CD4+CD25highforkhead box protein 3+ regulatory T lymphocytes suppress interferon-γ and CD107 expression in CD4+ and CD8+ T cells from tuberculous pleural effusions



Tuberculous pleural effusion is characterized by a T helper type 1 (Th1) profile, but an excessive Th1 response may also cause tissue damage that might be controlled by regulatory mechanisms. In the current study we investigated the role of regulatory T cells (Treg) in the modulation of Th1 responses in patients with tuberculous (TB) pleurisy. Using flow cytometry we evaluated the proportion of Treg (CD4+CD25highforkhead box protein 3+), interferon (IFN)-γ and interleukin (IL)-10 expression and CD107 degranulation in peripheral blood (PB) and pleural fluid (PF) from patients with TB pleurisy. We demonstrated that the proportion of CD4+CD25+, CD4+CD25highFoxP3+ and CD8+CD25+ cells were increased in PF compared to PB samples. Mycobacterium tuberculosis stimulation increased the proportion of CD4+CD25low/negIL-10+ in PB and CD4+CD25low/negIFN-γ+ in PF; meanwhile, CD25high mainly expressed IL-10 in both compartments. A high proportion of CD4+CD107+ and CD8+CD107+ cells was observed in PF. Treg depletion enhanced the in-vitro M. tuberculosis-induced IFN-γ and CD4+ and CD8+ degranulation responses and decreased CD4+IL-10+ cells in PF. Our results demonstrated that in TB pleurisy Treg cells effectively inhibit not only IFN-γ expression but also the ability of CD4+ and CD8+ cells to degranulate in response to M. tuberculosis.


Among the many clinical manifestations of tuberculosis (TB), pleuritis is of particular interest as it may be resolved without therapy, and patients are known to undergo a relatively effective immune response against Mycobacterium tuberculosis. However, half of patients recidivate during the next 5 years in the absence of chemotherapy [1]. TB pleural effusion is caused by a severe delayed-type hypersensitivity reaction in response to the rupture of a subpleural focus of M. tuberculosis infection, and although it often occurs in conjunction with pulmonary infiltrates typical of post-primary tuberculosis, it may also manifest as a primary feature of the disease [2]. TB pleurisy results in an increased pleural vascular permeability leading to the accumulation of protein-enriched fluid and recruitment of specific inflammatory leucocytes into the pleural space leading to the clearance of mycobacteria from this cavity. Lymphocytes are the predominant cell type, and in-vivo and in-vitro studies have demonstrated that M. tuberculosis-specific T cells are recruited and expand clonally at the site of infection [3, 4]. An increase of CD4+ T cells belonging to the T helper type 1 (Th1) subset with selective concentrations of interferon (IFN)-γ, tumour necrosis factor (TNF)-α and interleukin (IL-12) [5-9], as well as IFN-γ-inducible chemokines [10], have been detected in pleural fluid (PF).

Although Th1 responses are necessary for the effective control of TB infection they may also cause tissue damage. In line with this, IFN-γ may mediate the clinical manifestation of TB pleuritis, such as fever, exudative pleural effusion and tissue necrosis [11]. For this reason, it has been proposed that regulatory T cells (Treg) and Th2 responses might play important regulatory functions in protecting the host from collateral tissue damage by regulating excessive Th1 responses.

High levels of CD4+CD25highforkhead box protein 3 (FoxP3+) Treg cells have been detected in peripheral blood mononuclear cells (PBMC) from TB patients [12-15], and an in-vivo enrichment of Treg cells has been observed in tuberculous PF [12, 16]. In-vitro depletion of circulating and local site-derived CD4+CD25high T lymphocytes suggest a role of Treg in the suppression of systemic and local immune responses to M. tuberculosis antigens [12, 16]. Also, it has been demonstrated recently that the alveolar lung compartment is enriched in CD4+CD25+FoxP3+ Treg cells that suppress the ability of alveolar and monocyte-derived macrophages to restrict the growth of M. tuberculosis, suggesting that Treg cells also subvert anti-mycobacterial immunity in human TB [17].

