Effects of Tocilizumab Therapy on Circulating B Cells and T Helper Cells in Patients With Neuromyelitis Optica Spectrum Disorder

Tocilizumab, a humanized anti-IL-6 receptor monoclonal antibody, showed its therapeutic efficacy on neuromyelitis optica spectrum disorder (NMOSD). To assess the immunological effects of this drug on B cells, follicular T helper (Tfh) cells, and peripheral T helper (Tph) cells in patients with NMOSD, peripheral B cell and Tfh cell phenotypes were evaluated in 26 patients with NMOSD before and after tocilizumab treatment by nine-color flow cytometry, as well as the expression of costimulatory and co-inhibitory molecules on B cells. Results showed that the frequency of CD27+IgD− switched memory B cells, CD27-IgD- double-negative B cells, and CD27highCD38high antibody-secreting cells was increased in patients with NMOSD. Tocilizumab treatment led to a significant shift of B cells to naïve B cells from memory B cells after 3 months. Three markers on B cells associated with T-cell activation (i.e., CD86 CD69, and HLA-DR) were downregulated after tocilizumab treatment. The frequencies of total Tfh and Tph cells were decreased, whereas that of follicular regulatory T cells tended to increase. Intrinsic increased PD-L1 and PD-L2 expression was characteristic of B cells in patients with NMOSD. Tocilizumab selectively restored PD-L1 on B-cell subsets. These results provided evidence that tocilizumab enhanced B- and T-cell homoeostasis by regulating B-cell differentiation and inhibiting lymphocyte activation in patients with NMOSD.

Co-inhibitory signals characteristic of programmed cell death 1 (PD-1) are expressed on T cells as feedback for activated responses (10,11). Engagement of PD-1 ligand 1 (PD-L1) and 2 (PD-L2) also plays important regulatory roles in immune responses (12). However, little is known about the kinetics of PD-1, PD-L1, and PD-L2 expression on B cells from patients with NMOSD.
Pleiotropic pro-inflammatory interleukin (IL)-6 drives disease activity and may contribute to abnormalities in B and T cells (13). Therapeutic agents targeting the IL-6 axis are effective in settings of acute and chronic inflammation (14). Recently, we reported a phase 2 randomized controlled trial showing that tocilizumab, a humanized anti-IL-6 receptor monoclonal antibody (mAb), significantly reduced the risk of relapse compared to azathioprine in NMOSD. Tocilizumab also decreases serum AQP4-IgG titers in patients with highly active NMOSD (15). However, the underlying effects of tocilizumab treatment on circulating B and T cells are unclear. Accordingly, in this study, we performed a detailed analysis of immunological phenotypes in B and T cells before and after tocilizumab treatment in patients with NMOSD.

Recruitment and Enrollment of Patients and Controls
Patients with NMOSD were diagnosed according to 2015 International Panel for Neuromyelitis Optica Diagnosis criteria.
Patients who received tocilizumab treatment at Tianjin Medical University General Hospital between November 2017 and May 2019 were enrolled. The inclusion criteria in this study were as follows: those who received routine infusion of tocilizumab at a dose of 8 mg/kg every 4 weeks for at least 3 months. At study entry, patients were permitted to continue the baseline treatment of corticosteroids. The exclusion criteria were as follows: 1) patients who were concomitantly treated with oral immunosuppressants, such as azathioprine, mycophenolate mofetil, and tacrolimus; and 2) patients who received rituximab treatment for less than 6 months from initial tocilizumab infusion. We also included 20 age-and sex-matched healthy controls (HCs) who had no malignancy or autoimmune disorder and did not receive immunosuppressive therapy. The Institutional Review Board of Tianjin Medical University General Hospital reviewed and approved this study. Each participant provided written informed consent.

Flow Cytometry Analysis
All patients who received routine tocilizumab treatment were registered for immunophenotyping analysis. Peripheral blood was collected at baseline, 1 month, and 3 months after initiation of treatment. All fresh samples were immediately processed. Briefly, peripheral blood mononuclear cells (PBMCs) were prepared by density-gradient centrifugation, resuspended in phosphate-buffered saline (PBS)/3% human IgG (Bax International Inc., Vienna, Austria) to block Fc receptors and prevent nonspecific antibody binding, and then incubated for 15 min at 4°C in the dark. Subsequently, the cells were washed with PBS containing 1% bovine serum albumin. Background fluorescence was assessed using appropriate isotype-and fluorochrome-matched control mAbs.
Prior to the measurements, 1 mL of 300 nM 4′,6-diamidino-2phenylindole (Invitrogen, Carlsbad, CA, USA) was added to exclude dead cells. The stained samples were assessed by flow cytometry using a FACS Aria III flow cytometer (BD Biosciences, San Jose, CA, USA). The results were analyzed using FlowJo software (version 10).

