Abstract
BACKGROUND Multiple sclerosis (MS) is an autoimmune disorder of the central nervous system (CNS) that has traditionally been considered T cell-mediated. However, accumulating evidence points to a crucial role for B cells in disease processes. Experimental autoimmune encephalomyelitis (EAE) is a well-established model to study the immune aspects of CNS autoimmunity.
METHODS In order to examine the collaboration of B cells and T cells in EAE, we studied non-obese diabetic (NOD)-background IgH[MOG] mice, whose B cells express a transgenic IgH chain derived from a myelin oligodendrocyte glycoprotein (MOG)-specific antibody. We immunized these and NOD WT controls with the MHC class II-restricted peptide MOG[35-55], which induces a CD4+ T cell-driven response. CNS tissue inflammation and demyelination were assessed histopathologically, and the phenotype of CNS-infiltrating mononuclear cells was studied by flow cytometry. The capacity of IgH[MOG] B cells to present antigen to CD4+ T cells was assessed using in vitro priming assays with MOG[35-55] as the antigen.
RESULTS MOG[35-55]-immunized IgH[MOG] mice rapidly developed severe EAE characterized by leukocytic infiltration and demyelination in the brain, spinal cord and optic nerve. Notably, while the frequency of CD4+ T cells was increased in the CNS of IgH[MOG] with severe disease relative to controls, no differences were observed with respect to the frequency of B cells. Further, IgH[MOG] CNS-infiltrating CD4+ T cells produced significantly higher levels of Th17-associated cytokines GM-CSF and IL-17 compared to those from controls. Mechanistically, IgH[MOG] B cells were better able than WT B cells to elicit inflammatory cytokine production from MOG[35-55]-specific CD4+ T cells in in vitro priming assays.
CONCLUSION These data show that MOG-specific B cells contribute to CD4+ T cell-driven EAE by promoting CD4+ T cell inflammation and recruitment to the CNS.
Background
Multiple sclerosis (MS) is a chronic neurodegenerative disease in which the adaptive immune system launched an attack against central nervous system (CNS) proteins, such as myelin, ultimately resulting in neurological dysfunction and death. MS affects more than 2 million people worldwide [1]. Approximately 80% of patients present an initially relapsing-remitting (RR) disease course for which there are now more than 10 disease-modifying therapies available. However, 30-60% of these (RR) patients will eventually transition to a chronically worsening secondary progressive (SP) phase, for which treatment options are limited [2]. Pathophysiological mechanisms in progressive MS are thus of intense current interest [3].
T cells, and CD4+ T cells in particular, have been the most intensively studied players in the immune pathogenesis of MS. However, it has become increasingly clear that B cells additionally play important roles in MS pathogenesis. Clonally expanded B cells are present in the cerebrospinal fluid (CSF) and MS plaques [4-6], and the presence of meningeal follicles adjacent to cortical lesions is associated with disease progression [7,8]. Further, antibodies against myelin oligodendrocyte glycoprotein (MOG), a key component of myelin, were found in active MS lesions [9]. Crucially, the B cell targeting reagent ocrelizumab (anti-CD20) results in striking improvements in RR-MS, and is the only FDA approved drug for primary progressive MS [10,11]. Intriguingly, antibody-secreting B cells do not express CD20 [12]. This suggests that the principal pathogenic role of B cells in MS might be unrelated to the generation of autoantibodies, and rather to their capacity to interact with other immune cell types, such as T cells.
Experimental autoimmune encephalomyelitis (EAE) is an animal disease that models many of the immune aspects of MS pathogenesis. Use of this model has helped us to understand the role of T cells, and CD4+ T cells in particular, in the initiation and maintenance of autoreactive inflammation in the CNS [13]. However, studies using the transgenic IgH[MOG] mouse strain have indicated that B cells may also play a crucial role in EAE pathology. These mice express a knock-in transgenic IgH chain derived from a MOG-specific antibody; thus, around 30% of their B cells are therefore specific for MOG protein [14]. IgH[MOG] animals develop severe EAE when immunized with either whole MOG protein [14] or with its extracellular domain (MOG[1-125])[15], indicating an important role for MOG-reactive B cells in neuroimmune processes. However the potential mechanisms by which MOG-reactive B cells facilitate T cell-driven pathogenicity, such as in class II-restricted peptide immunization models of EAE, remain incompletely understood.
