Introduction

Systemic lupus erythematosus (SLE) is a prototypic systemic autoimmune disease characterized by various immunological abnormalities, including dysregulated activation of both T and B lymphocytes with overt production of autoreactive antibodies [1, 2]. Aberrant production and imbalance of T helper (Th) cell cytokines have been implicated in the pathogenesis of autoimmunity [3]. However, the cytokines and chemokines for the activation and migration of B cells have not been well investigated. CXCL13/B lymphocyte chemoattractant (BLC) is a small cytokine belonging to the CXC chemokine family that is produced by cells in the omentum, peritoneal macrophages, and dendritic cells [4, 5]. CXCL13 is selectively chemotactic for B cells including both the B1 and B2 subsets by interacting with specific chemokine receptor CXCR5 [4, 6]. CXCL13 and CXCR5 together control the organization of B cells within follicles of lymphoid tissues [7] and is highly expressed in the liver, spleen, lymph nodes, and gut of humans [6]. Previous studies have revealed that CXCL13 can induce the trafficking of distinct CXCR5+ T cells designated as follicular helper T cells (TFH) which are specifically involved in high-affinity IgG production in germinal centers developed within B cell follicles of secondary lymphoid tissues including lymph nodes, spleen, and tonsils [8–11]. CD4+TFH cells, located at B cell follicles, provide a T helper function to B cells and represents one of the most numerous and important subsets of effector T cells in lymphoid tissue [12, 13]. Zinc finger transcription factor BCL-6 may play certain role for driving naïve Th cells to the TFH cell lineage [12]. IL-21 is a pleiotropic cytokine produced by CD4+Th cells that modulate the differentiation and function of T, B, natural killer (NK), and dendritic cells by binding to the receptor composing of the IL-21 receptor-α (IL-21Rα) and the common γ chain [14, 15]. Recent study has intimated that IL-21 can mediate the differentiation and generation of TFH cells [10, 15]. Nevertheless, autocrine production of IL-21 from TFH cells can potently stimulate the differentiation of B cells into antibody-forming cells through IL-21R [12]. As a result, dysregulation of TFH cell function may relate to the pathogenesis of SLE. IL-21 has been shown to contribute to the development of autoimmune diseases in different animal models of SLE, experimental autoimmune encephalomyelitis, and rheumatoid arthritis [14]. Although genetic association of IL-21 polymorphisms has been demonstrated in SLE [16], the detailed mechanisms of the subsequent B cells activation in SLE patients requires further characterization.

B1 cells have different origin and function from the major recirculating conventional B (B2) cells by their unique tissue distribution, cell surface phenotype, and function [17]. Cell surface expression of CXCR5 on CD5+B1 cells was significantly higher than B2 cells. The accumulation of B1 cells in the peritoneal cavity and spleen are responsible for the body cavity immunity and the production of autoantibody for the development of autoimmune disease in the murine model [4, 17, 18]. Elevated levels of B1 cells have been documented in patients with autoimmune disorders such as Sjogren’s syndrome and rheumatoid arthritis [19, 20]. Previous studies using murine model of SLE showed that CXCL13 is highly produced by CD11b+CD11c+ dendritic cells (DC) in the target organs including thymus and kidney for the chemoattraction of B1 cells into target organ [18, 21–23]. Therefore, the elevated expression of CXCL13 by myeloid DC in the target organs may play a crucial role in breaking the immune tolerance in the thymus leading to the activation of self-reactive CD4+Th cells and the recruitment of autoantibody producing B cells in the development of murine lupus [18, 22, 23]. Tumor necrosis factor (TNF)-α expression is increased in the SLE murine model and plays crucial role in the recruitment and maturation of DC into CXCL13 producing DC during disease development [23]. Apart from CXCL13, B cell activation factor of the TNF family (BAFF) is a TNF family member that binds to B cells and potentiates B cell receptor-mediated proliferation [24]. BAFF was reported to increase the in vitro chemotactic response of primary human B cells to CXCL13 [25]. BAFF actually processes more potent effect on memory B cells to CXCL13 than that of naive B cells [25]. Transgenic mice over-expressing the BAFF exhibited impaired B cell tolerance and altered T cell differentiation for the development of an autoimmune disorder similar to SLE [26]. In order to confirm the immunopathological roles of CXCL13 and characterize the TFH dysregulation in SLE, the plasma concentrations of CXCL13, BAFF, TFH-cells-related cytokine IL-21, and cell surface expression of TFH-related receptor CXCR5 and IL-21R on CD4+Th and CD19+ B cells were investigated in the present study.