Statistical Analysis
All statistical analyses were performed using GraphPad Prism v8 and R software. Continuous and ordinal variables are presented with medians (interquartile ranges [IQRs]) or means (standard deviations [SDs] or standard errors of the means [SEMs]). Shapiro-Wilk tests were used to determine whether the variables had a normal distribution. For comparisons between patients and HCs, unpaired, two-tailed Student's t tests or Mann-Whitney U tests were used. Paired Student's t tests (parametric) or Wilcoxon matched pairs tests (nonparametric) were used to compare baseline and post-treatment results in patients with NMOSD. For multiple comparisons, the Bonferroni-Dunn adjustment was used for analysis of repeated-measures. P values less than 0.05 were considered significant.
Relapse was defined as new onset of neurological symptoms or worsening of existing neurological symptoms with an objective change on neurological examination that persisted for more than 24 h, with signs and symptoms attributable solely to NMOSD, and preceded by at least 30 days of clinical stability. By the end of the study, three (11%) of 26 patients in the tocilizumab group had a relapse.
To sum up, the frequencies of total Tfh and Tph cells were decreased, whereas that of follicular regulatory T cells tended to increase. The subtypes of Tfh1, Tfh17 and Tph1 cells decreased significantly.

Tocilizumab Inhibited the Activation of B Cells in Patients With NMOSD
Next, we studied the expression of the costimulatory molecules CD86 and CD69 as markers of activated B cells that modulate Tcell signaling. No significant changes in the mean fluorescence intensity (MFI) of CD86 or CD69 were observed in CD19 + B cells after tocilizumab treatment for 1 month (not shown); however, the MFI of CD86 was decreased after 3 months (488.2 ± 30.47 versus 392.3 ± 32.62, respectively; P = 0.0205)  and that of CD69 decreased at 3 months (366.5 ± 25.87 versus 291.6 ± 22.50, respectively; P = 0.0387; Figures 5A-F). There were no significant differences in the expression of BAFF-R (2384 ± 124.9 versus 2336 ± 96.75, respectively; P = 0.7620) before and after treatment with tocilizumab ( Figures 5G, H). In contrast, HLA-DR expression decreased significantly after 3 months of tocilizumab treatment (19017 ± 1174 versus 14550 ± 1033, respectively; P = 0.0125). In short, three markers on B cells associated with T-cell activation (i.e., CD86 CD69, and HLA-DR) were downregulated after tocilizumab treatment.

Effects of Tocilizumab Treatment on the Expression of PD-1 and Its Ligands in Peripheral B Cell Subsets
First, we compared the expression of PD-1 and its ligands PD-L1 and PD-L2 on B-cell subsets in HCs and patients with NMOSD. SWM B cells and ASCs exhibited significantly higher PD-1 expression in patients with NMOSD than that in HCs (NMOSD  (Figures 6C-F).
Next, we compared the expression of PD-1 and its ligands on B-cell subsets in the patients with NMOSD after 3 months of treatment with those in HCs. SWM B cells and ASCs exhibited significantly higher PD-1 expression in patients with NMOSD after treatment than that in HCs  Finally, we compared the expression of PD-1 and its ligands on B-cell subsets in the patients with NMOSD before and after 3 months of treatment. Statistical analysis shows that the expression of PD-L1 on SWM cells and ASCs after treatment was significantly lower than those before treatment (592.9 ± 32.93 versus 503.7 ± 28.82, P=0.0469 and 695.0 ± 45.53 versus 571.4 ± 36.32, P=0.0388 respectively) ( Figure 6B). Additionally, there was no significant differences in the expression of PD-1 and its ligands on B-cell subsets (Figures 6D, F).
As a whole, tocilizumab-treated patients showed little changes in PD-1 expression among all B-cell subsets compared with the baseline. By contrast, treatment with tocilizumab does have a tendency to reduce PD-L1 and PD-L2 expression on B-cell subsets. For instance, the expression of PD-L1 on SWM cells and ASCs after treatment was significantly lower than those before treatment ( Figures 6D, F).