Here, we studied the co-operative role of B cells and T cells in CNS autoimmunity using IgH[MOG] mice on the non-obese diabetic (NOD) genetic background. We immunized these, and wildtype (WT) NOD mice, with MOG[35-55], a MHC class II-restricted peptide that obligatorily drives a CD4+ T cell response. While WT NOD mice gradually develop a relapsing chronic form of EAE over the course of 80-100 days when immunized with MOG[35-55] [16], we observed that IgH[MOG] mice developed a severe form of EAE within 14 days that was characterized by demyelination and inflammation of the CNS. Disease in these animals was accompanied by an increase in CNS-infiltrating T cells that expressed higher levels of the inflammatory cytokines IL-17 and GM-CSF as compared to wildtype controls. Importantly, IgH[MOG] B cells could prime MOG[35-55]-reactive T cells to produce increased inflammatory cytokines in vitro. Hence, we demonstrate that the collaborative role of B cells and T cells are important for severe CNS autoimmunity.
Methods
Animals
IgH[MOG] mice on the NOD background were a gift from Dr. Hartmurt Wekerle and 1C6 mice were a gift from Dr. Vijay Kuchroo. NOD/ShiLtJ mice were purchased from Jackson Laboratories. The sex of the mice used is indicated in each figure legend.
EAE induction and scoring
NOD and IgH[MOG] mice were immunized subcutaneously with 200µg MOG[35-55] (Feldan), emulsified in incomplete Freund’s adjuvant (BD Difco) that was supplemented with 500 µg M. tuberculosis extract (BD Difco). On day 0 and day 2 post-immunization, mice received 200 ng pertussis toxin (List Biological Laboratories) intraperitoneally. Mice were monitored daily for signs of EAE, which were assessed using a semi-quantitative 0-5 scale: 0; no disease, 0.5; ruffled fur, 1; limp tail, 1.5; mild impairment in gait, 2; severe impairment in gait, 2.5; partial hind limb paralysis, 3; hind limb paralysis, 4; forelimb paralysis, 5; moribund [17]. Pre-onset analyses were conducted a minimum of 5 days post-immunization but before the onset of symptoms. Endpoint analyses were conducted at d14.
Histopathology
EAE mice were euthanized, and CNS tissue and optic nerves were immediately fixed in 4% paraformaldehyde solution. After 24 hours, the tissues were transferred into PBS for paraffin embedding. Paraffin embedded sections were made at 4 µm thickness and stained with Hematoxylin & Eosin (H&E) to detect the infiltration of immune cells, or Luxol fast blue (LFB) to detect demyelination. The images were taken at 10X and 20X magnifications with Nikon Eclipse 80i microscope and were analysed using ImageJ software (NIH).
Measurement of serum immunoglobulin
Blood samples were collected from NOD and IgH[MOG] unimmunized mice, as well as from immunized mice at pre-onset and experimental endpoint, in an EDTA-coated Microvette (Sarstedt). They were centrifuged at 2000g for 20 minutes at 4°C to obtain plasma. The concentration of plasma IgM was analysed using the HRP C57BL/6J Mouse Clonotyping System (Southern Biotech) according to the manufacturer’s instructions.
Isolation of CNS-infiltrating mononuclear cells
Mice were euthanized and perfused intracardially with PBS. Brain and spinal cord were dissected from the skull and vertebral column respectively and were prepared as previously described [17]. Briefly, CNS tissues were digested with liberase (Roche) and DNAse I (Sigma) and cells were enriched using a 35% Percoll (GE Healthcare) gradient.
