Background
Neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune disease of the central nervous system (CNS) characterized by severe optic neuritis, myelitis, and the presence of anti-aquaporin 4 (AQP4) antibody (AQP4-Ab) [
1,
2]. The pathogenicity of AQP4-Ab was further shown by several studies using rodents [
3‐
5] and supported by clinical findings including the therapeutic efficacy of plasma exchange, complement inhibitors, and B cell-depleting therapy [
6‐
8]. However, the mechanisms involved in autoantibody production remain unclear. Previous studies showed that the frequencies of plasmablasts in the peripheral blood and cerebrospinal fluid (CSF) of NMOSD patients were increased [
9,
10]. Recent studies indicated that naïve B cells from NMOSD patients produced autoantibodies or AQP4-Abs, suggesting a deficiency of central tolerance in NMOSD patients [
11,
12].
The production of autoantibodies and defects in B cell tolerance are associated with the pathomechanisms of several autoimmune diseases. Many studies have reported that autoreactive naïve B cells are increased in various autoimmune diseases including NMOSD, suggesting defects in early B cell tolerance checkpoints [
13,
14]. Recently, CD27
−IgD
−CXCR5
−CD11c
+ B cells (DN2 cells) expanded in patients with systemic lupus erythematosus (SLE) were shown to differentiate into antibody-secreting cells during the extrafollicular response [
15]. DN2 cells were reported to be derived from CD27
−IgD
+CXCR5
−CD11c
+ cells, termed activated naïve B cells. These studies suggested that abnormalities of early B cell development may exist in autoimmune pathology.
In this study, we found a decrease in the number of naïve B cells and an increase in switched memory B (SMB) cells and plasmablasts in NMOSD patients compared with healthy controls (HC). Transcriptome analysis of B cell subsets in NMOSD patients revealed that the profiles of B cell lineage transcription factors were skewed towards an antibody-secreting cell-like phenotype. In accordance with this finding, SMB cells from NMOSD patients had a higher potential to differentiate into antibody-secreting cells when cocultured with T cells compared with those from HC. Furthermore, transcriptome analysis revealed that IL-2 signaling was activated, particularly in naïve B cells from NMOSD patients. Indeed, numbers of naïve B cells expressing CD25, a receptor of IL-2, were increased in NMOSD patients and CD25+ naïve B cells exhibited a higher potential to differentiate into antibody-secreting cells compared with CD25− naïve B cells, suggesting that CD25+ naïve B cells are committed to differentiate into antibody-secreting cells. Our results indicated that CD25+ naïve B cells are a novel candidate precursor antibody-secreting cell.
Methods
Patients and controls
Blood was obtained from 24 patients with AQP4-Ab positive NMOSD, 22 patients with multiple sclerosis (MS), 11 patients with myelin oligodendrocyte glycoprotein antibody-associated disease (MOG-AD) and 27 HC. CSF was obtained from 8 patients with NMOSD, 9 patients with MS, and 12 non-inflammatory disease controls (DC) (Table
1). All NMOSD patients met the 2015 NMOSD diagnostic criteria [
16]. All patients with MS had relapsing–remitting MS and had clinically definite MS according to the 2017 McDonald MS Diagnostic Criteria [
17]. Nine patients with MOG-AD fulfilled the 2015 NMOSD diagnostic criteria. The other three did not meet the NMOSD criteria, because they did not develop myelitis
and optic neuritis (optic neuritis alone
n = 2; myelitis alone
n = 1). Serum MOG-Ab levels were measured using a cell-based assay at Tohoku University (Sendai, Japan). For the flow cytometric analysis of CD25
+ or CD25
− populations in B cell subsets, we recruited nine female patients with NMOSD (median age 48.0 years, interquartile range 38.0–54.5) and nine age- and sex-matched HC (median age 47.0 years, interquartile range 39.0–52.5).