Materials and Methods

SLE Patients, Control Subjects, and Blood Samples

Thirty-five Chinese SLE patients (31 females and four males) were recruited at the Rheumatology Out-patient Clinic of the Prince of Wales Hospital, Hong Kong. Diagnosis of SLE was established according to the 1982 revised American Rheumatism Association criteria [27], and disease activity was evaluated by the SLE disease activity index (SLEDAI) score [28]. Twenty-three sex- and age-matched healthy Chinese volunteers were recruited as controls (NC group). Twelve milliliters of EDTA venous peripheral blood were collected from each patient and control subject. The above protocol was approved by the Clinical Research Ethics Committee of The Chinese University of Hong Kong New Territories East Cluster Hospitals, and informed consent was obtained from all participants according to the Declaration of Helsinki.

Assay of Human CXCL13, BAFF, and IL-21

Concentrations of CXCL13 and BAFF, and IL-21 in plasma were measured by enzyme-linked immunosorbent assay (ELISA) using Quantikine colorimetric reagent kit (R&D Systems, MN, USA) and Bender MedSystems GmbH, Vienna, Austria, respectively.

Quantitative PCR of BCL-6 Gene Expression

Total RNA of peripheral blood mononuclear cells (PBMC) was extracted by RNeasy Mini Kit (Qiagen Inc., Canada), following manufacturer’s instruction. All specimens were pre-treated with deoxyribonuclease I (Invitrogen Corp., CA, USA) and then stored at −70°C. For each reaction, approximately 0.5 μg of total RNA was reversely transcribed to complementary DNA (cDNA) with TaqMan Reverse Transcription Reagents (Applied Biosystems Inc., CA, USA). The mRNA expression of BCL-6 and β-actin (endogenous control) in SLE patients’ PBMC was quantified by real-time polymerase chain reaction (PCR) with the use of Applied Biosystems 48-well StepOne™ Real Time PCR System. The primers and probes for human BCL-6 and β-actin were purchased from Applied Biosystems for TaqMan gene expression assay. Real-time PCR was performed in a 20-μL reaction mixture containing TaqMan Fast Universal PCR Master Mix (Applied Biosystems), TaqMan β-actin/BCL-6 Detection Reagent and cDNA sample. The real-time PCR reaction was performed at fast mode: 95°C for 1 s to denature cDNA and 60°C for 20 s to allow the TaqMan probe and primers to anneal to the denatured cDNA. The cycles were repeated 30 times after an initial 20 s denaturation at 95°C. A relative BCL-6 mRNA expression was obtained by comparing the relative expression of BCL-6 and β-actin.

Multiparametric Flow Cytometric Analysis of CXCR5 and IL-21R on CD4+Th and CD19+B Cells

Peripheral blood mononuclear cells were prepared by the centrifugation of blood using a density gradient (Ficoll-Paque Plus; GE Healthcare Corp., NJ, USA). Cells were incubated with fluorescein isothiocyanate (FITC)-conjugated mouse anti-human CXCR5 monoclonal antibody (mAb; R&D Systems) or unconjugated mouse anti-human IL-21R mAb (R&D Systems) followed with secondary FITC-conjugated goat anti-mouse immunoglobulin (Ig) antibody (Invitrogen). The immunophenotyping of CD4+Th cells and CD19+B cells was labeled with phycoerythrin (PE)-conjugated anti-CD4 and allophycocyanin (APC)-conjugated anti-CD19 antibody (Becton Dickinson Pharmingen Corp., CA, USA). PE, FITC, or APC-conjugated mouse IgG (BD Pharmingen) was used as isotypic control accordingly. Cell surface expression of CXCR5 and IL-21R on CD4+Th and CD19+B lymphocytes was analyzed by flow cytometry as presented in mean fluorescence intensity (MFI) on 5,000 viable cells (BD FACSCalibur).

Statistical Analysis

Numerical data were expressed as median (interquartile range, IQR) if they were not in Gaussian distribution. Difference in plasma concentration among groups was compared with Mann–Whitney rank sum test. Spearman’s rank correlation test was used to assess the correlation of plasma chemokines and cytokine with SLEDAI score. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) statistical software for Windows, Version 16.0 (SPSS Inc, IL, USA). Probability values (p) < 0.05 were considered as significant.

Results

SLE Patients and Control Subjects

The age, sex, SLEDAI score, duration of diagnosis, and drug treatment of the study groups are summarized in Table 1.

Table 1 Characteristics of SLE Patients and Control Subjects
Full size table

Elevated Plasma Concentrations of CXCL13, BAFF, and IL-21 in SLE Patients

As shown in Fig. 1a and b, plasma CXCL13 [median (interquartile range), SLE, 95.3 (69.9–159.4) vs control, 45.3 (38.3–63.1) pg/mL, p < 0.0001] and BAFF [SLE, 1,627 (1,248–1,981) vs control, 923 (809.5–1,087) pg/mL, p < 0.0001] concentration was significantly higher in SLE patients than in control subjects. Similarly, plasma TFH cytokine IL-21 concentration was also significantly higher in SLE patients than in control subjects [SLE, 445.4 (411.8–503.1) vs control, 398.0 (382.2–432.2) pg/mL; p < 0.001] (Fig. 1c). Table 2 shows that there was a significant and positive correlation between CXCL13 concentration and SLEDAI score in SLE patients (r = 0.399, p = 0.032). However, there were no significant correlations between plasma BAFF or IL-21 concentrations with SLEDAI score in SLE patients (Table 2, both p > 0.05).