Flow cytometry
Single cell suspensions were obtained from spleens, lymph nodes and CNS of EAE mice. For detection of surface antigens, cells were stained with Fixable Viability Dye (eBioscience) and incubated with Fc Block (Biolegend) prior to staining with antibodies against surface antigens (CD45, CD4, CD8, CD19, CD11b, CD11c, Ly6G; details in following section). For detection of intracellular cytokines, cells were first stimulated with 50ng ml−1 PMA (Sigma), 1µM ionomycin (Sigma) and 1µL mL−1 GolgiStop (BD) for 4 hours at 37°C, prior to being labeled with viability indicator, Fc Block and relevant surface antigens as above. They were then fixed and permeabilized (Fixation Buffer and Intracellular Staining Perm Wash Buffer, both Biolegend) and stained for intracellular cytokines (IFN-γ, TNF-α, IL-17A, GM-CSF; details in following section). Samples were analyzed on a FACS Aria (BD) and data were analyzed using FlowJo software (Treestar).
Flow cytometry antibodies
The following monoclonal antibodies against mouse antigens were used: CD45, clone A20 (Biolegend); CD11b, clone M1/70 (eBioscience); CD11c, clone N418 (Biolegend); Ly6G, clone 1A8 (BD Biosciences); CD4, clone RM4-5 (eBioscience); CD8, clone 53-6.7 (Biolegend); CD19, clone 1D3 (eBioscience); IFN-γ, clone XMG1.2 (eBioscience); TNF-α, clone MP6-XT22 (eBioscience); IL-17a, clone TC11-18H10.1 (Biolegend); GM-CSF, clone MP1-22E9 (eBioscience); CD40, clone 3/23 (Biolegend); CD80, clone 1610A1 (Biolegend).
Antigen presentation assay
Single cell suspensions were obtained from the spleens of unimmunized NOD and IgH[MOG] mice. Cells were labeled with CD43 (Ly-48) Microbeads (Miltenyi), and CD43+ leukocytes (all leukocytes except resting B cells) were depleted on a magnetic MACS column (Miltyeni). Unlabelled CD43− cells were collected and were subsequently stained with anti-mouse CD19. CD19+ B cells were purified using high-speed cell sorting. In parallel, CD4+ T cells were purified from the spleens of 1C6 mice using mouse CD4 MicroBeads (Miltenyi) and labeled with CellTrace Violet (CTV; Thermo Fisher Scientific). CTV-labeled 1C6 CD4+ T cells were cultured with B cells at a ratio of 1 CD4: 1 B, with 0, 1 or 10µg ml−1 MOG[35-55] for 72 hours. CTV dilution and intracellular cytokine production were assessed by flow cytometry.
Statistical analysis
For comparison of EAE scores on individual days, Mann-Whitney U test was used. Fisher’s exact test was used to measure the frequency of mice attaining ethical endpoints. For immunoglobulin analyses and ex vivo cytokine profiling, unpaired Student t-tests were used. For cytokine production in the in vitro priming experiment, two way ANOVA was used. Two-tailed analyses were used in all instances. All statistical analyses were conducted using Prism software (GraphPad).
Results
IgH[MOG] mice develop severe EAE upon active immunization with MOG[35-55]
When immunized with MOG[35-55], wildtype (WT) NOD strain gradually develop an initial relapse remitting pattern that ultimately transitions to a chronically worsening phase over the course of 60-100 days [16]. NOD-EAE has thus been considered a possible model of SPMS [18,19]. To assess whether the presence of substantial numbers of MOG-reactive B cells could alter this disease pattern, we immunized IgH[MOG] and NOD controls with MOG[35-55]. Strikingly, immunized IgH[MOG] mice rapidly developed severe EAE within 14 days (Figure 1), while WT NOD mice either were disease-free or had only mild symptoms at this timepoint (mean maximal severity, 3.1±0.2 for IgH[MOG], n=27; 0.07±0.07 for WT NOD, n=19; p<0.0001). Strikingly, 55% of immunized IgH[MOG] mice (15/27) attained ethical endpoints and had to be euthanized by d14; no WT NOD attained disease of this severity (0/19; p<0.0001). For the remainder of the study, d14 was therefore treated as the endpoint of immunization protocols.
We next examined CNS tissue damage in actively immunized mice. Immunized IgH[MOG] mice displayed immune infiltration the cerebellum (Figure 2A) and pons (Figure 2B), as well as the spinal cord (Figure 2C) and optic nerve (Figure 2D). Demyelination was also observed in these tissues (Figures 2E-H). By contrast, there was little to no infiltration or demyelination of CNS tissue observed in immunized WT NOD mice sacrificed in parallel. Our data thus show that the presence of myelin-reactive B cells can exacerbate CNS autoimmunity and tissue damage when EAE is induced in a CD4+ T cell dependent manner.
