Regulatory B cell imbalance correlates with Tfh expansion in systemic sclerosis

Objective. Systemic sclerosis (SSc) is an autoimmune disease with fibrosis, microangiopathy and immune dysfunc tion. B cell abnormalities characterised by autoantibody production and poly clonal B cell activation play an impor tant role in the pathogenesis of SSc. We previously identified an expansion of functional and activated circulating T follicular helper (cTfh) cells in SSc patients. The aim of this study was to analyse the frequency of regulatory B (Breg) cell subsets and the correlation with Tfh in SSc patients. of SSc.


Introduction
Systemic sclerosis (SSc) is an autoimmune disease characterised by fibrosis, microangiopathy and immune dysfunction. The exact pathophysiology of SSc is not well established, and mechanisms leading to dysimmunity remain to be established (1,2). B lymphocytes are involved in the pathophysiology of SSc and several autoantibodies can be detected in the sera of SSc patients including anti-DNA topoisomerase I (anti-topo I) and anti-centromere antibodies (3). It has been previously reported that B cell homeostasis is disturbed in SSc patients, with increased naïve and reduced numbers of memory B lymphocytes (4,5); as well as an increased level of the B cell activating factor (BAFF) cytokine in the blood of SSc patients (6,7). In addition, B lymphocytes in patients with SSc secrete high levels of IL-6 and TGF-β and are able to promote skin fibrosis (8). The beneficial effect of B cell depletion with rituximab in SSc patients further supports the involvement of B cells in disease development and/or progression (9). The maintenance of immune tolerance and prevention of autoimmune induction is exerted by various regulatory B cell subpopulations which in humans, include CD24 hi CD38 hi transitional B cells and CD24 hi CD27 + B cells, the human equivalent of murine B10 cells (10,11). Another subset of B lymphocytes, the CD21l ow B cells (CD19 + CD27 -CD 21 low ) has been found to be expanded in several autoimmune diseases such as lupus erythematosus, Sjögren's syndrome and SSc, and was involved in the development of chronic graft-versus-host disease (5,(12)(13)(14)(15). We recently reported that circulating T follicular helper (cTfh) cells were increased in SSc patients and they presented an activated phenotype with the were significantly decreased in SSc compared to healthy controls (Fig. 1B-D). The absolute numbers and frequencies of plasmablast B cells were similar in both groups (Fig. 1D). When examining the CD21 low B cell subset, we observed that the frequency and absolute numbers were significantly increased in SSc patients compared to healthy controls: 5.3±4.3% versus 2.3±1.3% (p<0.001) and 10.50±8.9 cells/µL versus 3.7±1.8 cells/μL (p<0.001), respectively ( Fig. 1E-F). By contrast, the frequency of Breg cells CD24 hi CD38 hi and CD24 hi CD27 + was significantly reduced in patients with SSc as compared to controls: 0.25±0.22% versus 0.34±0.23% (p=0.02); and 19.0±12.6% vs. 29.0±8.9% (p<0.001), respectively ( Fig. 1G-H).

Correlation of Breg cell levels with clinical severity
Among the 50 SSc patients, there were 16 with dSSc and 34 with lSSc. The median mRSS was significantly higher in patients with dSSc, as expected (data not shown). There was no difference in total B cells, CD21 low , CD24 hi CD38 hi and CD24 hi CD27 + Breg proportions in patients with dSSc and lSSc (data not shown).

Regulatory B cell imbalance and Tfh expansion in SSc / L. Ricard et al.
We then analysed the frequencies of CD21 low B cells and CD24 hi CD38 hi and CD24 hi CD27 + subsets according to SSc severity. The CD24 hi CD27 + Breg cell frequency was significantly decreased in SSc patients with PAH (8.6±4% vs. 20.6±13%; p=0.014), but not in patients with ILD (15±10% vs. 23±14%; p=0.058) (Suppl. Fig. S1). Breg CD24 hi CD27 + frequency was not different in SSc patients with active digital ulcers and did not correlate with the mRSS scale or BNP levels. No difference was observed for CD24 hi CD38 hi and CD21 low B cell frequencies with respect to the presence of ILD, of active digital ulcers, of arterial hypertension, and no correlation was detected with mRRS or BNP levels (data not shown). Futhermore SSc patients with high plasmablast cells (above 1%) had more PAH than those with plasmablast cells above 1% (25% vs. 0%; p=0.03) (data not shown).