Table 1
Characteristics HCs, DCs, and patients with MS, NMOSD, and MOG-AD
Number | 27 | 22 | 24 | 11 | 12 | 9 | 8 |
Age | 38 (22–60) | 34 (22–54) | 46 (24–65) | 36 (17–66) | 51 (30–72) | 39 (23–53) | 43 (39–74) |
Male: female | 6:21 | 5:17 | 2:22 | 3:8 | 6:6 | 2:7 | 1:7 |
disease duration (months) | | 56 (0–118) | 53 (0–173) | 35 (0–78) | | 51 (0–118) | 47 (0–162) |
Age of symptom onset | | 29 (19–52) | 38 (20–63) | 50 (16–66) | | 33 (19–52) | 39 (26–63) |
Medications | | | | | | | |
Naïve | | 15 | 0 | 4 | | 6 | 1 |
IFNβ, n | | 4 | | | | 2 | |
Glatiramer acetate, n | | 2 | | | | 0 | |
Dymethyl fumarate, n | | 1 | | | | 1 | |
Prednisone, n | | 0 | 22 | 7 | | 0 | 5 |
Daily dose of prednisone, median (mg/day) | | | 7 | 5 (3–7) | | | 8 (5–14) |
Tacrolimus, n | | 0 | 4 | 0 | | 0 | 3 |
Azathioprine, n | | 0 | 4 | 0 | | 0 | 0 |
Mizoribine, n | | 0 | 1 | 0 | | 0 | 0 |
| | | | | Normal pressure hydrocephalus n = 1 Cerebral venous thrombosis n = 1 Old cerebral infarction n = 4 causeless sensory abnormality n = 2 Parkinson disease n = 4 | | |
Flow cytometry and cell sorting
Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood by density centrifugation using Histopaque-1077 (Merck, Darmstadt, Germany). PBMCs were stained with antibodies listed in Additional file
4 Table S1. Data were acquired on a BD LSRFortessa (BD Bioscience, Franklin Lakes, NJ, USA) and analyzed using FlowJo 10.6.1 (FlowJo LLC, Ashland, OR, USA).
Cell culture
B cells were isolated from PBMCs by anti-CD19 magnetic bead (Miltenyi Biotec, Bergisch Gladbach, Germany) positive selection. The cells were sorted on a BD FACSAria Fusion (BD bioscience). Sorted B cell populations were cocultured with autologous T cell populations at a ratio of 5:1 in 200 μl of RPMI-1640 supplemented with 10% fetal bovine serum, 2 mM l-glutamine, 50 U/ml penicillin, and 50 mg/ml streptomycin (all from Thermo Fisher, Waltham, MA, USA), and stimulated with lipopolysaccharide (5 μg/ml, Chondrex, Woodinville, WA, USA) and staphylococcal enterotoxin B (1 μg/ml, Sigma-Aldrich, St. Louis, MO, USA) for 7 days. For blocking experiments, IL-21R-Fc Chimera Protein (10 μg/ml, R&D Systems, Minneapolis, MN, USA) and Human IgG1-Fc Protein (10 μg/ml, R&D Systems) as a control, were used. For B cell cultures, sorted cells were stimulated with F(ab’)2 fragment goat anti-human IgG + IgM (H + L) (Jackson ImmunoResearch, West Grove, PA, USA), MEGACD40L (Enzo Life Sciences, Farmingdale, NY, USA), and recombinant human IL-21 (BioLegend, San Diego, CA, USA) for 7 days.
Measurement of IgG
IgG secreted into the culture supernatant was quantitated by sandwich ELISA using an IgG (Total) Human ELISA Kit (Thermo Fisher) according to the manufacturer’s instructions.