It is widely accepted that control of human TB depends on fully activated CD8+ T cells that not only release Th1 cytokines such as IFN-γ and TNF-α, but also lyse M. tuberculosis-infected macrophages [18, 19] and reduce M. tuberculosis viability [20]. In a previous work, we demonstrated an in-vitro M. tuberculosis-induced expansion of Treg cells in PBMC from active TB patients that efficiently suppress antigen-induced IFN-γ expression and the ability of CD8+ cytotoxic T lymphocytes (CTL) to lyse M. tuberculosis-pulsed macrophages. Moreover, Treg depletion enhanced surface CD107a expression on CTL [21]. CD107a and CD107b are intracellular proteins normally found in lysosomes that are expressed transiently on CTL surfaces upon exocytosis of cytotoxic granules during target cell lysis [22, 23]. Furthermore, CD107a-expressing CD8+ T cells are shown to mediate cytolytic activity in an antigen-specific manner [23], demonstrating that CD107a expression on the cell surface is a marker of cytotoxic CD8+ T cell degranulation/activation. In the current study we investigated the role of Treg cells in the modulation of IFN-γ and CD107 expression in human tuberculous pleurisy. Our results demonstrated that upon encountering M. tuberculosis, Treg are functionally active in pleural effusions by modulating CD4+ and CD8+ T cell-mediated IFN-γ and CTL responses.

Materials and methods

Ethical approval

Ethical approval for the study was obtained from Ethic Committee of Hospital F. J. Muñiz, according to the principles laid down in the Declaration of Helsinki recommendations guiding physicians in biomedical research involving human subjects. The study was conducted in accordance with the standards for clinical research of the Universidad Nacional de Buenos Aires, Argentina.


Patients with newly diagnosed moderate and large pleural effusions were identified at the Hospital F. J. Muñiz, Buenos Aires, Argentina. Informed, written consent was obtained from all patients and healthy individuals and the study was approved by the Hospital F. J. Muñiz Ethic Committee. Patients were evaluated by a history and physical examination, with routine investigations including testing for HIV infection, chest radiography, microbiological sputum examination, whenever possible, and aspiration of pleural fluid for biochemical, cytological and detailed microbiological (Ziehl–Nielsen stain and culture for bacterial pathogens) evaluation. Exclusion criteria included HIV positivity or the presence of concurrent infectious diseases. Pleural effusions and blood samples were obtained from patients during thoracentesis before the initiation of chemotherapy. Diagnosis of pleural effusion from TB aetiology was based on demonstration of M. tuberculosis on pleural fluid smear (by the Ziehl–Neelsen stain) and growing M. tuberculosis in pleural fluid specimens (by Lowenstein–Jensen-positive culture) and/or histopathology corroboration. A total of 16 patients were included; among them, six also had pulmonary disease. Table 1 summarizes selected clinical and laboratory data of TB patients. Blood samples were also obtained from 10 purified protein derivative (PPD+) healthy volunteers (four males and six females, aged between 26 and 60 years).

Table 1. Clinical, cytological and laboratory profile from patients with tuberculous pleurisy
Parameter PBMCPFMC
  1. AFB: acid-fast bacillus in sputum; PPD: purified protein derivative skin test; ADA: adenosine deaminase; PBMC: peripheral blood mononuclear cells; PFMC: pleural fluid mononuclear cells.
Age25 (22·5–30·75)  
Gender10 male/6 female  
Pulmonary disease6/16  
ADA  96·6 (71·0–106·0)
Cell counts, cells/mm3 7400 (6100–8500)1600 (800–2500)
Lymphocyte, % 19 (16·5–21·5)77 (67·5–85)
Lymphocyte counts, cells/mm3 1445 (1075–1699)1190 (668–1875)

Thoracocentesis and pleural biopsy

PF was collected by therapeutic thoracocentesis, as described previously [24]. Specimens were subjected to routine biochemical analysis, including tests for total protein, glucose, lactate dehydrogenase and differential cell counts. Bacterial cultures and cytological examinations were performed on all PF in the central laboratory at Hospital F. J. Muñiz. A second sample was dispensed into 50-ml polystyrene tubes containing heparin to obtain mononuclear cells. Peripheral blood samples were also collected on the same day of thoracocentesis.

