|Home | About | Journals | Submit | Contact Us | Français|
Multiple sclerosis (MS) is a clinically and pathologically heterogeneous inflammatory/demyelinating disease of the CNS. In the MS variant Devic disease, lesions are predominantly found in the optic nerves and spinal cord but not the brain. The immunological bases of the different forms of MS are unknown. We previously generated myelin oligodendrocyte glycoprotein–specific (MOG-specific) TCR transgenic mice (TCRMOG mice; also referred to as 2D2 mice) and reported that a large proportion of these mice develop spontaneous isolated optic neuritis. We have now crossed the TCRMOG mice with MOG-specific Ig heavy-chain knock-in mice (IgHMOG mice; also referred to as Th mice), in which one-third of the B cells are specific for MOG. In these mice, MOG-specific B cells are very efficient in presenting MOG to the transgenic T cells and undergo class switching to IgG1 in the presence of the transgenic T cells. Sixty percent of TCRMOG×IgHMOG mice spontaneously developed a severe form of experimental autoimmune encephalomyelitis (EAE). Histological examination of the CNS revealed a selective distribution of meningeal and parenchymal inflammatory lesions in the spinal cord and optic nerves. Thus, CNS antigen–specific T and B cells cooperate to induce a distinct clinicopathologic EAE pattern that closely replicates human Devic disease.
MS is a clinically and pathologically heterogeneous disease most often characterized by inflammatory and demyelinating lesions throughout the CNS. There are several subtypes of the disease that differ in their CNS lesion distribution (1). Many MS patients first develop isolated optic neuritis, in which lesions are predominantly present in the optic nerves (2, 3). In Asian populations, Devic disease, or neuromyelitis optica (NMO), is the predominant subtype, representing one-third of MS cases (4). Patients with NMO often have a poor prognosis compared with patients with typical MS and develop severe and early relapses (5, 6). Devic disease is characterized by a specific lesion distribution in which inflammatory foci are restricted to optic nerves and the spinal cord with a remarkable absence of lesions in the brain (7). Antibodies (e.g., IgG and IgM) and products of the complement activation cascade have been found in and around CNS lesions, but a pathogenic role for autoantibodies in this disease subtype remains to be formally demonstrated (8). Interestingly, plasmapheresis has been reported to improve the neurologic outcome in patients with severe NMO (9), thereby supporting a role for humoral factors (possibly antibodies) in its pathogenesis.
The cellular and molecular bases for the different subtypes of MS are not well understood, but many features of the human disease have been recapitulated in the EAE animal model. EAE can be induced by immunization of susceptible animals with myelin antigens such as myelin basic protein (MBP), proteolipid protein (PLP), and myelin oligodendrocyte glycoprotein (MOG) (10–12); it is characterized by lymphocytic and mononuclear cell inflammatory infiltrates and demyelination in the brains and spinal cords of affected animals. Clinical manifestations of the disease depend on the antigen and the genetic background of the animal used. Immunization of SJL mice with the PLP aa 139–151 peptide, for example, induces episodes of relapsing-remitting disease in which the mice have a paralytic attack followed by periods of remission (11). On the other hand, immunization of C57BL/6 mice with MOG aa 33–55 produce a chronic form of EAE in which the animals stay paralyzed after their initial attack (12). Therefore each EAE model recapitulates some aspects of the human disease and provides valuable insight into the heterogeneous manifestations of MS.
EAE can be induced by myelin antigen–specific Th1 cells (13, 14); more recently it has been suggested that Th-17 cells are major inducers of EAE (15, 16). By contrast, initial studies with immunization of B cell–deficient mice with the encephalitogenic peptide MOG aa 35–55 suggested that B cells may not be essential for the development of certain forms of EAE (17, 18). However, studies of immunization of B cell–deficient mice with recombinant whole MOG indicated a pathogenic role for B cells in the initiation of EAE (19–21), and other reports suggested that myelin antigen–specific B cells may act as regulatory cells and control the development of EAE (22). Therefore, the role of myelin antigen–specific B cells and antibodies in the pathogenesis of EAE as well as of MS are not well understood.
The result of interactions between myelin antigen–specific B cells and myelin antigen–specific T cells on the development and the phenotype of EAE have not previously been investigated. Here we describe a mouse EAE model in which both T and B cells are specific for the same autoantigen MOG. The presence of MOG-specific B and T cells in the peripheral repertoire resulted in the development of severe and spontaneous EAE in the majority of the mice. Histopathological analysis of the CNS of the affected mice showed a preponderance of inflammatory lesions in the leptomeninges and parenchyma of the spinal cords and optic nerves but not in the brains of affected mice. Our results suggest that an active cooperation between MOG-specific T and B cells results in a spontaneous and severe form of EAE with a lesion distribution pattern typical of human Devic disease.
