Skewed expression prole of receptors for BAFF on peripheral blood B lymphocytes in Graves’ Disease

Previous studies have demonstrated altered serum B lymphocyte activating factor (BAFF) level and expression of BAFF and BAFF receptor (BAFF-R) in thyroid inltrating lymphocytes in Graves’ disease (GD) patients. The aim of the present study was to investigate expression of receptors for BAFF on the peripheral blood B lymphocytes in addition to serum BAFF concentrations. Fifty-two GD patients and twenty-three healthy controls (HC) were recruited in this study. Serum BAFF concentrations were measure by ELISA. Expression of receptors for BAFF on peripheral blood B lymphocytes, including BAFF-R and transmembrane activator and calcium-modulating and cyclophilin ligand interactor (TACI), was analyzed by ow cytometry. Serum BAFF levels, expression of BAFF-R and TACI on peripheral blood B lymphocytes were compared between GD and HC groups. Correlations between serum BAFF concentrations and thyroid function and thyrotropin receptor autoantibody (TRAb) titers were analyzed in GD patients. In 10 patients with Graves’ orbitopathy (GO) receiving steroids therapy, effects of steroids on serum BAFF levels and expression of BAFF-R and TACI were observed.


Introduction
Graves' disease (GD) is an autoimmune disease mediated by thyrotropin receptor speci c autoantibody (TRAb) [1]. TRAb secreted from the autoreactive B lymphocytes binds to thyrotropin receptor and stimulates the over production of thyroid hormone [1]. The persistent existing of the TRAb secreting B lymphocytes lead to the high relapse rate of GD. Targeting these autoreactive B lymphocytes could be a new strategy to treat GD [1].
B lymphocyte activating factor (BAFF, also known as Blys), a member of the tumor necrosis factor (TNF) family, promotes B lymphocytes proliferation, differentiation and survival [2]. BAFF rescue low-a nity auto-reactive transitional B lymphocytes at tolerance checkpoints and promotes their maturation [2,3].
Overexpression of BAFF in mice induced a dramatic expansion of activated autoreactive B lymphocytes and autoantibody production [4]. In humans, over produced BAFF was also linked to autoimmune diseases. Serum BAFF concentrations are elevated in autoimmune diseases including systemic lupus erythematosus (SLE) [5], autoimmune hepatitis [6] and primary Sjögren's syndrome [7].
Accumulated evidences support BAFF is associated with autoimmune thyroid diseases (AITD). BAFF gene polymorphisms are susceptibility variants for the occurrence of GD [8,9]. Patients with either GD or autoimmune thyroiditis showed elevated serum BAFF concentrations [10][11][12]. BAFF also associated with the onset of thyroid autoimmunity in chronic hepatitis C patients receiving interferon alpha therapy [13].
Blocking BAFF function with soluble BAFF receptor fusion protein relief hyperthyroidism in a murine model [14]. Finally, immune suppressive effect of steroids in patients with Graves' orbitopathy (GO) is mediated by blocking BAFF secretion [11].
BAFF exerts its biological effect through interacting with three cell surface molecules: BAFF receptor (BAFF-R, CD268; also known as TNFRSF13C or BR-3) [15], transmembrane activator and calciummodulating and cyclophilin ligand interactor (TACI, CD267; also known as TNFRSF13B) [16] and B cell maturation antigen (BCMA; also known as TNFRSF17) [17]. They have the different expression pro les based on B cell developmental stages [3,18]. BAFF-R is the dominant BAFF receptor expressed on almost all B lymphocytes expressing functional B cell receptor (BCR), including naïve B cells, marginal zone B cells and switched memory B cells. TACI is mainly expressed by CD27 + memory B cells and a small part of plasma cells. Expression of BCMA is restricted on plasmablasts and plasma cells. Campi et al. found thyroid in ltrating lymphocytes (TIL) in AITD expressed more BAFF and BAFF-R than those in multinodular goiter [19]. However, it is not clear whether expression of receptors for BAFF on peripheral blood B lymphocytes are altered in the background of GD.
In this study, to disclose how the interactions between BAFF and its different receptors promote the autoimmunity of GD, we measured serum BAFF levels and the expression of two BAFF receptors, BAFF-R and TACI, on the peripheral blood B lymphocytes in GD patients. Furthermore, effects of steroid on expression of BAFF and its receptors in GO patients were also observed.

Power Calculation and Patients
Power was calculated by R package. We set the Type I Error=0.05 and the Type II Error=0.2. Based on data of our initial assay, at least 14 subjects should be enrolled in each group to achieve the hypothesis being tested and these error rates, so 52 newly diagnosed GD patients and 23 healthy controls (HC) were recruited from Jiangsu Province Hospital of TCM/the A liated Hospital of Nanjing University of Chinese Medicine.