Mononuclear cells

Peripheral blood and pleural fluid mononuclear cells were isolated from heparinized blood and PF (PBMC and PFMC, respectively) by Ficoll-Hypaque gradient centrifugation and suspended in RPMI-1640 tissue culture medium (HyClone®; Thermo Scientific, Rockford, IL, USA) containing 100 U/ml penicillin, 100 μg/ml streptomycin and 10% heat-inactivated fetal calf serum (Invitrogen Gibco®, Carlsbad, CA, USA) (complete medium).


In order to deplete lymphocytes with high CD25 expression, 1 × 107 PBMC or PFMC were incubated with a limiting concentration of anti-CD25 antibody (0·1 μg/107 cells; eBioscience, San Diego, CA, USA) for 30 min at 4°C, washed with phosphate-buffered saline (PBS) and mixed with goat anti-mouse immunoglobulin (Ig)G-coated magnetic beads (Invitrogen Dynal, Oslo, Norway) by gentle rolling at 4°C for 30 min. Non-rosetted cells (CD25-depleted PBMC) were separated using a magnet. Generally, one cycle of treatment was sufficient for an effective depletion as assessed by flow cytometry. Approximately 90 ± 3% of CD4+CD25high T cells were eliminated from PB and PF samples after depletion, while CD4+CD25low and CD8+CD25+ T cells were slightly reduced (11 ± 4% and 10 ± 5%, respectively). To compare cytokine production and CD107 expression, CD25-depleted PBMC or PFMC were suspended in complete medium ensuring that the number of cells/ml of each subset was the same as in total cultured PBMC.


The gamma-irradiated M. tuberculosis H37Rv strain used in this study was provided by J. Belisle (Colorado University, Co, USA). Mycobacteria were suspended in PBS free of pyrogen, sonicated and adjusted at a concentration of 1 × 108 bacteria/ml [optical density (OD)600 nm = 1].

PBMC and PFMC cultures

In-vitro antigen-specific effector T cells were evaluated by culturing total or CD25-depleted PBMC or PFMC (2 × 106 cells/ml in complete medium) for 18 h in polystyrene tubes (BD FalconTM, San Jose, CA, USA) at 37°C in 5% CO2 atmosphere with or without M. tuberculosis in a 2:1 M. tuberculosis to PFMC/PBMC ratio. Cells were then tested for cytokine, CD25, FoxP3 and CD107 expression.

Immunofluorescence analysis

Surface membrane expression

The following monoclonal antibodies (mAbs) and the corresponding isotypes were used to evaluate surface marker expression in fresh and 18-h cultured PBMC or PFMC: phycoerythrin-cyanin 5 (PE-Cy5) or fluorescence-activated cell sorter (FITC)-conjugated anti-CD4 and anti-CD8, FITC-anti-CD127 (eBioscience) and PE-anti-CD25 (BD Bioscience).

Intracellular expression of FoxP3 transcription factor

FoxP3 expression was detected using the FITC anti-human FoxP3 staining set (eBioscience), according to the manufacturer's instructions. An isotype-matched antibody was used as control (eBioscience).

CD107 surface expression

The frequency of CD4+ and CD8+CD107+ cells was evaluated in total or CD25-depleted PBMC and PFMC cultured for 18 h with or without M. tuberculosis, with the addition of FITC-anti-CD107 mAb (BD Bioscience) for the last 4 h. Cells were then stained with PE-Cy5 or PE-anti-CD4 or anti-CD8 mAb (BD Bioscience).

Intracellular cytokine expression

Intracellular IL-10 and IFN-γ expression was determined in 18-h cultures from PBMC and PFMC with the addition of brefeldin A (5 μg/ml; Sigma Chemical Co., St Louis, MO, USA) for the last 4 h; thereafter cells were stained with anti-CD4, anti-CD8 and anti-CD25, and fixed with 0·5% paraformaldehyde, permeabilized with FACSTM permeabilizing solution 2 (BD Bioscience) before FITC- or PE-labelled anti-IL-10 (BD Bioscience), and anti-IFN-γ (Invitrogen, Camarillo, CA, USA) or the corresponding isotypes were added.