Decrease in plasma IgM in immunized IgH[MOG] mice
Antigen-specific antibody (Ab) secretion is the primary function of B cells. Further, oligoclonal immunoglobulin (Ig) banding in cerebrospinal fluid (CSF) is an important diagnostic marker for MS [20]. Upon initial activation, B cells first secrete antigen-specific Abs of the IgM isotype [21]. The Ab secreted by a given B cell clone can later “switch” to a different isotype based on the presence of differentiation cues in the local milieu, such as cytokines secreted by T cells [22]. We therefore wanted to examine whether changes in total serum IgM might reflect the rapid nature of the disease observed in IgH[MOG] mice. We collected sera from immunized NOD versus IgH[MOG] at pre-onset (PO) and at disease endpoint (EP), and additionally from unimmunized (UI) mice of both strains. We then analyzed these sera for the presence of IgM. We found that the concentration of plasma IgM levels was significantly reduced in IgH[MOG] mice at endpoint relative to NOD controls, indicative of increased switching to secondary isotype subclasses in the context of severe disease (Figure 3). These data indicated that exacerbated disease in IgH[MOG] mice is accompanied by isotype class switching and therefore pointed towards the potential collaboration of MOG-reactive B cells with other immune cell types.
Increased presence of immune cells in the CNS of immunized IgH[MOG] mice
We next wanted to examine whether there were differences between NOD and IgH[MOG] mice in the composition of the immune cells infiltrating the CNS. We first sacrificed mice at the pre-onset stage (d5-post immunization) and analyzed the percentage of immune cells (CD19+ B cells, CD4+ T cells, CD8+ T cells, CD11b+CD11c+ dendritic cells, CD11b+CD11c−Ly6G− macrophages and CD11b+Ly6G+ neutrophils) in both CNS and spleen. At this early timepoint, differences in the frequency of such cells were modest in both CNS (Figure 4A) and spleen (Figure 4B), although we did observe an increase in the relative proportion of macrophages infiltrating the CNS of IgH[MOG] mice (Figure 4A).
We then assessed the frequency of immune cell proportions at disease endpoint (d14). Strikingly, the proportion of CD4+ T cells was significantly increased in the CNS of IgH[MOG] mice relative to NOD (Figure 4C). CD8+ T cells, dendritic cells and neutrophils were also more prevalent in the CNS of IgH[MOG] (Figure 4C), though overall frequency of these cells were low in all mice studied (<5%). The frequency of CD4+ T cells was also higher in the spleens of IgH[MOG] mice at endpoint, suggesting their expansion in these animals (Figure 4D). Interestingly, despite having an antigenic repertoire heavily skewed towards MOG reactivity, immunized IgH[MOG] mice did not display an increased frequency of B cells in either the CNS (Figure 4C) or spleen (Figure 4D) at endpoint. These findings suggested that while IgH[MOG] B cells might be important in driving T cell expansion and infiltration into the CNS, they might themselves be of secondary importance to the development of severe EAE in IgH[MOG] mice.
CNS-infiltrating IgH[MOG] CD4+ T cells produce increased levels of IL-17 and GM-CSF
As we had observed an elevated frequency of CD4+ T cells in the CNS of sick IgH[MOG] mice, we next examined the capacity of these cells to produce inflammatory Th1 and Th17 cytokines by flow cytometry, due to the well-established role of these CD4+ effector T cell subsets in EAE [23]. We observed no differences between IgH[MOG] and WT CD4+ T cells in their production of the Th1 signature cytokine IFN-γ in the CNS (Figure 5A). By contrast we saw a striking upregulation of IL-17 production from CNS-infiltrating CD4+ T cells from IgH[MOG] at disease endpoint (Figure 5B). Notably, production of GM-CSF, a key pathogenic cytokine implicated in Th17-driven tissue inflammation, was also augmented in IgH[MOG] CD4+ T cells (Figure 5C), though no differences were noted in the production of TNFα (Figure 5D). In sum, our data showed that Th17 cells were preferentially recruited to the CNS of IgH[MOG] mice with severe EAE.