Therapy
All patients received methimazole therapy. Initial methimazole dosage was determined by serum free tetraiodothyronine concentration according to the American thyroid association (ATA) guideline for management for hyperthyroidism [20]. Ten patients having clinical activity score (CAS) above 3/7 at the rst examination were diagnosed with active GO according to the European Group of Graves' Orbitopathy (EUGOGO) classi cation system [21]. These patients underwent methylprednisolone iv injection in addition to administration of methimazole. Methylprednisolone was administered 500mg weekly in the rst 6 weeks and 250 mg weekly in the following 6 weeks as Zhu et al. described [22].

Subjects characteristics
Demographic data of the two groups, thyroid function and TRAb titers of the GD group were listed in Table 1. Although more females were recruited in the GD group, there were no differences of age and gender ratio between the two groups.
2. Serum BAFF levels were elevated in GD patients and positively correlated with free thyroxine concentrations Serum BAFF levels were elevated in autoimmune diseases including GD in serial studies. To con rm these ndings, we measured serum BAFF concentrations in both GD patients and healthy individuals by ELISA. As shown in Figure 1A, serum BAFF levels in patients with GD were signi cantly higher (1.18 ± 0.33 ng/ml) than those in healthy individuals (0.93 ± 0.24 ng/ml, P=0.0011, Figure1A). This was in line with previous studies.
Despite previous reports showed serum BAFF levels were associated with activity of autoimmune disease, there were con icting conclusions on the relation between BAFF expression and autoantibody titers. We thus plotted correlations between BAFF levels with either thyroid function or TRAb titers. Our data showed in GD patients, serum BAFF levels positively correlated with serum free thyroxine concentrations (Spearman r=0.2460, P=0.0394, Figure 1B), but not the TRAb titer (P =0.4734, Figure1C).

Skewed expression pro le of receptors for BAFF on peripheral blood B lymphocytes in Graves' disease
Receptors for BAFF include BAFF-R, TACI and BCMA. Because BCMA is mainly expressed in plasmablasts and plasma cells, which are very few in peripheral blood, we only detected BAFF-R and TACI in this study.
Both two molecules were expressed on peripheral blood B cells. Majority of the peripheral B lymphocytes expressed BAFF-R in our assay. Expression of BAFF-R on peripheral blood B lymphocytes in GD patients was higher comparing with that in healthy controls (mean MFI: 4.52 ± 2.06 in GD vs 3.00 ± 0.87 in HC, P<0.0001, Figure 2A). TACI was not expressed on all CD19 + cells. There was no difference of percentages of B cells expressing TACI between healthy controls and GD patients (data not shown). However, as shown in gure 2B, expression of TACI on peripheral blood B lymphocytes decreased in GD patients (mean MFI: 7.96 ± 4.06 in GD vs 9.10 ± 3.37 in HC, P =0.0331). These data suggested skewed expression of receptors for BAFF on peripheral blood B lymphocytes in GD patients.

Skewed expression pro le of receptors for BAFF was corrected after steroids therapy in GO
Steroids therapy was shown to suppress serum BAFF concentrations in GO patients, its effect on BAFF receptors expression needs explored. We therefore measured serum BAFF concentrations, expression of BAFF-R and TACI on peripheral blood B lymphocytes before and after steroids therapy in 10 GO patients. Mean serum BAFF concentrations decreased from 1.18 ± 0.27 ng/ml to 0.97 ± 0.10 ng/ml after a 12- week iv injection of methylprednisolone (P=0.0364, Figure 3A). Steroids also decreased BAFF-R expression on peripheral blood B lymphocytes in GO patients from 6.26 ± 4.91 to 4.05 ± 1.58 (P=0.0083, Figure 3B). There was a trend of increased TACI expression in B lymphocytes after steroids therapy (TACE MFI: 6.67 ± 1.96 before steroids therapy vs 7.05 ± 2.32 after steroids therapy), however, the difference was not statistical signi cance (P=0.1974, Figure 3C). These results indicated immunol regulation effect of steroids on B lymphocytes is not only on the B cell surviving ligand BAFF, but also on its receptors, mainly BAFF-receptor.