Surface or intracellular stained cells were analysed by flow cytometry; 20 000 events were acquired on a FACSCan cytometer (BD Bioscience) using CellQuest. FCS express software (De Novo Software, Los Angeles, CA, USA) was used for the analysis. Lymphocyte gates were set according to forward- and side-scatter parameters, excluding cell debris and apoptotic cells. Results were expressed as percentages of positive cells in a lymphocyte population or within CD4+ or CD8+ T cells.


Data were analysed using Graphpad Prism version 5·0 (Graphpad Software Inc., San Diego, CA, USA). Results were expressed as medians and 25–75th percentiles. The non-parametric Kruskal–Wallis test was used to compare data from TB patients and healthy individuals, followed by Mann–Whitney U-test. Friedman's test was performed to compare data from pleural and peripheral blood samples, followed by Wilcoxon's test. Correlations were performed by the non-parametric Spearman's rank test. All statistical analyses were two-sided, and the significance level adopted was for P-values of <0·05.


The proportion of CD4+CD25+ T cells is increased in TB-PFMC

As shown in Table 2, the percentage and absolute number of CD4+ T cells were decreased in PBMC from TB (TB-PB) when compared to PBMC from N (N-PB), while no differences were observed in CD8 T cells. The percentages of CD4+CD25+ and CD8+CD25+ cells were higher in TB-PB that in N-PB. Conventional and regulatory CD4+ T cells were analysed discriminating low or high CD25 expression on the basis of CD25 mean fluorescence intensity (MFI) within the CD4 population. We observed that TB-PB showed a higher percentage of CD4+CD25low and CD4+CD25high cells when compared to N-PB. When we evaluated the expression of FoxP3, considered a marker of natural Treg cells, a higher proportion of CD25highFoxP3+ was detected within the CD4+T subset from TB-PB than in N-PB, while no differences were found in the proportion of CD25+FoxP3+ within the CD8+ subset (Table 2).

Table 2. Increased levels of CD4+CD25+ cells in pleural fluid (PF) and peripheral blood (PB) from patients with tuberculous pleurisy
  1. Mononuclear cells isolated from paired samples of PF and PB obtained from 16 patients with tuberculous pleurisy (TB-PF and TB-PB, respectively) as well as PB from 10 healthy individuals (N-PB) were tested for the expression of surface CD4, CD8 and CD25 and intracellular forkhead box P3 (FoxP3) by flow cytometry. Results are expressed as percentage (%) of (a) total CD4+ and CD8+ cells, CD4+CD25+ and CD8+CD25+, CD4+ expressing high or low CD25 expression (CD4+ CD25high and CD4+CD25low) within the lymphocytes gates, (b) percentage of CD25+FoxP3+ cells within the CD4+ and CD8+ subsets (CD25highFoxP3+ in CD4 or CD8). Absolute numbers (cells/mm3) were also calculated. Median and 25–75% percentiles are shown. Significant differences between TB-PB versus N-PB: *P < 0·05; TB-PB versus TB-PF: #P < 0·05.
Total CD4+ T cells60·0 (55·1–65·2)#477 (366–771)34·2 (22·3–42·8)*500 (280–853)*53·1 (45·8–56·0)1113 (962–1176)
CD4+CD25+13·0 (7·6–18·2)#157 (54–234)5·6 (3·8–8·1)*168 (74–238)*4·7 (4·1–6·7)135 (48–188)
CD4+ CD25 high2·5 (2·0–3·0)#30 (26–77)1·7 (1·1–2·1)*56 (33–72)*1·1 (0·7–1·6)43 (16–54)
CD4+CD25low12·1 (8·4 ± 15·6)#123 (77–156)5·6 (3·8–8·1)*152 (89–281)*3·4 (2·7–5·2)115 (57–150)
CD25highFoxP3+ in CD42·3 (1·5–2·8)#31 (15–61)1·5 (1·0–2·0)*59 (36–83)*1·0 (0·6–1·4)23 (11–27)
Total CD8+ T cells23·4 (19·6–25·0)#147 (101–192)39·3 (25·7–43·1)441 (333–742)31·5 (29·1–35·0)662 (611–735)
CD8+CD25+4·0 (3·5–5·9)#34 (4–169)1·8 (1·5–2·3)*13 (6–109)0·6 (0·4–1·3)13 (8–17)
CD25+ FoxP3+ in CD80·6 (0·5–1·6)#5 (2–13)0·3 (0·1–0·8)5 (3–14)0·4 (0·2–0·5)3 (2–4)