We previously generated MOG-specific TCR transgenic mice on a C57BL/6 background (TCRMOG mice; also referred to as 2D2 mice) in which the transgenic TCR-bearing cells are efficiently selected in the thymus and are seeded to the peripheral immune compartment (23). Although over 95% of the peripheral T cells express transgenic TCR, only a small percentage (4%) of TCRMOG mice develop spontaneous EAE. Nevertheless, a significant number of the transgenic mice develop spontaneous optic neuritis (47%). Immunization of TCRMOG mice with the MOG peptide in complete Freund’s adjuvant plus pertussis toxin resulted in the development of full-blown EAE, suggesting that EAE can be induced in the transgenic mice if there is appropriate activation of transgenic TCRMOG T cells. Since antigen-specific B cells can be efficient APCs, we hypothesized that MOG-specific B cells could activate TCRMOG T cells and induce spontaneous EAE in the TCRMOG mice. Therefore, we crossed the TCRMOG mice with MOG-specific Ig heavy-chain knock-in mice on a C57BL/6 background (IgHMOG mice; also referred to as Th mice), in which 20% of peripheral B cells are specific for the autoantigen MOG (24), thus providing B and T cells specific for the same MOG antigen in the peripheral repertoire of the mice.
To determine whether the presence of MOG-specific B cells could change the T cell response of TCRMOG T cells in TCRMOG×IgHMOG mice, we first analyzed the proliferation of splenocytes from TCRMOG×IgHMOG and control TCRMOG mice crossed with nitrophenyl-specific (NP-specific) Ig heavy-chain knock-in mice on a C57BL/6 background (TCRMOG×IgHNP mice; IgHNP also referred to as B1.8 mice) in response to MOG aa 35–55 peptide or whole recombinant MOG aa 1–125 (rMOG) protein. Splenocytes from TCRMOG×IgHMOG mice, carrying MOG-specific T and B cells, and TCRMOG×IgHNP mice, carrying MOG-specific T cells and NP-specific B cells, proliferated equally well in response to MOG aa 35–55 peptide (Figure (Figure1A).1A). As a control, splenocytes from IgHMOG mice, IgHNP mice, and OVA-specific TCR transgenic mice on a C57BL/6 background (TCROVA mice; also referred to as OT-II mice) crossed with IgHMOG mice (TCROVA×IgHMOG mice), all lacking the MOG-specific T cells, did not respond to MOG aa 35–55 peptide (Figure (Figure1A).1A). Similarly, splenocytes from IgHNP, IgHMOG, or TCROVA×IgHMOG mice did not proliferate in response to whole rMOG (Figure (Figure1A).1A). However, splenocytes from TCRMOG×IgHMOG mice, which express MOG-specific T and B cells, proliferated more strongly and at lower concentrations to rMOG than the control TCRMOG×IgHNP mice (Figure (Figure1A). 1A).
The enhanced response of TCRMOG×IgHMOG splenocytes to rMOG could be attributed to the proliferation of either T or B cells or both. To determine which cell type was responsible for the enhanced proliferation, we first tested the response of T cells in the presence of B cells of different specificities. We initially separated T cells from the spleens of TCRMOG mice and stimulated them with purified B cells from either IgHNP or IgHMOG mice together with MOG aa 35–55 peptide or rMOG protein as a source of specific antigen. Whereas MOG aa 35–55 peptide induced a strong proliferation of MOG-specific T cells independently of the presence of specific B cell populations, whole rMOG protein led only to dose-dependent proliferation in the presence of MOG-specific B lymphocytes (Figure (Figure1B).1B). To confirm that the observed response corresponded to T cell proliferation, we stimulated CFSE-labeled splenocytes of TCRMOG mice in the presence of MOG-specific B cells (IgHMOG) or NP-specific B cells (IgHNP) and different concentrations of MOG aa 35–55 peptide or rMOG and measured T cell proliferation by CFSE dilution on CD4+ T cells. Using this approach we observed similar numbers of dividing TCRMOG T cells in response to MOG aa 35–55 peptide independent of the antigen specificity of the B cells used as APCs (Figure (Figure1C).1C). In contrast, more TCRMOG T cells divided in response to rMOG in the presence of IgHMOG MOG-specific B cells than in the presence of IgHNP NP-specific B cells, confirming that provision of MOG-specific B cells to T cells of the same specificity enhanced their proliferative response (Figure (Figure1C1C and Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI28334DS1).
Next we analyzed T cell cytokines secreted under these culture conditions. Stimulation of TCRMOG T cells with rMOG and B cells led to a dose-dependent secretion of IFN-γ. IFN-γ production in response to rMOG was greater in the presence of MOG-specific IgHMOG B cells than in the presence of NP-specific B cells (Figure (Figure1D).1D). Intracellular anti–IFN-γ staining of CD4+ T cells confirmed that MOG-specific CD4+ T cells were responsible for the increased IFN-γ production observed (data not shown). In contrast, IFN-γ levels were comparable among T cells stimulated with MOG aa 35–55 peptide, IgHMOG B cells, and IgHNP B cells (Figure (Figure1D).1D). These data suggest that the MOG protein is selectively processed (25) and presented by MOG-specific B cells, resulting in more efficient activation of MOG-specific T cells when both cell types are present in the peripheral repertoire. The Th2-associated cytokines IL-4 and IL-10 and the pathogenic cytokine IL-17, which is the signature of Th-17 cells, could not be detected in the supernatants (data not shown).