RNA-seq analysis
Total RNA was isolated and purified from sorted cells using an RNeasy Micro Kit (Qiagen, Hilden, Germany). RNA-seq libraries were generated with the Ovation SoLo RNA-Seq System, Human kit (NuGEN, Redwood City, CA, USA) using 5 ng of total RNA. The cDNA libraries were sequenced by 50-base single-read sequencing on an Illumina HiSeq 2500 sequencer (Illumina, San Diego, CA, USA). The sequencing run and the base call analysis were performed according to the HiSeq 2500 System Guide with TruSeq SBS kit v3-HS. Raw sequence data were generated with processing by CASAVA-1.8.4 with version RTA 1.17.20.0. Reads were mapped to the hg38 genome with Tophat2. Normalized FPKM values and differential gene expression analyses were generated with Cuffdiff2. Q values (the FDR-adjusted p value after Benjamini–Hochberg correction for multi-testing) lower than 0.05 were considered significant.
RT-qPCR
cDNA was synthesized from 500 ng total RNA using ReverTra Ace qPCR RT Master Mix (Toyobo, Osaka, Japan). Real-time quantitative PCR (RT-qPCR) was performed using a 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with Fast SYBR Green Master Mix (Thermo Fisher Scientific). mRNA levels were normalized to beta-actin (ACTB) in each sample. The specific primers used in this study are listed in Additional file
4: Table S2.
Statistical analysis
Data were analyzed using Prism 7 software (GraphPad Software, La Jolla, CA, USA) and differences between groups were analyzed using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test, the Mann–Whitney U test, or the Wilcoxon matched-pairs test. The significance level was set at P < 0.05. Correlations between two variables were analyzed using Spearman’s rank correlation test.
Discussion
In this study, we showed that the expressions of the key transcription factors of B cell lineage, BCL6, PAX5, and BACH2 were reduced and the expression of PRDM1, an essential transcription factor for differentiation into antibody-secreting cells, was significantly upregulated in NMOSD patients. BCL6 was reported to repress the expression of PRDM1 to maintain B cell fate [
26,
27]. PAX5 is an important transcription factor that maintains B cell identity and induces the expression of transcription factors, such as BACH2, which is associated with maintaining B cell fate [
28,
29]. BACH2 inhibits differentiation into antibody-secreting cells by repressing the expression of PRDM1 [
30]. Our results suggested that the regulation of transcription factors that control B cell fate was altered, and that the transcriptional profiles of non-memory B cells in NMOSD patients were already skewed towards antibody-secreting cells. Recently, it was reported that naïve B cells from NMOSD patients differentiated into antibody-secreting cells, which secreted AQP4-Ab in vitro [
11]. Another study showed that naïve B cells from NMOSD patients contained significantly higher frequencies of autoreactive B cells compared with HCs, suggesting impaired B cell tolerance in the early phase of B cell maturation [
12]. In accord with these studies, the abnormality of transcriptional factors in naïve B cells suggested the expansion of an autoreactive subpopulation in NMOSD naïve B cells, which are committed to differentiate into antibody-secreting cells.
IL-2 regulates the differentiation and function of CD4
+ T cells, especially regulatory T cells [
31]. However, the roles of IL-2 in B cell-differentiation and function, particularly in autoimmune conditions, are unknown. IL-2-stimulation following stimulation with BCR cross-linking by CpG and CD40L repressed the expression of BACH2 and primed naïve B cells obtained from HCs to differentiate into plasma cells in vitro [
25]. The culture conditions required for the production of autoantibodies from naïve B cells obtained from NMOSD patients included IL-2 as well as CD40L, TNF-α, IL-1β, IL-21, and a Toll-like receptor 7 agonist [
11]. In accordance with these findings, we demonstrated CD25
+ naïve B cells differentiated more efficiently than CD25
− naïve B cells when stimulated with BCR cross-linking, CD40L, IL-21, and IL-2. These results suggested that CD25
+ naïve B cells are committed to differentiate into antibody-secreting cells, rather than a “naïve” population. Furthermore, their expansion may result in an increase of SMB cells and plasmablasts and the continuous production of autoantibodies in NMOSD patients. Autocrine IL-2 was reported to be involved in the survival and differentiation of T cells [
32‐
34]. However, to the best of our knowledge, there is no evidence for the production of IL-2 from B cells. In addition, our RNA-seq analysis showed that B cells did not express IL-2 mRNA (data not shown). Therefore, we assumed that T cells were the major source of IL-2 in the differentiation of naïve B cells.