PFMC from TB (TB-PF) showed a higher percentage but a similar absolute count of CD4+T cells than TB-PB (Table 2). The proportion of total CD4+CD25+, CD4+CD25high and CD4+CD25low cells were higher in TB-PF than in TB-PB. Also, an increased percentage of CD25+FoxP3+ within CD4+ cells was observed in TB-PF compared to TB-PB (Table 2); 65–97% were CD127NEG in PF and in PB (data not shown), confirming their condition of Tregs. However, no differences in the absolute number of CD4+CD25+ and CD4+CD25highFoxP3+ cells were observed between PB and PF samples. The percentage and absolute number of CD8+ T cells were diminished in TB-PF when compared to TB-PB; however, the percentage of CD25+ cells was enhanced in TB-PF. Although the percentage of CD25+FoxP3+ cells was also enhanced in the CD8+ subset, it represents only 17·3% (2·9–25·2) of CD8+CD25+ cells from TB-PF, suggesting that CD8+CD25+ cells were mainly classically activated T cells.

IL-10 and IFN-γ are expressed differentially in PBMC and PFMC from TB patients

In order to delineate the cytokine profile of CD4+ and CD8+ T cells, the intracellular expression of IL-10 and IFN-γ was determined in PB and PF from TB patients upon stimulation with M. tuberculosis for 18 h. As shown in Fig. 1a,c, M. tuberculosis enhanced the proportions of CD4+IL-10+ and CD8+IL-10+ cells in PF and PB samples, the highest values being observed in PB from TB patients. Conversely, the percentage of CD4+ and CD8+ T cells expressing IFN-γ was increased upon M. tuberculosis stimulation, but the highest values were observed in TB-PF (Fig. 1b,c). Although 85–97% of CD4+CD25high cells expressed IL-10 in PF and PB samples, the main IL-10 expression was found in CD4+CD25low/neg cells from PB. Furthermore, IFN-γ was expressed mainly by CD4+CD25low/neg cells in PF (Fig. 1d).

Figure 1.

Interleukin (IL)-10 and interferon (IFN)-γ are expressed differentially in peripheral blood mononuclear cells (PBMC) and pleural fluid mononuclear cells (PFMC) from tuberculosis (TB) patients. TB-PF and TB-PB cells from 14 TB patients were cultured for 18 h alone (control, C) or with Mycobacterium tuberculosis. Then cells were tested for the intracellular expression of IL-10 and IFN-γ by flow cytometry. (a) Percentage of CD4+ IL-10+ and CD8+ IL-10+ cells within the lymphocyte gates. Results are expressed as median and 25–75 percentiles. (b) Percentage of CD4+ IFN-γ+ and CD8+ IFN-γ+ cells within the lymphocyte gates (median and 25–75 percentiles). (c) Representative dot-plots from one TB patient are shown. Numbers in the upper-right quadrant represent the percentage of CD4+IL-10+, CD8+IL-10+, CD4+IFN-γ+ and CD8+IFN-γ+ cells within live lymphocyte gate in PB (gate PB-Ly) and PF samples (gate-PF-Ly). (d) Percentage of IL-10+ and IFN-γ+ cells within the CD4+CD25high, CD4+CD25low and CD4+CD25neg subsets from non-stimulated (control, C) and M. tuberculosis-stimulated TB-PB and TB-PF cells. (a–c) Statistical differences: *P < 0·05 for M. tuberculosis-stimulated versus control cells, aP < 0·05 for TB-PF versus TB-PB values, #P < 0·05 for CD4+CD25low/negversus CD4+CD25high.