Fig. 1
figure 1

Plasma CXCL13, BAFF, and IL-21 concentrations of SLE and normal control subjects. The Mann–Whitney rank sum test was used to assess the differences of concentration of a CXCL13, b BAFF, and c IL-21 between the SLE and NC groups. Results are presented as box-and-whisker plots showing the median (IQR). Whitney rank sum test was used to assess the differences of concentration between patients and control subjects. ***p < 0.001

Full size image
Table 2 Correlation Between Plasma Concentrations of CXCL13, BAFF, and IL-21 and SLEDAI Score
Full size table

No Difference of mRNA Expression of BCL-6 in PBMC

Since BCL-6 is the transcription factor for TFH cells development [12], we used TaqMan quantitative PCR method for the assay of the mRNA expression of BCL-6 in PBMC from SLE patients and control subjects. As shown in Fig. 2, there was no significant difference of BCL-6 gene expression between SLE subjects and control subjects (p > 0.05).

Fig. 2
figure 2

BCL-6 expression in SLE patients and control subjects. Total RNA was extracted from PBMC and subjected for the quantitative PCR assay of the relative mRNA expression of BCL-6. Results are presented as box-and-whisker plot showing the median (IQR). Whitney rank sum test was used to assess the differences of concentration between patients and control subjects

Full size image

Down-Regulated Cell Surface Expression of CXCR5 and IL-21R on Th Cells and B Cells

Since we found that B cell chemokine CXCL13 and TFH cytokine IL-21 was significantly higher in SLE than controls, it prompts us to evaluate the expression of receptors for CXCL13 and IL-21 on their target cells. As shown in Fig. 3, the cell surface expression of CXCR5 on Th cells and B cells and IL-21R on B cells was significantly down-regulated in SLE patients while comparing with controls [median (interquartile range), CXCR5 on Th cells, SLE, 0.01 (0.00–0.09) vs controls, 0.1 (0.04–0.20); CXCR5 on B cells, SLE, 4.5 (2.48–6.61) vs controls, 11.2 (9.25–14.02); IL-21R on B cells, SLE, 3.6 (0.98–11.71) vs controls, 17.8 (4.33–38.44), all p < 0.01]. However, the expression of IL-21R on Th cells did not have any significant difference between SLE and control subjects [SLE, 0.29 (0.09–0.57) vs controls, 0.03 (0.0–0.5), p > 0.05].

Fig. 3
figure 3

Expression of CXCR5 and IL-21R on CD4+Th cells and B cells in SLE patients and control subjects. The expression of CXCR5 and IL-21R are shown as MFI subtracting corresponding isotypic controls with box-and-whisker plot showing the median (IQR). Mann–Whitney rank sum test was used to assess the differences of concentration between patients and control subjects. **p < 0.01, ***p < 0.001

Full size image

Discussion

SLE has been postulated to be an autoantibody, immune complex, and Th cytokine-mediated disease [29]. Previous studies demonstrated that B cell chemokine CXCL13 is ectopically and highly expressed in thymus and kidney in murine model for SLE [23]. In the present study, we have confirmed the significant increase in plasma concentration of CXCL13 in SLE patients and the elevated plasma concentration of CXCL13 correlated significantly with SLE disease severity (Fig. 1 and Table 2). The results are in concordance with the recent publication of the elevation of serum CXCL13 levels in SLE patients with different ethical group [30]. Apart from SLE, previous study has demonstrated the elevation of CXCL13 in active rheumatoid arthritis (RA) patients comparing with quiescent disease and healthy controls, correlating with Disease Activity Score in 28 joints [31]. Anti-TNF treatment was found to reduce the plasma level of CXCL13 in RA patients [31]. Therefore, CXCL13 can act as a disease activity marker for both RA and SLE patients. Production of CXCL13 is not higher in other immunological disorders such as allergic asthma and psoriasis; this indicates the specific role of the elevated CXCL13 concentration in autoimmune disease [32].