IgH[MOG] B cells have an augmented capacity to prime inflammatory MOG[35-55]-specific CD4+ T cells
Thus far, we had found evidence of increased expansion and cytokine production from CD4+ T cells in IgH[MOG] mice. As B cells can express MHC class II and are thus capable of presenting antigen to CD4+ T cells [24], we wanted to assess whether IgH[MOG] B cells intrinsically possess an enhanced capacity to prime MOG[35-55]-specific T cell responses. We purified splenic B cells from unimmunized IgH[MOG] or WT NOD mice and, in the presence or absence of MOG[35-55], co-cultured these cells with antigen-inexperienced CD4+ T cells from transgenic 1C6 mice. 1C6 mice are on the NOD background and have class II-restricted T cell receptor specificity to MOG[35-55] [25]. IgH[MOG] and WT B cells did not differ in their capacity to induce CD4+ T cell proliferation in response to MOG[35-55] (Figure 6A). However, IgH[MOG] B cells elicited a greater frequency of CD4+ T cells that expressed IFN-γ, IL-17, and the autocrine T cell growth factor IL-2 (Figure 6B). Altogether, our findings indicated that IgH[MOG] B cells may shape the initial Ag-specific activation of T cells by promoting their generation of inflammatory cytokines.
Discussion
The role of CD4+ T cells in the autoimmune pathogenesis of MS is well-established. Notably, genome-wide association studies (GWAS) have revealed that polymorphisms in the human leukocyte antigen class II region, which restricts the CD4+ T cell repertoire, are strongly associated with MS susceptibility [26]. Further, many of the current treatments of MS are believed to target T cell responses [27-29]. However, the success of the B cell-depleting drug ocrelizumab in both RRMS and PPMS has intensified recent interest in the contribution of B cells to disease processes [30].
IgH[MOG] mice were initially described on the C57BL/6 (B6) and SJL/J genetic backgrounds. On the B6 background, IgH[MOG] mice showed an increased incidence of EAE, relative to WT, when immunized with whole MOG protein [14]. Intriguingly, IgH[MOG] SJL/J mice developed EAE of greater severity than controls when immunized with the myelin-derived epitope proteolipid protein (PLP)[139-154] [14], which induces a relapsing/remitting disease pattern in SJL/J background mice [31]. This suggested that the presence of MOG-reactive B cells could contribute to EAE pathology that was driven by a class II-restricted peptide; however, the nature of a putative collaboration between B cells and T cells in disease processes remained incompletely defined. In our study, we actively immunized IgH[MOG] mice on NOD background with the class II restricted peptide MOG[35-55]. This permitted us to study the contribution of MOG-reactive B cells in a model of EAE that is initiated by CD4+ T cells. While WT NOD mice developed a gradual, chronic MOG[35-55]-driven disease course, with severe symptoms appearing only many weeks after immunization, we found that 80% of IgH[MOG] NOD mice develop severe and frequently lethal disease within 14 days, that was characterized by inflammation and demyelination in brain, optic nerve and spinal cord. Cellular characterization of mononuclear infiltrate showed an increased frequency of CD4+ T cells, but not B cells, in the CNS of IgH[MOG] relative to controls. It was previously shown that IgH[MOG] × 1C6 double-transgenic mice on the NOD background develop spontaneous EAE, while single-transgenic IgH[MOG] or 1C6 mice do not [25]. These findings indicated that the collaboration of both myelin-specific B and T cells in the same animal could induce CNS autoimmunity; however, it was difficult to determine whether B or T cell-driven responses were initially responsible for disease induction in this model. Here, by using a myelin-derived, class II-restricted, immunogen, we show that B cells augment EAE even when CD4+ T cells initiate disease.