Next we analyzed the role of B cells in the enhanced response of TCRMOG×IgHMOG splenocytes to rMOG. B cells were purified from either IgHMOG (MOG-specific) mice or IgHNP (NP-specific) mice, labeled with CFSE, and cultured in the presence of purified T cells from either TCRMOG (MOG aa 35–55–specific) mice or TCROVA (OVA aa 323–339–specific) mice. B cell proliferation was evaluated by CFSE dilution on B220+ B cells, and different cell surface activation molecules were assessed by flow cytometry analysis. Provision of MOG protein to MOG-specific B cells was not sufficient to trigger B cell division (Figure (Figure2A).2A). Coculture with activated TCRMOG T cell lines and MOG led to a dose-dependent proliferation of MOG-specific IgHMOG B lymphocytes (Figure (Figure2A)2A) and upregulation of activation markers such as CD69, CD80, and CD86 (Supplemental Figure 2 and data not shown). However, MOG-specific B cells also proliferated upon coculture with activated OVA-specific T cell lines and rMOG, as did NP-specific B cells upon coculture with activated MOG-specific T cells (Figure (Figure2B).2B). These results show that, in vitro, the presence of activated T cells alone rather than antigen-specific T cells is sufficient for the induction of B lymphocyte proliferation.
We previously reported that very few (4%) TCRMOG mice develop spontaneous EAE, but a significant portion develops isolated optic neuritis (23). We also found that the development of full-blown EAE in TCRMOG mice was in part dependent on the immunization regimen, suggesting that isolated optic neuritis may progress to EAE if there is appropriate activation of MOG-specific T cells. Based on our in vitro data indicating that antigen-specific B cells efficiently activated MOG-specific T cells, we followed the development of spontaneous EAE in TCRMOG×IgHMOG and control mice.
Consistent with previous reports, no spontaneous EAE was observed in IgHMOG mice (24), and TCRMOG mice developed a very low incidence of spontaneous EAE (6%; Table Table1).1). The presence of transgenic T cells of a different specificity (OVA specific) also did not result in the development of spontaneous EAE in TCROVA×IgHMOG mice (Table (Table1).1). In sharp contrast, 59% of TCRMOG×IgHMOG mice carrying MOG-specific T and B cells developed severe EAE (Table (Table1).1). The disease had an abrupt onset with rapid progression to a clinical score of 2.5–3 over 1–2 days in most mice with disease (Supplemental Figure 3). We did not observe any gender bias for the development of spontaneous disease in TCRMOG×IgHMOG mice. Incidence as well as mean clinical scores among TCRMOG×IgHMOG mice were higher than in TCRMOG mice with spontaneous EAE. Spontaneous disease was also accelerated in that it developed earlier in TCRMOG×IgHMOG mice (44.1 days after birth) than in TCRMOG mice (73.5 days after birth). Mortality among TCRMOG×IgHMOG mice was 10% (3 of 29), compared with 0% (0 of 5) in TCRMOG mice. The severity of the disease was also apparent histologically, as TCRMOG×IgHMOG mice had a mean of 188.4 CNS inflammatory foci compared with 16.6 in the TCRMOG mice (Table (Table1).1). No histologic disease was observed in TCRMOG×IgHMOG mice without clinical disease, TCRMOG×IgHNP mice, or IgHMOG mice (Table (Table1).1). These results suggested that the presence of MOG-specific B cells is crucial for increased incidence and severity of spontaneous EAE in TCRMOG mice.
We determined the anatomic localization of inflammatory lesions in the different areas of the CNS of TCRMOG and TCRMOG×IgHMOG with EAE. In TCRMOG mice with EAE, most of the lesions (40%) were localized in the cerebral hemispheres (Table (Table2).2). The remaining foci were equally distributed among the midbrain-brainstem, the cerebellum, and the spinal cord. These mice also had optic neuritis. In contrast, most of the lesions of TCRMOG×IgHMOG mice were localized in their spinal cords and optic nerves, with very few in the brain (less than 3% in cerebral hemispheres, midbrain-brainstem, or cerebellum; Table Table22 and Figure Figure3A).3A). This lesion distribution pattern is similar to that of Devic disease. This selective distribution of inflammatory foci in the spinal cords and optic nerves was observed in 100% of the affected TCRMOG×IgHMOG mice analyzed histologically (Table (Table1),1), but not in TCRMOG mice with EAE. The inflammatory lesions were composed of lymphocytes and monocytes and were present in both leptomeninges and parenchyma, particularly the subpial regions where there was marked associated demyelination and edema (Figure (Figure3A).3A). Peripheral nerve roots were intact. Phenotypic analysis on cryosections demonstrated infiltration of T cells (CD4+ and TCR Vβ11+) and large numbers of B cells (B220+; Figure Figure3B).3B). The accumulations of lymphocytes were particularly marked in the subarachnoid spaces around the spinal cord of TCRMOG×IgHMOG mice, and the aggregates focally formed follicle-like structures that were made evident using reticulin stains (Figure (Figure3C).3C). Similar patterns of germinal center–like structures have previously been described in the leptomeninges of patients with MS (26, 27). We did not identify substantial numbers of eosinophils or neutrophils in these lesions. Collectively, these data indicate that TCRMOG×IgHMOG mice have a selective Devic-like distribution of inflammatory lesions, i.e., in the spinal cord and the optic nerves, which is characterized by a close and well-organized follicule-like architecture that may allow close interaction between antigen-specific T and B cell populations in the target tissue.