CD19
+CD27
−IgD
− DNB cells are another candidate antibody-secreting cell precursor. DNB cells were reported to be increased in the peripheral blood of SLE patients [
35,
36]. Recently, autoreactive CD27
−IgD
−CXCR5
−CD11c
+ DN2 cells were reported to be expanded in SLE patients and differentiated into antibody-secreting cells under Toll-like receptor 7 signals [
15]. DN2 cells are thought to be differentiated from CD27
−IgD
+CXCR5
−CD11c
+ activated naïve B cells, which are also increased in SLE patients. In our analysis, CD25
+ naïve B cells did not express CD11c, suggesting they are a distinct population from the previously reported activated naïve B cells. In autoimmune diseases, multiple pathways might be involved in the differentiation of naïve B cells to antibody-secreting cells, and further studies are needed to elucidate which pathway may be dominant in each disease.
T
PH cells are a recently reported novel subpopulation of CD4
+ helper T cells, which induce the differentiation of B cells into antibody-secreting cells [
21] and were identified in the peripheral blood and synovial tissues of rheumatoid arthritis patients. T
PH cells produce IL-21 and activate B cells to produce antibodies, similar to T
FH cells. However, unlike T
FH cells, T
PH cells do not express CXCR5, which is required to enter germinal centers, but do express other chemokine receptors, CCR2, CXCR1, and CCR5, allowing migration into inflamed tissues [
21]. These findings suggest that T
PH cells are involved in extrafollicular T–B cell interactions and B cell differentiation. Several studies recently demonstrated an increase of T
PH cells in the peripheral blood of patients with various autoimmune diseases [
22‐
24,
37,
38]. In the current analysis, we report a significant increase of T
PH cells in the peripheral blood of NMOSD patients, which was not observed in MS and MOG-AD patients. Furthermore, T
PH cells from NMOSD patients induced the differentiation of SMB cells into antibody-secreting cells more efficiently than T
PH cells from HC, and this reaction was dependent on IL-21. These results indicated that T
PH cell-SMB cell interactions in extra-follicle regions, such as the intrathecal space might be important for the production of pathogenic AQP4-Ab. In addition to the abnormalities of B cells, numbers of abnormal T cells are assumed to be increased and to enhance the production of pathogenic antibodies in autoimmune diseases including NMOSD.
MOG is a glycoprotein exclusively expressed on the plasma membrane of oligodendrocytes and myelin in the CNS. Recent advances in detection methods using cell-based assays revealed that MOG-Ab was present in the peripheral blood of various inflammatory CNS diseases, especially pediatric diseases including acute disseminated encephalomyelitis [
39]. However, the pathogenicity of MOG-Ab has not been fully demonstrated [
40]. MOG-Ab positive NMOSD patients were reported to have different clinical features from AQP4-Ab positive NMOSD patients in terms of disease course, responsiveness to therapy including corticosteroids, and prognosis, suggesting that different pathomechanisms may exist between the two diseases [
41]. Recently, it was reported that T
FH cells in the peripheral blood of MOG-Ab seropositive patients were significantly increased compared with HCs [
42]. However, we did not observe a significant difference in the frequencies of T
PH, T
FH, and B cell subsets between MOG-AD patients and HCs in contrast to AQP4-Ab seropositive NMOSD patients. Our results suggest that the pathomechanisms of MOG-Ab production are different from those of AQP4-Ab positive NMOSD, in which the pathogenic autoantibody is continuously produced and other coexisting autoimmune diseases occur more frequently than MOG-AD [
41,
43]. Because MOG was reported to not be expressed in the thymus [
44], MOG-Ab may be produced secondarily during inflammation of the CNS. Further studies are needed to clarify the role of MOG-Ab in the pathogenesis of neuroinflammatory diseases.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.