Intriguingly, CNS-infiltrating IgH[MOG] CD4+ T cells generated significant amounts of the type-17 cytokines IL-17 and GM-CSF. Several lines of evidence indicate that Th17 cells can shape B cell responses in vivo. Adoptive transfer of myelin-specific Th17 cells induced both B cell isotype class switching and germinal center formation in an IL-17-dependent manner [32]. Further, Th17 cells are crucial for the development of B cell-rich ectopic lymphoid structures [33,34], which are associated with rapid acceleration of the disease. Here, our data suggest that the inverse may also be true and that B cells may reinforce and augment Th17-mediated pathogenicity in the CNS.
A key antibody independent function of B cells is the antigen presentation to T cells, and B cells are particularly efficient at presenting their cognate antigen [35]. Interestingly, we found that in the presence of MOG[35-55], IgH[MOG] B cells were better able than WT B cells to induce the production of inflammatory cytokines from MOG[35-55]-reactive CD4+ T cells. These findings are broadly in line with previous observations that splenocytes from 1C6 × IgH[MOG] mice showed enhanced responses to MOG[35-55] relative to splenocytes from 1C6 single-transgenic animals [25], and they demonstrate that MOG-specific B cells have a greater intrinsic capacity to induce the differentiation of cognate antigen specific T cells. B cell antigen presentation plays a crucial role in autoimmune responses to whole MOG protein, as mice specifically lacking MHC class II expression on B cells (B-MHC-II−/−) are resistant to human MOG-induced EAE [24]. Notably, however, B-MHC-II−/− mice develop EAE in response to MOG[35-55]. Thus, in mice with an unbiased B cell repertoire, B cell-dependent antigen presentation to T cells may play an important role in the processing and presentation of secondary MOG-derived epitopes. However, when a large frequency of MOG-reactive B cells are initially present, as is the case in IgH[MOG] mice, antigen presentation of MOG[35-55] itself by B cells may exacerbate disease severity.
Conclusion
In this study, we provide evidence that in the presence of a B cell repertoire that is skewed towards MOG, NOD background mice develop unusually severe EAE upon immunization with the classic class II-restricted peptide MOG[35-55]. Disease is characterized by an influx of highly inflammatory CD4+ T cells into the CNS. Our findings support a role for myelin-reactive B cells in augmenting T cell-driven CNS autoimmunity.
Declarations
Ethics approval
All mouse breedings and experiments are approved by the Animal Protection Committee of the Centre de recherche du CHU de Quebec – Université Laval (protocols 13-070-2 and 17-090-2).
Consent for publication
Not applicable
Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.
Competing interests
The authors declare no competing interests.
Funding
The work was supported by a Biomedical Operating Grant from the MS Society of Canada (#3036) to MR. PMIAD is supported by a doctoral studentship from the Multiple Sclerosis Society of Canada and PG holds a Ph.D. fellowship from the Fonds recherche du Québec – Santé (FRQS). NB is Junior-1 scholar, and MR is a Junior-2 scholar, of the FRQS.
Authors’ contributions
PMIAD directed the project, conducted experiments and wrote the manuscript. APY and JB conducted experiments. BM and SL conducted histological analyses. PG and NB assisted with serum detection of immunoglobulins. MR supervised the project and wrote the manuscript. All authors read and approved the final manuscript.
Acknowledgements
We thank Hartmut Wekerle and Vijay Kuchroo for providing us with IgH[MOG] and 1C6 mice respectively. We thank Alexandre Brunet and Stéphanie Fiola for technical assistance with flow cytometry and Andre Machado Xavier for ImageJ analysis.
Abbreviations
- Ab
- antibody
- CNS
- central nervous system
- CSF
- cerebrospinal fluid
- CTV
- CellTrace Violet
- EAE
- experimental autoimmune encephalomyelitis
- EP
- endpoint
- GM-CSF
- granulocyte and macrophage colony stimulating factor
- H&E
- hematoxylin & eosin
- IFN
- interferon
- Ig
- immunoglobulin
- IL
- interleukin
- LFB
- Luxol fast blue
- MHC
- major histocompatibility complex
- MOG
- myelin oligodendrocyte glycoprotein
- MS
- multiple sclerosis
- NOD
- non-obese diabetic
- PLP
- proteolipid protein
- PO
- preonset,
- RR
- relapsing/remitting
- SP
- secondary progressive
- TNF
- tumor necrosis factor
- UI
- unimmunized
- WT
- wildtype