To obtain insight into the mechanisms responsible for the development of Devic-like disease in TCRMOG×IgHMOG mice, we analyzed the phenotype of both T and B cells ex vivo. Between 6 and 8 weeks of age, CD4+ T cells from TCRMOG mice were mostly naive, and only 1.8% of them were activated/memory cells in the spleen (Figure (Figure4A).4A). Healthy TCRMOG×IgHMOG mice had similar numbers of activated cells (2.4%). In contrast, an increased proportion of CD4+ T cells (4.7% and 8.2% in spleens and lymph nodes, respectively) from diseased TCRMOG×IgHMOG mice were activated and expressed CD69 (Figure (Figure4A).4A). Analysis of the number of regulatory cells, as determined by Foxp3 intracellular staining in CD4+ T cells, showed a small but comparable number of CD4+Foxp3+ T cells (2%) in the spleens and lymph nodes of healthy and diseased TCRMOG×IgHMOG mice (data not shown).
Next we looked at the expression of different markers on the surface of B cells from the various mouse strains. We did not find any difference in the expression of MHC class II molecules, B7.2, or IgMa on the surface of B cells from IgHMOG or healthy or diseased TCRMOG×IgHMOG mice (Figure (Figure4B4B and data not shown). However, we observed an increase in the percentage of B220+B7.1+ cells in the spleens of TCRMOG×IgHMOG mice with disease compared with healthy TCRMOG×IgHMOG mice. The upregulation of the B7.1 molecule on the surface of B cells in TCRMOG×IgHMOG mice with Devic-like disease suggests that these B cells have been activated. To determine whether MOG-specific B cells differentiated into effector B cells, we measured levels of MOG-specific antibodies in serum samples of IgHMOG, TCRMOG, TCRMOG×IgHMOG, TCROVA×IgHMOG, and TCRMOG×IgHNP mice. No MOG-specific antibodies were detected in the absence of IgHMOG MOG-specific B cells, i.e., in TCRMOG mice and in TCRMOG×IgHNP mice. Without any immunization, IgHMOG mice and TCROVA×IgHMOG mice produced large amounts of MOG-specific IgM, IgG2a, and IgG2b antibodies but low levels of IgG1 (Figure (Figure4C).4C). In contrast, serum samples from TCRMOG×IgHMOG mice, in which MOG-specific T cells and B cells are present in the peripheral repertoire, contained lower amounts of MOG-specific IgM (P = 0.04) but higher amounts of MOG-specific IgG1 antibodies (P = 0.03) than the IgHMOG and TCROVA×IgHMOG mice. MOG-specific antibody isotype switching to IgG1 was consistently found in all TCRMOG×IgHMOG mice analyzed, whether they had clinical EAE or were asymptomatic (Figure (Figure4C).4C). Isotype switching was not dependent on age, since serum samples obtained from 4-week-old and adult TCRMOG×IgHMOG mice showed high levels of anti-MOG IgG1 antibodies.
To further characterize the MOG-specific T cell responses involved in the development of Devic-like disease, we purified mononuclear cells from WT, IgHMOG, and healthy and diseased TCRMOG×IgHMOG mice. Consistent with the absence of disease in the WT, IgHMOG, and healthy TCRMOG×IgHMOG mice, there were small numbers of mononuclear cells present in the CNS of these animals (data not shown). In these 3 strains of mice, CD4+ T cells represented 3% of the few infiltrating monocytes. However, in TCRMOG×IgHMOG mice with Devic-like disease, there was an increase in the number of infiltrating monocytes (data not shown) and a 5-fold increase in the percentage of CD4+ T cells in the CNS tissues (Figure (Figure5A).5A). Since both Th1 and Th-17 cells have been implicated in the development of EAE (15, 16, 28, 29), we determined whether T cells from TCRMOG×IgHMOG mice with EAE could secrete these cytokines. Although IL-17 production was not detected in our in vitro cultures (data not shown), intracellular cytokine staining of CD4+ T cells obtained from the CNS of TCRMOG×IgHMOG mice showed that CD4+ T cells involved in the development of Devic-like disease secreted IFN-γ and IL-17 (Figure (Figure5B).5B). Since only 4% of TCRMOG mice develop spontaneous EAE, it has not been possible to obtain CNS samples from these animals to perform FACS analysis together with those of TCRMOG×IgHMOG mice with disease. However, we have prepared RNA from the spinal cords of TCRMOG×IgHMOG mice with Devic-like disease and TCRMOG mice with spontaneous EAE and analyzed the expression of IL-17 mRNA by real-time PCR. This analysis showed that there was an increase in expression of both IL-17 and IFN-γ mRNA in the spinal cords of diseased TCRMOG×IgHMOG mice compared with TCRMOG mice with EAE (Figure (Figure5C).5C). Although we could not determine which cell type was responsible for the increased IL-17 and IFN-γ production observed in the spinal cords of TCRMOG×IgHMOG mice, together with the intracellular cytokine staining, these data suggest that more Th1 and Th-17 cells are infiltrating the spinal cords of TCRMOG×IgHMOG mice than do so in TCRMOG mice.
