre Itga4^fl/fl mice immunized with rhMOG have a modest reduction in EAE susceptibility (16). In our study, we have immunized mice with MOG_35–55 peptide, which does not elicit a strong humoral response and does not generate pathogenic B cells because EAE development is not abrogated in B cell–deficient animals (3, 13). In contrast to CD19^Cre Itga4^fl/fl mice immunized with rhMOG (16), CD19^Cre Itga4^fl/fl mice immunized with MOG_35–55 peptide develop more severe and sustained EAE than do control mice. Similarly to the previous study, we observed a requirement of Itga4 for the entry of B cells in the CNS. However, our data suggest that Itga4 is important to maintain Breg activity and that a significant portion of this Breg activity is occurring in the periphery. Indeed, enhanced CD4⁺ T cell proliferation and pathogenic cytokine production were observed in the peripheral lymphoid organs of CD19^Cre Itga4^fl/fl mice immunized with MOG_35-55 and were associated with a decrease in MZ and MZ CD1d^high B cells in the spleen (Fig. 2A, 2B, 2E, 2F). During EAE, the number of CNS-infiltrating CD1d^high B cells was also reduced in the CNS of CD19^Cre Itga4^fl/fl mice compared with CD19^Cre control mice (Fig. 2F), consistent with a general decrease in CNS-infiltrating B cells (Fig. 1B). However, the CD1d^high B cell population was 100-fold less abundant in the CNS than in the spleen of CD19^Cre mice (Fig. 2F). Therefore, although we cannot completely rule out Breg activity in the CNS, we propose that an important part of the Breg immunoregulatory effects takes place in peripheral lymphoid organs. This would explain why some studies have failed to detect numerous Bregs in the CNS (2). Alternatively, we have established a clear requirement of Itga4 for MZ (especially CD1d^high) B cell and IL-10– and IL-35–producing B cell presence in the spleen because these populations are dramatically compromised in the spleen of CD19^Cre Itga4^fl/fl mice. Furthermore, we have established that the regulatory cytokines IL-10 and IL-35 are preferentially expressed in the MZ B cell subset. Although MZ B cells are mostly spleen resident, they have been shown to recirculate (19). Upon immunization, a significant proportion of Breg express CD138 in association or not with BLIMP1 (4, 5) and are present in the spleen and LN where they can limit T cell responses. In support to this hypothesis, the frequencies of CD138⁺ (BLIMP1⁺ and BLIMP1⁻) B cells present in the LN and spleen of CD19^Cre Itga4^fl/fl mice after immunization with MOG_35–55 were significantly decreased compared with CD19^Cre Het mice (data not shown). Therefore, the lack of B cell subsets with regulatory potential in the lymphoid organs of CD19^Cre Itga4^fl/fl mice together with the increased proliferative response of LN and spleen cells to MOG_35–55 peptide indicate that both lymphoid organs can be the site of active control of these responses by Bregs. They further indicate that the regulation of the immune response requires properly organized lymphoid tissues and/or the maturation of MZ B cells in the spleen.

Despite its efficacy and overall good safety profile, natalizumab therapy has been associated with the development of PML, a severe disorder caused by JC virus infection of the CNS (20). MZ B cells are often considered innate immune cells capable of mounting a quick and efficient T cell–independent response against blood-borne pathogens (19). Although our study does not address this point directly, it is possible that the retention of MZ B cells in the spleen through the expression of Itga4 might be required for efficient JC virus containment and latency. This will be an important area of investigation and may shed light on the complex dissemination of this virus in the body of immunocompromised individuals.

Importantly, our study, which links Itga4 to MZ and Bregs, should promote further investigation on the role of MZ B cells in the ontogeny of Bregs during CNS autoimmunity and infection to further improve current MS-modifying therapies.

We Thank Dr. T. Papayannopoulou and Dr. Oukka for providing the CD19^Cre Itga4^fl/fl mice.

The authors have no financial conflicts of interest.