Our present study also confirmed that the expression of BAFF increases in SLE, thereby suggesting the pathogenic role of BAFF in SLE by stimulating T-cell-dependent B cell production of autoantibodies [33, 34]. BAFF is actually a fundamental survival factor for transitional and mature B cells and its overexpression leads to an expanded B cell compartment and autoimmunity in mice, and elevated serum concentration of BAFF can be found in autoimmune patients with lupus and rheumatoid arthritis [34]. Since BAFF can increase the chemotactic response of primary and memory human B cells to CXCL13 [25], the synergy between the elevated CXCL13 and BAFF produced by stromal cells and follicular dendritic cells can have important implications for B cell activation in SLE [25]. Elevated plasma IL-21 in SLE is produced by TFH cells for the germinal centers formation [10]. The central role of IL-21 in germinal centers formation actually related to its effects on TFH cell generation [10]. Recent animal study has revealed that elevated production of IL-21, TFH dysfunction within germinal centers and aberrant positive selection of germinal center B cells are required for the production of autoantibodies and systemic autoimmunity [35, 36]. Therefore, our results of the elevated CXCL13, BAFF, and IL-21 may relate with the immunopathogenesis mediated by the function of TFH cells in autoimmune disease. As also observed by us, the plasma CXCL13, BAFF, and IL-21 concentrations in SLE patients did not show any correlation with the dosages of prednisolone, hydroxychloroquine, and azathioprine. Although we could not find the correlation between the plasma CXCL13, BAFF, and IL-21 concentrations and serum autoantibody titers, the concentration of autoantibody could be highly produced at lymphoid tissues in SLE patients with elevated plasma CXCL13 concentration [12, 13]. Nevertheless, Schiffer et al. have recently reported that serum CXCL13 concentration correlated with dsDNA titer in SLE patients with active disease [30].

Since our present study cohorts only recruited SLE patients with relatively inactive disease (SLEDAI mean ± SD, 3.1 ± 3) without end-organ manifestation, the correlation of CXCL13 levels with end-organ disease manifestations such as renal or central nervous system (CNS) lupus were not studied. However, previous animal studies have indicated that CXCL13 is increased in aged kidney of SLE mouse model [22]. Moreover, aberrant CXCL13 can induce the formation of ectopic lymphoid tissues in non-lymphoid organs and related to the accumulation of inflammatory cells in kidneys in SLE [22, 37]. Recent clinical study also confirmed that CXCL13 concentrations were higher in SLE patients with nephritis than SLE patients without nephritis [30]. Although CXCL13 plays a role in experimental autoimmune encephalomyelitis of animal model of multiple sclerosis [38], the correlation of CXCL13 level and CNS lupus activity in patient study has not been reported.

Tfh cells retain intense expression of CXCR5, which directs these cells toward CXCL13 localized areas within germinal centers [9]. We observed that the cell surface expression of TFH-related CXCR5 and IL-21R was significantly down-regulated on Th and B cells in SLE patients, it may indicate that most differentiated TFH cells can migrate out from the circulation and localize in the lymphoid organ upon activation during the disease development of SLE. Elevated lymphoid tissue homing chemokine CXCL13 and B cell survival factor BAFF actually synergistically enhance the chemotactic response of B cells and CD4+CXCR5+ T cells to the lymphoid follicles and CXCL13 can be produced by follicular dendritic cells at germinal center [39, 40]. Activated TFH cells upon migration to the B cell follicles can interact with antigen-primed B cells that have specifically migrated outward. This physical interaction then facilitates CD40-CD40L-dependent B cell activation [12]. As a result, it may account for the decreased circulating number of CXCR5+IL-21R+TFH cells in SLE in our study. We could not detect any significant difference of transcription factor BCL-6 gene expression in PBMC between SLE and control subjects. Although BCL-6 is preferentially expressed by TFH cells but not Th1 or Th2 cells for the development of TFH cells [12], only 10–15% of CD4+T cells in germinal center at follicles appear to express BCL-6 protein [41]. This may account for the non-significant difference of BCL-6 expression in PBMC between SLE and control subjects.

In view of the recent development and clinical trials of monoclonal antibody such as belimumab against BAFF that effectively deplete pathogenic B cells and target pathways essential for B-cell-dependent autoantibody-mediated autoimmune response, B-cell-directed therapies represent promising treatments for autoimmune disorders [42, 43]. Using murine model, neutralization of IL-21 by IL-21R-Fc fusion protein treatment can minimally attenuate SLE disease progression [44], and anti-CXCL13 treatment inhibit the development of collagen-induced arthritis and reduce follicular response in both lymphoid and non-lymphoid tissues [45]. In conjunction with above studies, our results of the pathological role of elevated production of TFH-related CXCL13, BAFF, and IL-21 in SLE therefore provides further evidences that the human pathological TFH and B cell responses can be suppressed by targeted approaches including the anti-B cell activating factor (BAFF and CXCL13) or anti-IL-21 for the treatment of SLE and other autoimmune diseases. Moreover, circulating concentration of CXCL13, the signature molecule of TFH cells, may act as a surrogate disease marker of SLE.