While the enhanced and selective activation of TCRMOG T cells in the presence of MOG-specific B cells can explain the high incidence of spontaneous disease observed in TCRMOG×IgHMOG mice compared with TCRMOG mice, it is not clear why TCRMOG×IgHMOG mice have Devic-like disease and not typical EAE. To begin to address this issue, we compared the expression of the autoantigen MOG in different CNS regions of C57BL/6 mice. Consistent with our previous observation (23), we found that MOG mRNA was more abundant in the optic nerve than the spinal cord (Figure (Figure6).6). We also found that MOG mRNA was more abundant in the spinal cord than in the brain. This selective distribution of MOG mRNA might contribute at least in part to the preferential localization of lesions in the spinal cords and optic nerves of TCRMOG×IgHMOG mice.
Here we report the generation of what we believe to be a new mouse model in which both T and B cells are specific for the same myelin antigen, MOG. We demonstrated that there was active cooperation between T and B cells in these animals: the presence of MOG-specific T cells led to massive production of MOG-specific IgG1 antibody, and MOG-specific B cells also enhanced MOG-specific T cell proliferation and activation. Strikingly, this cooperation resulted in the development of a spontaneous and severe form of EAE characterized by the specific presence of inflammatory foci in the spinal cords and optic nerves of the animals. This lesion distribution pattern is very similar to that of human NMO.
B lymphocytes have been implicated in several autoimmune diseases, mostly because of their capacity to produce specific autoantibodies. B cells and plasma cells are often present in MS lesions (30, 31), and oligoclonal Ig bands have been observed in the cerebrospinal fluid of many but not all MS patients (32, 33). The presence of Ig and products of complement activation cascade deposited around the vessels suggests a pathogenic role for autoantibodies in NMO (8). In addition, a serum IgG autoantibody (NMO-IgG) serves as a specific marker for NMO (34). Recent work has shown that the NMO-IgG is not specific for myelin antigens but selectively binds to the aquaporin-4 water channel, a component of the dystroglycan protein complex located in astrocytic foot processes at the blood-brain barrier (35).
The high expression of class II MHC and costimulatory molecules by activated B cells suggest that they could also play a role in driving the T cell responses as a result of antigen presentation to large numbers of antigen-specific T cells (36, 37). In vivo, depletion of B cells from normal mice can reduce the magnitude of the immune response, suggesting that B cells contribute to the T cell response (38–41), although primary T cell responses can also develop in B cell–deficient mice (42). Therefore, a number of reports suggested that antigen-specific B cells induce naive CD4+ T cell proliferation in vivo (43–45). The importance of B cells as APCs in some autoimmune diseases such as proteoglycan-induced autoimmune arthritis has recently been demonstrated (46). In contrast with this arthritis model, EAE is generally considered to be mainly T cell driven (28). Nevertheless, it is important to note that B lymphocytes are required for the induction of EAE when MOG protein, but not MOG peptide, is used for the induction of EAE (19). Using a double antigen-specific transgenic model, we demonstrate here that, in contrast to wild-type splenocytes containing the classical APCs, MOG-specific B cells were able to stimulate the proliferation of MOG-specific T cells in the presence of a limited amount of rMOG. Since IgHMOG B cells only recognize conformationally dependent epitopes of MOG but not MOG aa 35–55 (47), the enhanced T cell response in the presence of IgHMOG B cells and MOG most likely resulted from the capture of MOG through B cell surface Ig that caused enhanced presentation of the antigen rather than direct presentation of MOG aa 35–55 epitope.
Ex vivo analysis of T cells from TCRMOG×IgHMOG mice revealed that unless mice exhibited clinical disease, there was no marked increase in the number of activated T cells in the periphery (Figure (Figure4).4). Development of Devic-like disease correlated with the infiltration of both T and B cells into the spinal cord (Figure (Figure3B3B and Figure Figure5A).5A). The disease was also associated with the presence of lymphoid follicle–like structures in the spinal cords of TCRMOG×IgHMOG mice with Devic-like disease, suggesting an active cooperation between T and B cells in situ (Figure (Figure3C).3C). Similar to actively induced EAE, we observed that both IL-17– and IFN-γ–producing T cells infiltrated the CNS of TCRMOG×IgHMOG mice with Devic-like disease. Because of the low frequency of spontaneous disease observed in the TCRMOG mice, it was not possible to perform parallel intracellular cytokine staining of CD4+ T cells infiltrating the brains of TCRMOG mice with spontaneous EAE. However, IL-17 mRNA expression was greater in the spinal cords of TCRMOG×IgHMOG mice with Devic-like disease than in TCRMOG mice with EAE. Therefore the increase in IL-17 expression in the CNS of TCRMOG×IgHMOG mice with Devic-like disease might reflect an increase in the number of pathogenic Th-17 cells. Alternatively, we cannot exclude the possibility that other cell types might be responsible for the enhanced IL-17 production. The role of the CNS cytokine milieu and MOG-specific B cells in the differentiation of Th-17 cells is currently under investigation.