Discussion
More than half of newly generated B lymphocytes have some degree of autoreactivity [23]. These immature autoreactive B cells are extreme sensitive to apoptosis because they express high levels of death receptor Fas, a mechanism to eliminate autoreactive B cells [24]. However, the over expressed BAFF and its receptors may subvert the destiny of apoptosis of these autoreactive B lymphocytes by inducing the expression of anti-apoptotic factors and by blocking the function of pro-apoptotic molecule [25]. The rescued autoreactive B lymphocytes could mediate autoimmune disease through either the secreted autoantibodies or antigen presenting [25]. Altered serum BAFF levels and intra-thyroid BAFF and its receptors expression levels have been observed in autoimmune thyroid disease [10,12,19]. Different to these researches, we measured BAFF receptors BAFF-R and TACI expression on peripheral blood B lymphocytes in addition to serum BAFF concentrations in this study.
We found elevated serum levels of BAFF in GD patients, suggesting a possible role of BAFF in thyroid autoimmunity. This is in line with other studies describing increased serum BAFF levels in thyroid autoimmune diseases including GD and Hashimoto's thyroiditis (HT) [10,12]. Our data showed serum BAFF concentrations correlated positively with free thyroxine concentrations, con rming serum BAFF levels positively correlating with activity of autoimmune diseases. There was no correlation between serum BAFF concentrations and TRAb titers based on our data. The correlation between serum BAFF levels and titers of autoantibodies are controversial as different results have been reported [7,12,26]. In our assay, TRAb was measured by a competing electrochemiluminescence immunoassay method [27]. TRAb measured by this method was only that having the same epitope with monoclonal antibody M22, but not the whole TSH receptor autoantibodies in the blood. BAFF levels may correlate with the whole autoantibodies titers, but not the titer of autoantibody with the speci c epitope, this could be the reason we did not observe correlation between BAFF levels with TRAb titers in our study.
The most signi cant nding of our work was the skewed expression of receptors for BAFF on peripheral blood B lymphocytes in GD patients: the BAFF-R expression was increased while the BAFF receptor TACI expression was decreased. Our results were not in line with reports of reduced BAFF-R expression on peripheral blood B cells in other autoimmune disease as SLE [15]. Con icting results of TACI expression on peripheral B lymphocytes in SLE patients have been reported [5,15]. However, B lymphocytes in ltrating in thyroid expressed more BAFF-R, these data combined with our results [19], indicating different expression pro les of receptors for BAFF among autoimmune diseases.
The skewed expression of BAFF-R and TACI in GD may re ect the different functions of these two molecules [2,3]. BAFF-R and TACI have different binding a nity to BAFF [3]. BAFF-R and TACI oppositely regulates B cell homeostasis [2,16,28,29]. BAFF-R promotes B cells survival while TACI sensitizes B cells to apoptosis. Knocking out TACI led to elevated B cells number and SLE-like symptoms in the presence of over expressed BAFF in mice [28]. TACI activation also mediates immunosuppression through inducing interleukin (IL)-10 production in B lymphocytes [30,31]. IL-10 is an immunosuppression cytokine. IL-10 secreting B cells are named Bregs (regulatory B cells) or B10 cells. Overexpressed TACI promoted more IL-10 expression in B lymphocytes in SLE patients [30]. Loss of Bregs has been described in GD patients [32]. We may speculate in GD patients, in the context of high levels of BAFF and skewed expression of BAFF-R and TACI on B lymphocytes, BAFF-R pathway activity was enhanced while TACI pathway activity was relatively diminished, which leads to the accumulation of autoreactive B cells and loss of Bregs. The skewed expression of BAFF-R and TACI is the key mechanism leading to autoimmunity in GD. Totally blocking both BAFF-R and TACI signal pathway will only augment in ammation due to the inhibited IL-10 secretion, as observed in multiple sclerosis (MS) patients [33]. To selective inhibit the BAFF-R pathway and to restore the TACI pathway function could be a novel strategy to cue GD. This strategy may eliminate the autoreactive B lymphocytes, thus with the potential to prevent relapse of GD.
Our data indicated steroids targets BAFF and BAFF-R, but not TACI. There were no previous studies reporting effects of steroids on BAFF-R expression in autoimmune diseases in the literature to our knowledge. In large B cell lymphoma, only BAFF-R, rather than BAFF, was an independent prognostic factor for steroids treatment [34], highlighting BAFF-R is an important target for steroids in eliminating B lymphocytes. Due to the small sample size of patients receiving steroids, the effect of steroids on TACI expression need be studied in the future with larger sample size.

Conclusions
In conclusions, skewed expression of receptors for BAFF were associated with autoimmunity of GD. Restoring the expression pro le of receptors for BAFF could be the new strategy to treat and cue GD.    Decreased TACI expression on peripheral B lymphocytes in GD patients. Mean MFI of TACI on peripheral B lymphocytes in GD patients was 7.96 ± 4.06, lower than that in HC (9.10 ± 3.37), P =0.0331. MFI: mean uorescence intensity. Bars: mean MFI. Trend of increased TACI expression on the peripheral blood B lymphocytes after steroids therapy in GO patients. Mean MFI of TACI on peripheral blood B lymphocytes was elevated from 6.67 ± 1.96 before steroids therapy to 7.05 ± 2.32 after steroids therapy, without statistically signi cant (P=0.1974).