CD107 expression is enhanced in CD4+ and CD8+ T cells from PFMC

Considering that CD107a cell surface expression has been described as a marker of cytotoxic T cell degranulation/activation [22, 23] and is closely associated with the ability of T cells to produce IFN-γ [23], CD107 expression was evaluated on CD4+ and CD8+ T cells upon stimulation with M. tuberculosis for 18 h. As shown in Fig. 2a,b, M. tuberculosis induced CD107 expression on CD8+ and CD4+ T cells from TB-PB and TB-PF. The percentages of CD4+CD107+ and CD8+CD107+ cells in PF were higher than in their PB counterpart and confined to CD4+/CD8+CD25low/neg cells (data not shown). The percentage of CD107+ cells also correlated with the percentage of IFN-γ+ cells in M. tuberculosis-stimulated PB and PF from TB (Fig. 2c).

Figure 2.

CD107 expression is enhanced in CD4+ and CD8+ T cells from tuberculosis-pleural fluid (TB-PF). (a) TB-PF and TB-periheral blood (PB) cells from 14 TB patients were cultured for 18 h alone (c) or with Mycobacterium tuberculosis and then tested for their CD107 surface expression by flow cytometry. Results are expressed as percentage of CD4+CD107+ and CD8+CD107+ cells within the lymphocyte gates (median and 25–75 percentiles). Statistical differences: *P < 0·05 for M. tuberculosis-stimulated versus control cells; aP < 0·05 for TB-PF versus TB-PB. (b) Representative dot-plots from one TB patient are shown. Numbers in the upper-right quadrant represent the percentage of CD4+CD107+ and CD8+CD107+ in live lymphocyte gate in PB (gate PB-Ly) and PF samples (gate-PF-Ly). (c) Correlation between %CD4+/CD8+ CD107+ and %CD4+/CD8+ IFN-γ+ in control and M. tuberculosis-stimulated TB-PB (C-TB-PB, M. tuberculosis-TB-PB), TB-PF (C-TB-PF and M. tuberculosis-TB-PF). Individual data and Spearman's rho coefficients are shown.

Treg cells suppress M. tuberculosis-induced degranulation and IFN-γ expression in CD4+ and CD8+ T cells from TB-PF

Finally, we wondered whether CD4+CD25high Treg were able to modulate in-vitro M. tuberculosis-induced CD4+ and CD8+ T cell responses in PF. For this purpose, PBMC and PFMC were depleted of CD25high cells by magnetic methods and cultured for 18 h alone or with M. tuberculosis; cells were then tested for their expression of IL-10, IFN-γ and CD107. As shown in Fig. 3a, CD25high depletion decreased the percentage of M. tuberculosis-stimulated CD4+IL-10+ and CD8+IL-10+ cells from PF and PB. In contrast, M. tuberculosis-stimulated PB and PF cells depleted of CD25high cells showed increased percentages of CD4+ and CD8+ cells expressing IFN-γ (Fig. 3b) and degranulating CD4+ and CD8+ (Fig. 3c). Again, a direct correlation was observed between CD107+ and IFN-γ+ cells in PB and PF samples (Spearman's r, PB samples: CD4 = 0·7727, P = 0·0053 and CD8 = 0·6894, P = 0·0064; PF samples: CD4 = 0·6713, P = 0·0168 and CD8 = 0·8912, P < 0·0001). These results suggested that Treg efficiently suppressed M. tuberculosis-induced IFN-γ in CD4+ and CD8+ T cells, a fact that impacted directly on their ability to degranulate in response to M. tuberculosis even in a Th1 microenvironment such as observed in PF.

Figure 3.

Regulatory T cells (Treg) suppress Mycobacterium tuberculosis-induced degranulation and interferon (IFN)-γ expression in CD4+ and CD8+ T cells from tuberculosis-pleural fluid (TB-PF). TB-peripheral blood (PB) and TB-PF from 10 TB patients were depleted of CD25+ cells by magnetic methods. CD25high-depleted and non-depleted cells were cultured for 18 h alone or with M. tuberculosis and tested for intracellular interleukin (IL)-10 (a) and IFN-γ (b) or and surface CD107 expression (c). Results are expressed as a percentage of CD4+ and CD8+ cells expressing IFN-γ, IL-10 and CD107 within the lymphocyte gates (median and 25–75 percentiles). For all graphs, statistical differences between non-depleted and CD25high-depleted PBMC: *P < 0·05 (Wilcoxon's rank sum test).