Th cells are important for the differentiation of B cells into Ig-secreting plasma cells or long-lived memory B cells. In addition, interaction between antigen-specific T and B cells is required for B cell maturation and Ig isotype class switching. Consistent with the role played by Th cells in the maturation of B cells, we also observed that B cells from TCRMOG×IgHMOG mice underwent class switching from IgM to IgG1 in vivo. Ig switch is dependent on the subset of T cells providing help to B cells, and IgG1 is regarded as a Th2-dependent isotype (48). In TCRMOG×IgHMOG mice, however, the T cell response seems to be biased toward Th1 cells, since we only detected IFN-γ and not Th2 cytokines in cultures of TCRMOG or TCRMOG×IgHMOG mouse spleen cells. In addition, we observed an increase in the number of B7.1-expressing B cells in TCRMOG×IgHMOG mice with Devic-like disease. B7.1 is a costimulatory molecule associated with enhanced Th1 responses. Therefore, the requirement for Th2 responses in IgG1 class switching might not be absolute in this model. In support of this hypothesis are previously published results showing that mice deficient in the Th2-specific cytokine IL-4 have diminished but still robust production of IgG1 (48). Alternatively, other factors present in the microenvironment might contribute to the IgG1 class switching observed in TCRMOG×IgHMOG mice. Taken together, the TCRMOG×IgHMOG model provides an interesting and, to the best of our knowledge, unique model to study the induction and cooperation of effector autopathogenic T and B cell responses in vivo and in autoimmunity.
In TCRMOG×IgHMOG mice it is unlikely that MOG-specific antibodies themselves are responsible for the initiation of spontaneous Devic-like disease, since we did not observe a substantial difference in the level of MOG-specific antibodies between healthy and diseased TCRMOG×IgHMOG mice (Figure (Figure4C),4C), and the transfer of serum from TCRMOG×IgHMOG mice to TCRMOG mice did not induce EAE or Devic-like disease (data not shown). However, the combination of enhanced antibody production and a robust MOG-specific pathogenic T cell response might be responsible for the fulminant disease observed in TCRMOG×IgHMOG mice. In addition, the antibody and T lymphocyte specificity might play a role in determining the specific lesion distribution pattern as follows: First, the amount of antigen might be important. Previously we reported that there is relatively more MOG expressed in the optic nerves than in the spinal cords of C57BL/6 mice (23). Here we additionally demonstrated that mRNA for MOG was more abundant in the spinal cords than the brains of these mice. It is therefore possible that a differential expression of MOG protein results in predominant attack of optic nerves and spinal cords in TCRMOG×IgHMOG mice because there are relatively higher levels of the target antigen. Second, the accessibility of the autoantigen might be a critical factor. While the white matter in the spinal cord is on the outside and bathed in the cerebrospinal fluid, the gray matter is on the outside in the cerebellum and cerebrum. In the TCRMOG×IgHMOG mice, the inflammatory infiltrates in the subarachnoid space surrounding the spinal cord and around the optic nerves were more numerous than generally observed in induced EAE models and TCRMOG mice with spontaneous EAE. Therefore, because MOG antigen might be more accessible in the spinal cords and optic nerves, enhanced MOG-specific T and B cell responses generated in TCRMOG×IgHMOG mice might preferentially attack these CNS areas, resulting in selective lesion distribution similar to that of NMO.