DISCUSSION
The differential effects of disease-modifying drugs (DMDs) on peripheral blood B-cell subsets and differentiation-related Th cells have been investigated in multiple sclerosis (16)(17)(18). However, few studies have reported the effects of DMDs in NMOSD.
In this study, we found that in vivo blockade of the IL-6 receptor decreased lymphocyte activation, altered B-and T-cell homoeostasis in patients with NMOSD, and promoted the normalization of abnormal B-and T-cell subsets observed at baseline. Specifically, tocilizumab treatment led to an increase in the number of naïve B cells and decreases in the number of memory B cells and ASC B cells. Similarly, the frequencies and absolute numbers of abnormally reduced Tfh cells increased.
Decreases in the number of memory B cells were consistent with those reported with tocilizumab in patients with rheumatoid arthritis in a previous study (19). Unlike in rheumatoid arthritis, however, we also observed a significant decrease in the number of ASCs, probably because of the more pronounced abnormalities of these cells in patients with NMOSD. Naïve B cells are also dysregulated in patients with NMOSD, as reflected by the lower frequency of naïve B cells. Remarkably, the frequency of naïve B cells was normalized in response to tocilizumab, and the changes in naïve B cells were attributed to the recovery of the reduced number of memory B cells, ASCs, and DN B cells and, to a lesser degree, to a decline in the number of USW B cells. Second, it is known that IL-6 plays an important role in the terminal differentiation of B cells. Thus, IL-6 receptor blockade may lead to a failure of differentiation from naïve B cells to memory B cells.
Double-negative B cells are thought to be mature antigenexperienced B cells with developmental marker expression profiles similar to those of post-switch memory B cells (20). The proportion of DN B cells is increased in patients with multiple sclerosis, resulting in increased production of pro-inflammatory and cytotoxic cytokines following ex-vivo stimulation (21,22). Consistent with a previous study (23), we also observed expanded DN B cells in patients with NMOSD, suggesting the possible involvement of DN B cells in the pathogenesis of NMOSD (19,24). In this study, tocilizumab treatment effectively reduced the numbers of DN B cells and may facilitate the inhibition of inflammation in NMOSD. DN B cells were further divided into two subgroups in a previous study on systemic lupus erythematosus (SLE), DN1 cells (CXCR5 + CD19 + IgD -CD27 -) and DN2 cells (CXCR5 -CD19 + IgD -CD27 -). DN2 cells express T-bet and CD11c, and their activation is mediated by hyperresponsiveness to TLR7 and leads to the generation of autoreactive ASCs (25). The deeper characterization of these cells, via immunophenotyping, has the potential to identify novel SLE phenotypes that associate with disease-activity (26). Moreover, the frequency of DN B cells in rheumatic arthritis is inversely correlated to a subsequent good response (27). Additional larger scale studies are needed to determine whether DN B cells may serve as predictors of treatment response to tocilizumab.
In this research, blockade of the IL-6 receptor restores abnormally elevated SWM B cells and ASCs seen at baseline. The decrease of SWM B cells and ASCs may explain the reduction of the pathogenic AQP4-IgG titers found in our previous clinical trial (15), which showed that among AQP4-IgG positive patients, tocilizumab decreased sera AQP4-IgG titers significantly by 50% at the end of the study compared with those at baseline.
In NMOSD, the balance in Tfh cells and related subsets is impaired, making these cells a potential therapeutic target. Consistent with previous studies, we found an upregulation of Tfh cells, particularly Tfh1 cells, in NMOSD patients compared with HCs (6,28,29), while 3 months of treatment with tocilizumab reduced the frequency of Tfh cells, and we found that Tfh cell differentiation was associated with IL-6 and plasmablast formation. In line with our findings, specific targeting of IL-6 using tocilizumab therapy in patients with rheumatoid arthritis can significantly reduce circulating Tfh cell numbers, which are correlated with reduced plasmablast formation (30). In this study, we observed similar results in tocilizumab-treated patients with NMOSD, indicating the therapeutic efficacy of tocilizumab in these patients.
Tph cells may play a role in promoting B-cell responses and antibody production within pathologically inflamed nonlymphoid tissues. The possible mechanism lies in that CXCL13 and IL-21 produced by Tph cells may recruit both Tfh and B cells and thus promote local autoantibody production and perhaps modulate other B cell functions such as cytokine production (31). We found that Tph cells were increased in patients with NMOSD. Tocilizumab treatment resulted in reduction of the number of peripheral blood Tph cells. Further research is needed to investigate if this can offer a potential strategy for therapeutic targeting of tissue T cell-B cell interactions.