Association between Treg and effector T cells and clinical features

We wondered whether the proportion of Treg and CD4+ and CD8+ T cells that express intracellular IL-10, IFN-γ or surface CD107 correlated with clinical data, such as the presence of concomitant pulmonary tuberculosis and acid-fast bacillus (AFB) in sputum using Ziehl–Neelsen-stained smears, antigen load (number of bacilli/field), PPD-positive skin test and days of evolution. As shown in Fig. 4a, the proportion of CD4+CD25high cells in PB and PF from patients with combined pleural and pulmonary disease tuberculosis was higher than in patients with pleural TB, while no differences were observed in the proportion of CD4+CD25low cells. Furthermore, higher percentages of IL-10+ cells (Fig. 4b), together with lower IFN-γ+ and CD107+ cell levels (Fig. 4c,d, respectively), were observed in those patients with combined pleural and pulmonary TB. All patients with combined pleural and pulmonary TB had positive sputum smears and moderate to advanced disease at the time of the study, while most of the patients with pleural TB had unilateral PF and negative sputum smears; therefore, it is tempting to speculate that bacterial load and/or severity of the disease could be influencing Treg expansion and, in turn, the Th1 response in the pleural space.

Figure 4.

Patients with combined pleural and pulmonary tuberculosis (TB) showed differential regulatory T cells (Treg) and effector T cell frequencies in peripheral blood and pleural effusions. TB patients were grouped according to the presence (pleural and pulmonary) or absence (pleural) of combined pulmonary tuberculosis. Then, the medians and 25th to 75th percentiles of (a) %CD4+CD25high and CD4+CD25low cells, (b) %CD4+ IL-10+ and CD8+ IL-10+ cells, (c) %CD4+ IFN-γ+ and CD8+ IFN-γ+ cells and (d) %CD4+ CD107+ and CD8+ CD017+ cells were calculated for each group (mean and 25–75th percentiles). Statistical differences, patients with pleural and pulmonary tuberculosis versus patients with pleural tuberculosis: *P < 0·05.


Immune response to infection is a complex balance between successful induction of proinflammatory responses against the pathogen and anti-inflammatory responses required to limit damage to host tissues. In this context, Treg lymphocytes, which were initially found to control inflammation mediated by autoimmune diseases, play an important role in the control of the inflammatory process induced by infectious pathogens [25, 26]. Increased levels of CD4+CD25highFoxP3+ Treg cells have been detected in PBMC from pulmonary TB patients compared to uninfected [12, 15] or healthy infected [13] subjects. Ex-vivo depletion of circulating CD4+CD25high T cells from TB patients also resulted in enhanced IFN-γ production and CTL response upon stimulation with mycobacterial antigens, suggesting a role of Treg in the pathogenesis of TB [12, 14, 21]. Also, decreased or increased Treg proportions have been demonstrated in samples from different sites of M. tuberculosis infection such as pleural, ascitic or pericardial fluids and in the lung alveolar compartment [12, 16, 17, 27, 28]. In this study, we showed that the proportion of CD4+CD25highFoxP3+ Treg cells was increased in PBMC from patients with pleural TB compared to healthy infected controls, as was also observed in other studies [12, 16, 28]. In agreement with the results reported by other authors [16], CD4+CD25highFoxP3+ Treg cells were detected in TB-PF, indicating that these cells are recruited efficiently to the site of infection. However, as we did not include a control group, such as patients with non-tuberculous pleurisy, we cannot rule out that Treg recruitment could be a response to M. tuberculosis infection or a host mechanism to reduce inflammation in the pleural compartment.