Autoimmune diseases have a complex pathogenesis, and the TCR transgenic mice have proven to be valuable tools to investigate the behavior of self-reactive T cells, their interactions with their antigenic targets, and their induction of disease. Several TCR transgenic mice specific for autoantigens have been generated, and many of the autoantigen epitopes recognized by transgenic T cells have restricted expression in the CNS (49). Despite the striking increase in the frequency of self-reactive T cells in most of the TCR transgenic mouse lines, however, it is noteworthy that spontaneous autoimmune diseases do not occur in 100% of the animals. For example, 60% of PLP TCR transgenic mice develop spontaneous EAE (50), and the incidence of spontaneous EAE in MBP aa 1–11–specific TCR transgenic mice ranges from 0% to 14% depending on the line (51, 52). While 47% of TCRMOG mice develop spontaneous clinical optic neuritis, only 4% of these animals develop spontaneous EAE (23). The initiation of disease most likely requires adequate priming of the autoimmune response and/or control of regulatory T cell responses. In one of the MBP-specific TCR transgenic lines, spontaneous disease occurs in 15% of the animals housed in a pathogen-free facility and 43% of those housed in conventional facilities (51), suggesting a role for infectious agents as triggers of the disease. At the present time, the environmental factor(s) in specific pathogen–free facilities that may initiate spontaneous EAE and are responsible for differences in incidence of spontaneous EAE remain unidentified. Moreover, the mechanisms underlying initiation of spontaneous EAE by infectious agents remain to be elucidated. The inflammatory response generated by infection could trigger the autoreactive T cell response and initiate autoimmune diseases. Alternatively, molecular mimicry between microbial and self antigens has been proposed as a mechanism for the development of EAE (53). Of particular relevance in this regard to the present study is the fact that there are a large number of molecular homologs of the extracellular domain of MOG. Specifically, the butyrophilin gene products exhibit linear aa sequence homologies with MOG ranging from 35% to 50% and have in some cases been shown to exhibit molecular mimicry with MOG (54). In TCRMOG×IgHMOG mice, autoreactive MOG-specific T cells may recognize an antigen mimicking MOG that is present in the peripheral immune compartment and then get further activated and amplified by MOG presented by large numbers of MOG-specific B cells. This might then lead to a high frequency of spontaneous disease. Since the restricted expression of MOG to the CNS has recently been challenged (55), it is also possible that MOG itself expressed in the periphery or released into the peripheral circulation could be presented to autoreactive T cells by MOG-specific B cells very efficiently. Alternatively, the increased incidence of spontaneous EAE observed in MBP TCR transgenic mice crossed to RAG-deficient mice and therefore lacking regulatory T cells expressing endogenous TCR shows the importance of regulatory cells in keeping autoreactive T cells in check (49). Since only 4% of TCRMOG mice and 60% of TCRMOG×IgHMOG mice develop spontaneous EAE (23), we compared the number of regulatory cells between these 2 strains. Although we found that the number of CD4+Foxp3+ T cells was low in TCRMOG mice (around 2%), it was similar in TCRMOG×IgHMOG mice. Additional experiments are necessary to determine whether B cells can modulate the activity of regulatory cells.
In summary, TCRMOG×IgHMOG mice represent a useful model to determine the factors responsible for different lesion localization and to study the role of T cell/B cell cooperation in the induction of autoimmune pathologies. In addition, since TCRMOG×IgHMOG mice develop a high incidence of spontaneous disease involving both T and B cells, these mice may be useful for testing the biology of T cell/B cell cooperation in the genesis of autoimmune diseases and for testing novel therapies for MS and Devic disease.
C57BL/6 mice transgenic for a TCR with specificity for the peptide MOG aa 35–55 (TCRMOG mice; also referred to as 2D2 mice) were extensively described previously (23). As control for TCRMOG mice, we used TCROVA mice (also referred to as OT-II mice), which are transgenic for a TCR recognizing the peptide OVA aa 323–339 (The Jackson Laboratory).
IgHMOG mice (also referred to as Th mice), which contain MOG-specific B cells, were generated by replacing the germline JH locus with the rearranged Ig heavy-chain (IgH) variable (V) gene of a pathogenic MOG-specific monoclonal antibody (24). The B cells exclusively express the transgenic H chains, and one-third bind MOG. As control for the IgHMOG mice, we used NP-specific IgHNP mice (also referred to as B1.8 mice; ref. 56).
Mice were housed in a specific pathogen–free, viral antibody–free animal facility at the Harvard Institutes of Medicine. All breeding and experiments were reviewed and approved by the Institutional Animal Care and Use Committee of Harvard Medical School. Serums from 4- to 10-week-old mice were obtained at sacrifice by cardiac puncture.
MOG aa 35–55 peptide (MEVGWYRSPFSRVVHLYRNGK) and OVA aa 323–336 peptide (ISQAVHAAHAEINE) were obtained from the Biopolymer Facility, Center for Neurological Diseases, Brigham and Women’s Hospital. A plasmid construct encoding the extracellular domain of rat MOG protein (MOG aa 1–125) was generously provided by C. Linington (University of Aberdeen, Aberdeen, United Kingdom), and rMOG protein was purified from inclusion bodies (57). OVA was purchased from Sigma-Aldrich.
T and B lymphocytes were isolated from spleens and lymph nodes using anti-CD4 and anti-B220 columns, respectively, according to the manufacturer’s instructions (Miltenyi Biotec). Purity of the positively selected cell populations was >95% as assessed by flow cytometry. For the analysis of B cell proliferation and activation in the presence of antigen-specific T cells, we used T cell lines obtained by 2 rounds of stimulation of lymphocytes from TCR transgenic mice with 10 μg/ml MOG aa 35–55 or OVA aa 323–326 in the presence of irradiated splenocytes and 20 U/ml IL-2.
Cells were grown in RPMI supplemented with 10% FCS, β-mercaptoethanol, l-glutamine, gentamicin sulfate, and penicillin/streptomycin. For thymidine proliferation assay, either 5 × 106 cells/ml total splenocytes or 2.5 × 105 cells/ml purified CD4+ cells plus 2.5 × 106 cells/ml purified B220+ B cells were cultured for 72 hours in 96-well plates in the presence of different concentrations of the rMOG protein or MOG aa 35–55 peptide (range, 0.01–100 μg/ml). Cells were pulsed with 1 μCi 3H-thymidine for the last 16 hours of incubation. Mean thymidine incorporation in triplicate wells was measured using a β counter (LS 5000; Beckman Coulter).