The role of IL-6 in autoimmune diseases helps to understand the mechanism of the effect of tocilizumab on B-and T-cell subsets. For B cells, IL-6 can promote the terminal differentiation of B cells (32) and contribute to survival of plasma cells (33). IL-6 receptor blockade by tocilizumab is expected to inhibit the differentiation of B cells and survival of plasma cells, as both of them express the IL-6 receptor. For T cells, IL-6 is known to promote early Tfh cell differentiation by transiently inducing B cell lymphoma 6 (BCL-6) expression in CD4 + T cells (34). Further, Tfh cells can promote germinal center formation through the production of IL-21, which sustains BCL-6 expression on B cells and promotes B cell activation, classswitch recombination, and plasma cell differentiation (35). Tph cell is a recently discovered Th cell subset, which can also promote B cell differentiation by secreting IL-21 (36). In general, proinflammatory IL-6 promotes the formation of germinal center B cells and Tfh/Tph cells in cooperation with IL-21.
The PD-1 pathway is one of the most important immune checkpoints and is indispensable for maintaining the homeostasis and tolerance of the immune system. The expression of PD-1 family members on B cells in NMOSD has not yet been investigated in detail. Herein, we found that PD-1, PD-L1, and PD-L2 expression was increased on ASCs and memory B cells in NMOSD patient compared to that in healthy controls, consistent with previous studies in systemic lupus erythematosus (37). Because all B cells are proinflammatory in NMOSD, upregulation of PD-1 and its ligands on B cells may on account of cellular activation. Additionally, PD-L1-deficient B cells in an experimental autoimmune encephalomyelitis mouse model exhibit aggravated autoimmunity (38), suggesting that the PD-1 pathway may have roles in the ablation of autoimmunity. In our cohort, tocilizumab treatment selectively downregulated PD-L1 expression on ASCs and memory B cells in patients with NMOSD; this treatment could block or control autoimmune responses and help restore homeostasis. Although B cells expressing PD-1 may diminish after anti-CD19 or anti-CD20 mAb treatment, the effects of tocilizumab on PD-1-related coinhibitory signaling are persistent. This indicates that the mechanisms of tocilizumab differ from those of rituximab or inebilizumab regarding inhibition of disease activity. Little is known about the expression and kinetics of PD-1, PD-L1, and PD-L2 by B lymphocytes from NMOSD patients. Not only B cells that express PD-L1 and PD-L2 and T cells expressing PD-1 interact via this pathway, but also B cell-expressed PD-1, that could mediate B-B or B-T interactions. We postulate that the PD-1 system on B cells acts as a feedback of neuroinflammation in NMOSD. The elevated PD-1 system on B cells may play a protective role in controlling autoimmunity in active NMOSD. When autoimmune inflammation is suppressed by tocilizumab treatment, the PD-1 system tends to restore.
This study has several limitations. First, the follow-up period was relatively short, and the long-term effects on B-cell subsets after tocilizumab treatment were not evaluated. Second, the sample size of patients with tocilizumab treatment was extremely small. Additionally, 3 patients in this study experienced attacks during follow-ups, but the reason is not still clear. High-throughput single-cell sequencing may be warranted to explore heterogeneity of cellular immune function in patients who didn't respond well to tocilizumab. Finally, it is a limitation that the mechanisms of the effects of tocilizumab on Tfh/Tph cell inhibition are not investigated in our study. Further evidence will be warranted on the mechanistic study.

CONCLUSIONS
In summary, we showed that tocilizumab inhibited the activation of mature B cells and Tfh cells but maintained co-inhibitory PD-1 expression on B cells in the peripheral blood of patients with NMOSD. These findings provide important insights into the effects of tocilizumab on the immune mechanism of NMOSD.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by The Institutional Review Board of Tianjin Medical University General Hospital. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
CZ and YL designed the study, collected, analyzed and interpreted the data and drafted and revised the manuscript. YL, PZ, and CD performed the research. YL and HZ did statistical analysis. T-XZ, MY, ZH, DJ, and GY collected and analyzed the data. F-DS revised the manuscript critically for intellectual content. All authors contributed to the article and approved the submitted version.