Decrease in Breg cells correlates with the expansion of cTfh in SSc patients
In a previous study analysing cTfh in SSc patients, we observed an expansion of these cells, notably in dSSc patients. As Breg cells could be involved in the regulation of Tfh, we evaluated the correlation of cTfh frequency with Breg cell subsets. Interestingly, we observed a negative correlation between cTfh and CD24 hi CD27 + Breg cells in SSc patients but not in healthy controls ( Fig. 2A-B). We did not observe any correlation between cTfh and CD24 hi CD38 hi B cells or between cTfh and CD21 low B cells (Fig. 2C-D).

Discussion
In this study, we report the decrease of Breg cell subpopulations CD24 hi C-D38 hi and CD24 hi CD27+ in patients with SSc in comparison to healthy controls, with a more pronounced decrease of Breg lymphocytes in SSc with PAH impairments. In contrast, the CD21 low B cells were increased in SSc patients as compared to healthy controls. These CD21 low B cells have already been described in different autoimmune diseases such as rheumatoid arthritis (21), Sjögren's syndrome (12) and systemic lupus erythematosus (13,22). This particular B cell population is predominantly composed of memory B cells (23) and expresses high levels of activation markers, inhibitory receptors and a peculiar pattern of homing receptors. These cells are usually considered as having features of anergic and exhausted cells, as characterised by increased apoptosis and decreased proliferation after stimulation (12). These CD21 low B cells highly express autoreactive antibodies and thus can be enriched by autoreactive B cell clones that may have been selected by selfantigens. In our study, we did not detect differences in CD21 low B cells according to clinical severity as previously reported (5). Marrapodi et al. reported that CD21 low B cells in SSc were increased in comparison to healthy controls with a higher prevalence of PAH in those with more than 10% of CD21 low B cells (15,24). Regulatory B cells is a relatively newly recognised subset of B cells which have an immunoregulatory role by suppressing excessive inflammatory responses through the inhibition of T CD4 + Th1 and Th17 cell proliferation and the capacity to express inflammatory cytokines and the induction of regulatory T lymphocytes (Tregs) (25). Several studies have demonstrated decreased Breg cells in various autoimmune diseases, arguing for the potential involvement of these B cell subsets in the regulation of autoimmune diseases. Data about Breg cell impairment in SSc are still scarce but show decreased frequencies of several Breg cell subpopulations in various subtypes of SSc. Thus, Breg cell numbers could be lower in patients with severe and ex-tensive SSc, as found in dSSc, in patients with PAH and ILD impairments (26,27). Few studies have analysed the functional impairment of regulatory B cells in SSc and have mainly showed decreased numbers of IL-10 positive B cells (26,28). The TIM-1+ IL-10 expressing B cells are reduced in SSc patients and have reduced ability to suppress CD4 + T cell production of inflammatory cytokines (28). The results reported by Matsushita et al. showing that Breg cell levels correlated negatively with the titre of anti-topo I antibody (Ab) and anti-centromere Ab in SSc patients and are in line with our observations of a negative correlation between CD24 hi CD27 + Breg cells and cTfh (26). We did not detect a correlation between the CD24 hi CD38+ Breg cell subset and cTfh, suggesting that these two subsets display different functional characteristics as previously reported (29). It remains to be investigated whether these mechanisms are involved in the direct regulation of Tfh, which are likely IL-10 independent because both of these two subsets express IL-10 (29). It is interesting to highlight that BAFF inhibition therapies have been found to promote Breg cell numbers, by selectively depleting B effector cells producing IL-6 while sparing Breg cells (24). In addition, immunosuppressive therapies such as autologous haematopoietic stem-cell transplantation increased regulatory T and CD24 hi CD38 hi B cells at 6-and 12 months at higher levels in responders than in non-responders (30). Moreover, BAFF has been shown to regulate Tfh cells and promote their accumulation (31). Thus, a selective B cell depletion sparing Breg cells in SSc could be a potent therapeutic strategy (24). Taken together, increased numbers of CD21 low B cells in the blood of SSc patients instead of being eliminated, and impaired regulatory B cell function could be part of a favorable environment to break tolerance leading to the development and/or extension of SSc. Further studies are necessary to determine the functional profile of Breg cell subsets and their ability to regulate cTfh and disease severity.

Conclusions
These results suggest that Breg cell subsets may participate in the regulation of cTfh and disease severity. Decreased CD24 hi CD27 + Breg cell frequency may contribute to the development of SSc.

Take home messages
• The Breg cell subpopulations CD24 hi CD38 hi and CD24 hi CD27+ are decreased whereas the CD21 low B cells were increased in SSc patients.
• Regulatory B cells were negatively correlated with cTFH in systemic sclerosis.
• The regulatory B cells decrease is more pronounced in SSc with PAH.
• B reg cells seem to participate to cTFH expansion and to the disease severity in systemic sclerosis.