A total of 45 cases of TPE were enrolled from Shenzhen Third People's Hospital (Shenzhen, China). TPE was diagnosed if i) acid fast bacilli (AFB) staining or M. tuberculosis (MTB) cultures or MTB-DNA polymerase chain reaction of pleural effusion or pleural biopsy specimens showed positive; ii) or if parietal pleural biopsy specimens present typical histopathology characterized with tuberculous granuloma or caseous necrosis (18). A total of 40 cases of HC subjects who had received BCG vaccination at birth and showed a negative tuberculin skin test (TST) were recruited. All subjects were recruited from January 2016 to November 2016 in Shenzhen Third People's Hospital. Subjects with HIV infection, diabetes, cancer and autoimmune diseases were excluded from the study. At the time of sample collection, all of the TPE patients had not received any anti-TB therapy, corticosteroids or other non-steroidal anti-inflammatory drugs. The characteristics of both study cohorts are shown in Table I, there was no significant differences in terms of age range and gender ratio were noted between TPE patients (age range: 18–63 years; male/female: 1.0) and HCs (age range: 20–54; male/femal: 1.2). The study was approved by the Ethics Committee of Guangdong Medical University and Shenzhen Third People's Hospital, and written informed consent was obtained from all study subjects before their participation.

PBMCs and PFMCs were isolated and prepared as previously reported (19). Briefly, approximately 10 ml pleural effusion collected from TPE patients and 5 ml peripheral blood samples from TPE patients and HCs were centrifuged at 2000 × g for 10 min at 4°C. The supernatants were stored at −80°C for future analysis. Cell pellets from pleural effusion were suspended in PBS and cellular components of the blood samples were used for the PBMC isolation by standard Ficoll-Hypaque (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) density gradient centrifugation (2000 × g for 20 min at 4°C). PBMCs and PBFCs were then washed twice with pre-cooled PBS (pH 7.4; 4°C), and then re-suspended in complete RPMI-1640 medium with 20% heated-inactivation fatal calf serum (FBS; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA). A cell viability of >95% was seen in all experiments as determined by try-pan blue exclusion.

For quantitative ELISA assay, BAFF in plasma and supernatants of pleural effusion was tested using the Human BAFF Quantikine ELISA Kit (SBLYS0B; R&D Systems, Inc., Minneapolis, MN, USA) following the manufacturer's protocol.

The freshly isolated PBMCs and PFMCs were washed with PBS (4% FBS) and suspended at a concentration of 1×10⁷/200 µl, followed by staining with CD19-APC, IgD-FITC, CD27-PE-Cy7, CD38-APC-Cy7 (Biolegend, San Diego, CA, USA) for 30 min at 4°C in the dark. Cells were then washed twice and re-suspended in 200 µl PBS (4% FBS). Within 2 h, the samples were acquired on a modified BD Canto II™ flow cytometer (BD Biosciences, San Jose, CA, USA). Data analysis was performed using FlowJo software (Tree Star, Inc., Ashland OR, USA).

Statistical analysis was performed using GraphPad Prism 6 software (GraphPad Software, Inc., La Jolla, CA, USA). Differences in sex ratio of the two study cohorts were compared by Pearson's χ² test. Differences in age between the two study cohorts were evaluated by Student's t-test. Differences in BAFF level and the proportion of each B cell subset were evaluated by analysis of variance with Tukey's post hoc test for multiple comparisons. Correlations between two variables were analyzed by Spearman's analysis. P<0.05 was considered to indicate a statistically significant difference.

BAFF were previously reported to increase in the development of human active pulmonary TB (17). It's uncertain whether BAFF is similarly increased in the patients with TPE. In this study, we detected the level of BAFF in plasma from 40 cases HCs and 45 cases TPE patients using a sandwich ELISA kit. We found that the level of plasma BAFF in TPE patients was 2.8-fold higher than that in HCs (Fig. 1). Concomitantly, we investigated the levels of BAFF in pleural effusion of these TPE patients, and BAFF level was higher in pleural effusion compared to that in plasma (Fig. 1).

Gating strategies were set to evaluate B cell subsets (Fig. 2). Naïve B cells were classified as CD19⁺IgD⁺CD27⁻, while total memory B cells were defined as CD19⁺CD27⁺, including an unswitched IgD⁺ population and a switched IgD⁻ population. Plasma cells were identified as CD19⁺IgD⁻CD38⁺CD27⁺and transitional B cell as CD19⁺IgD^dim CD38⁺. Definitions of B cell subsets are also listed in Table II (18,20).

We analyzed B cell profile to evaluate whether B cell subsets in peripheral blood or pleural effusion were altered among the study groups. In PBMCs, the proportions of naïve B cells, total memory B cell, unswitched B cell, plasma B cell and transitional B cell were all similar between the two study groups (Fig. 3). Compared to PBMCs in TPE patients, the proportions of total memory B cell and unswitched B cell were similar in PFMCs of TPE patients, but the proportions of naïve B cells, plasma B cell and transitional B cell were much lower in PFMCs of TPE patients. It is noteworthy that the proportion of switched B cell was increased in PBMCs of patients with TPE, and higher switched B cell proportion in PFMCs than that in PBMC was also seen in these patients (Fig. 3C). Thus, these results suggest that the B cell compartment were different in the peripheral blood of TPE patients, especially in pleural effusion of TPE. The increased switched B cell may play a major role in acquired immunity against M. tuberculosis.

BAFF is a fundamental survival factor for the maturation and differentiation of B cell (13,14,21). To investigate how increased BAFF affects B cell survival in patients with TPE, we further analyzed the correlation of the BAFF level with the proportions of each B cell subset in PBMCs, PFMCs and both together (Fig. 4). We found that BAFF level had no correlation with naïve B cells, total memory B cell, unswitched B cell and transitional B cell in both PBMCs and PFMCs. Interestingly, BAFF level negatively correlated with plasma B cell when combining PBMCs and PFMCs, despite there was no significant correlation with either proportion alone (Fig. 4E). Moreover, the BAFF level had a high degree of correlation with the proportions of switched B cell both in PBMCs and PFMCs (Fig. 4C). These findings suggest an important role of BAFF in facilitating switched B cell proliferation and redistribution, and potentially inhibiting plasma B cell differentiation as a consequence of M. tuberculosis-induced immune activation in patients with TPE.