As demonstrated in this study, M. tuberculosis induced a high proportion of IL-10+ T cells in PB, whereas high percentages of IFN-γ+ T cells were found in PF samples, which is in concordance with other studies [7, 27, 29, 30]. Therefore, our results confirm that in spite of the systemic Th2 profile observed in TB patients a switch to a local Th1 profile is observed within the tuberculous PF. It is well known that a Th1 microenvironment favours CD4 and CD8 CTL differentiation and survival [31, 32]. In line with this, we found that the increased proportions of M. tuberculosis-induced CD4+CD107+ and CD8+CD107+ cells correlated closely with the proportions of M. tuberculosis-induced IFN-γ+ cells in PF, as we observed previously in circulating M. tuberculosis-induced CD8 CTL [21]. M. tuberculosis-induced CTL has been ascribed to the CD8 subset, although cytolytic CD4+ T cells specific for M. tuberculosis have also been reported [33-35]. Similarly, CD1-restricted CD4+CTL differentiation has been demonstrated in response to mycobacterial lipids [36]. Hence, the high proportion of CD4+CTL suggests that CD4+ cells expressing IFN-γ and/or with lytic capacity are also attracted at the site of infection and may contribute, in concert with CD8+CTLs, to the lysis of infected antigen-presenting cells. In this context, the negligible bacterial load observed in patients with tuberculous pleurisy could be related, among other factors, to the high CTL function mediated by CD4+ and CD8+ T cells.

Little is known about the role of Treg in a clear Th1 environment such as TB pleurisy. It has been demonstrated that early secreted antigenic target protein 6 (ESAT-6)- and bacilli Calmette–Guérin (BCG)-induced IFN-γ production by PB and PFMC from TB patients is correlated inversely with the percentage of CD4+CD25+FoxP3+ Treg cells [28] and their depletion with increased CD4+CD25neg proliferation [16, 37]. In this study we showed that Treg depletion decreased the percentage of CD4+IL-10+ and CD8+IL-10+ cells in PB as well as in PF samples. Although it is tempting to speculate that a 37–80% decrease in M. tuberculosis-induced IL-10 expression could be due to partial depletion of activated CD25low cells, it is likely that IL-10 was induced by Treg from CD4+CD25neg cells, as observed by other authors [38].

It has been proposed that Treg-mediated suppression could account for the lack of M. tuberculosis-induced IFN-γ and CTL effector functions in TB [12, 15, 21], the latter function being down-regulated by Treg through direct cell-to-cell contact [39]. In this way, suppression of the influenza vaccine-mediated CD8+CTL response by human circulating Treg cells involves down-regulation of IFN-γ and perforin expression as well as conjugate formation by CTL with target cells [40]. Herein, we showed that depletion of CD25high cells increased IFN-γ and CD107 expression in CD4+ and CD8+ cells from PB and from PF cells, although co-cultures of isolated Treg and non-Treg cells were not performed to confirm the suppressor functionality of PB and PF Treg cells. Both CD107 and IFN-γ expression in CD25high depleted cells were related closely to each other. Thus, in accordance with the suppressor ability described for peripheral Treg, these cells indeed inhibit IFN-γ expression, and in turn the ability of CD4+ and CD8+ cells to degranulate, in response to M. tuberculosis in TB pleurisy. Considering that CTL activity has been associated with control of mycobacterial growth through the killing of infected macrophages and also to tissue damage, Treg might contribute to down-regulating excessive CTL activity and protect the host from extensive tissue damage in the pleural space. In line with this, our study demonstrated that in PB and PF samples from patients with combined pleural and pulmonary TB there were increased Treg frequencies compared to those patients with pleural TB. It has been demonstrated that IL-4 induces CD4+CD25+ Treg cell from CD25neg PB-T cells [41], this cytokine being associated with the severity of the disease [42, 43] and detected in bronchoalveolar lavage (BAL) samples [44]. As all patients with combined pleural and active TB showed moderate to severe lung disease and positive sputum smear, an increased frequency of Treg in PB and PF might be associated with bacterial burden or inflammation related to active disease, as observed by other authors [21, 45-47]. In these patients, high IL-10 expression and suppression of IFN-γ and CTL response in PB and PF samples could be the consequence of a protective mechanism to limit pulmonary and extra-pulmonary inflammation.


This work was supported by grants from the Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT, PAE-PICT 07-2328, PAE-PICT 07-2329 and PICT 2010-0599).

Author contributions

L. G., J. I. B., N. Y. and C. S. B. were responsible for sample preparation and data collection. R. M. and J. C. were responsible for patient recruitment and selection and for sample collection. M. C. S. and S. de la B. were responsible for study design, data analysis and manuscript preparation.


None of the authors have either a commercial or other association that might pose a conflict of interest. No one involved in the publication process has a financial or other beneficial interest in the products or concepts mentioned in the submitted manuscript or in competing products that might bias his or her judgement.