For the assessment of proliferation by CFSE dilution assay, the cells were labeled with CFSE (Invitrogen) at 2 μM in PBS, washed, and cultured at 5 × 105 cells/ml for 2–4 days according to the specific stimulation protocol. Cells were then harvested and analyzed by flow cytometry.
Unlabeled or CFSE-labeled cells were blocked with antibodies to the FcγIII/II receptors (CD16/CD32) to avoid nonspecific staining and were subsequently labeled at 4°C with the following monoclonal antibodies (BD Biosciences): anti-CD23–PE, anti-CD40L–PE, anti-CD69–PE, anti-CD80–PE, anti-CD86–PE, anti-IgM–PE, anti-IgG1–PE, anti-IgG2a–PE, anti-IgG2b–PE, anti-IgG3–PE, anti-CD4–allophycocyanin, and anti-B220–allophycocyanin. Isotype- and concentration-matched control antibodies were used to assess nonspecific staining. Cells were then fixed in 0.3% paraformaldehyde and kept at 4°C until analysis.
For the detection of intracellular cytokines, the cells were restimulated for 5 hours with PMA (50 ng/ml) and ionomycin (1 μM), in the presence of 3 μM monencin for the last 3 hours. The cells were then stained for surface markers and fixed. Subsequently, the cells were permeabilized with saponin and stained with PE-labeled anti-cytokine antibodies resuspended in saponin buffer.
The cells were analyzed on a 4-color FACSort (BD Biosciences), and lymphocytes were gated on forward and side scatter as well as on CD4 and B220 for T and B cells, respectively.
IL-2, IL-4, IL-10, and IFN-γ concentrations were determined in cell culture supernatant by a standard sandwich ELISA by coating 96-well plates with the appropriate anti-cytokine monoclonal antibodies (BD Biosciences), incubating the plates with the culture supernatant overnight, and detecting the bound cytokines with a second biotinylated anti-cytokine antibody and avidin peroxidase. Results were reported in pg/ml according to a standard dilution curve.
For measuring MOG-specific antibodies in serum, 96-well plates were coated with 4 μg/ml MOG in PBS, blocked, and incubated with sera (1:40 and 1:200 dilution for IgM and 1:200 and 1:1,000 dilution for IgG isotypes) for 2 hours. After washing, the serum antibodies retained by the plate-bound MOG were detected by biotinylated, isotype-specific anti-Ig antibodies (BD Biosciences) and avidin peroxidase. Nonspecific binding was assessed using OVA-coated plates as negative control. Results were reported as OD.
Clinical assessment of EAE was performed daily, and mice were scored for disease according to the following criteria: 0, no disease; 1, decreased tail tone; 2, hind-limb weakness or partial paralysis; 3, complete hind-limb paralysis; 4, front- and hind-limb paralysis; 5, moribund state.
Animals were sacrificed after 3 months of observation for development of clinical signs of EAE or after at least 2 weeks of EAE. Brains, spinal cords, and optic nerves were removed and fixed in 10% formalin. Paraffin-embedded sections were stained with hematoxylin and eosin, Luxol fast blue, and reticulin and by Bielschowsky preparation for light microscopy. Immunohistochemistry was performed on cryosections using antibodies to CD4, TCR Vβ11, and B220.
Ig serum levels in TCRMOG×IgHMOG mice versus TCRMOG or IgHMOG control mice were compared by 2-tailed Student’s t test. P values less than 0.05 were considered to be statistically significant.
We thank Hartmut Wekerle, Andreas Holz, and Gurumoorthy Krishnamoorthy for providing IgHMOG mice; Klaus Rajewsky for providing IgHNP mice; Christopher Linington for providing the soluble MOG-encoding plasmid; and Michael Nazareno for assistance with screening of mice. This work has been supported by grants from the NIH (NS046414 to R.A. Sobel and NS38037, AI44880, NS30843, and NS45937 to V.K. Kuchroo and E. Bettelli) and the National MS Society (NIH grant RG2571 to V.K. Kuchroo and E. Bettelli). V.K. Kuchroo is a recipient of a Javitz Neuroscience Investigator Award from the NIH.
Nonstandard abbreviations used: IgHMOG mice, MOG-specific Ig heavy-chain knock-in mice on a C57BL/6 background; IgHNP mice, NP-specific Ig heavy-chain knock-in mice on a C57BL/6 background; MBP, myelin basic protein; MOG, myelin oligodendrocyte glycoprotein; NMO, neuromyelitis optica; NP, nitrophenyl; PLP, proteolipid protein; rMOG, recombinant MOG aa 1–125; TCRMOG mice, MOG-specific TCR transgenic mice on a C57BL/6 background; TCROVA mice, OVA-specific TCR transgenic mice on a C57BL/6 background.
Conflict of interest: The authors have declared that no conflict of interest exists.
Citation for this article: J. Clin. Invest. 116:2393–2402 (2006). doi:10.1172/JCI28334.
Estelle Bettelli and Dominique Baeten contributed equally to this work.
See the related Commentary beginning on page 2313.