Regulatory T Cells Attenuate Mycobacterial Stasis in ... - ATS Journals

Regulatory T Cells Attenuate Mycobacterial Stasis in Alveolar and Blood-derived Macrophages from Patients with Tuberculosis Patricia L. Semple1*, Anke B. Binder1*, Malika Davids1, Alice Maredza1, Richard N. van Zyl-Smit1, and Keertan Dheda1 1

Division of Pulmonology and UCT Lung Institute, Department of Medicine, University of Cape Town, Cape Town, South Africa

Rationale: There are hardly any data about the frequency of CD41 CD251Foxp31 regulatory T cells (T-Regs) in the lungs of patients with active tuberculosis (TB). Objectives: To obtain data about the frequency of CD41CD251 Foxp31 T-Regs, and their impact on mycobacterial containment, in the lungs of patients with active TB. Methods: Patients with pulmonary TB (n ¼ 49) and healthy volunteers with presumed latent TB infection (LTBI; n ¼ 38) donated blood and/or bronchoalveolar lavage (BAL) cells obtained by bronchoscopy. T-cell phenotype (Th1/Th2/Th17/T-Reg) and functional status was evaluated using flow-cytometry and 3H-thymidine proliferation assays, respectively. H37Rv-infected alveolar and monocytederived macrophages were cocultured with autologous T-Regs and purified protein derivative (PPD) preprimed T-Reg–depleted effector cells. Mycobacterial containment was evaluated by counting CFUs. Measurements and Main Results: In blood and BAL T-Reg levels were higher in TB versus LTBI (P , 0.04), and in TB the frequency of T-Regs was significantly higher in BAL versus blood (P , 0.001). T-Reg– mediated suppression of T-cell proliferation in blood and BAL was concentration-dependent. Restriction of mycobacterial growth in infected alveolar and monocyte-derived macrophages was significantly diminished, and by up to 50%, when T-Regs were cocultured with PPD-primed CD41 effector T cells. The levels of CD81 T-Regs (CD81 CD251Foxp31), IL-17–producing T-Regs (IL-171CD41CD251Foxp31), and IL-17–producing T cells were similar in BAL-TB versus BAL-LTBI. Within the TB group compartmentalization of responses was prominent (T-Reg, IFN-g, tumor necrosis factor-a, IL-17, and IL-22 significantly higher in BAL vs. blood). Conclusions: In patients with TB the alveolar compartment is enriched for CD41 T-Regs. Peripheral blood-derived T-Regs decrease the ability of alveolar and monocyte-derived macrophages to restrict the growth of Mycobacterium tuberculosis in the presence of effector cells. Collectively, these data suggest that CD41CD251 FoxP31 T-Regs subvert antimycobacterial immunity in human TB. Keywords: natural regulatory T cells; FoxP3; bronchoalveolar lavage; pulmonary tuberculosis; mycobacterial stasis

(Received in original form October 29, 2012; accepted in final form March 11, 2013) *These authors have contributed equally to the work. Supported by the EDCTP (TB-NEAT and TESA); the South African Medical Research Council; and the South African National Research Foundation (SARChI). Author Contributions: Conception and design, K.D., P.L.S., R.N.v.Z.-S. Laboratory experiments, P.L.S., A.M., M.D., R.N.v.Z.-S. Analysis and interpretation, K.D., P.L.S., A.B.B., A.M., R.N.v.Z.-S., M.D. Drafting the manuscript for important intellectual content, K.D., P.L.S., A.B.B., M.D. Correspondence and requests for reprints should be addressed to Keertan Dheda, M.D., Ph.D., Division of Pulmonology and UCT Lung Institute, Department of Medicine, Old Main Building, Observatory 7439, Cape Town, South Africa. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Crit Care Med Vol 187, Iss. 11, pp 1249–1258, Jun 1, 2013 Copyright ª 2013 by the American Thoracic Society Originally Published in Press as DOI: 10.1164/rccm.201210-1934OC on April 3, 2013 Internet address: www.atsjournals.org

AT A GLANCE COMMENTARY Scientific Knowledge on the Subject

There are hardly any human data about the frequency of CD41CD251FoxP31 regulatory T cells (T-Regs) in the alveolar lung compartment of patients with active tuberculosis (TB), and although T-Regs have been shown to attenuate antigen-specific T-cell responses, their role in mycobactericidal immunity in humans remains unclear. What This Study Adds to the Field

Our data indicate that the alveolar lung compartment in patients with active TB, relative to the blood compartment, is enriched for T-Regs. Furthermore, coculture of autologous T-Regs with alveolar and monocyte-derived macrophages infected with pathogenic strains of Mycobacterium tuberculosis resulted in decreased restriction of mycobacterial growth ex vivo. Collectively, these data suggest that CD41CD251FoxP31 T-Regs subvert antimycobacterial immunity in human TB.

According to the latest World Health Organization report, one person dies of tuberculosis (TB) every 15 seconds (1). However, despite the enormity of the problem and considerable research efforts, the mechanisms underpinning susceptibility and disease progression are poorly understood. Active TB is characterized by a suppression of Mycobacterium tuberculosis (M.tb)–specific T-cell responses (2, 3). However, the immune mechanisms allowing T cell–related permissiveness to infection have not been fully elucidated. The immune system has several regulatory mechanisms for suppressing the effector response to persistent antigens. In several diseases regulatory T cells (T-Regs) play a central role in the prevention of autoimmunity and in the control of immune responses by down-regulating the effector function of CD41 or CD81 T cells. Naturally occurring CD41CD251 FoxP31 T-Regs, a subset of CD41 T cells with immunosuppressive properties, have been implicated in intracellular pathogen host defense (4–6). The regulation for T-Reg is predominantly dictated by the expression of the forkhead box P3 transcription factor, Foxp3. This distinguishes T-Regs from activated effector cells within the CD251 T-cell population and is fundamental for the development and function of these suppressive cells (7, 8). T-Regs are increased in tuberculous pleural effusions and strongly suppress the proliferation of CD41CD252 T cells (9). M.tb-specific T-Regs have an increased propensity to proliferate and even small numbers of M.tb-specific T-Regs are capable of delaying the priming of murine effector CD41 and CD81 T cells in the mediastinal lymph nodes and their subsequent accumulation in the lung (10). T-Regs specifically recognize M.tb-derived antigens and inhibit Th1 antigen-specific cytokine effector responses (2). However, whether suppression of effector responses per se

1250

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 187

2013

Cytokine Profiles, T-Cell Phenotype, and Isolation of T-Reg

translates into attenuation of mycobactericidal immunity remains unclear. By contrast, T-Regs may accumulate in the face of chronic inflammation and also play a role in preventing immunopathology and limiting collateral damage to the host. The debate whether T-Regs enhance or attenuate mycobactericidal host immunity in humans persists, because overregulation of Th1 responses may lead to susceptibility, too little regulation may allow immunopathology, and inappropriate regulation may encourage nonprotective pathways. Although T-Regs are increased in the peripheral blood (PB) compartment of patients with TB (3, 4, 11, 12) there are hardly any data about the comparative frequency at the primary site of disease (i.e., the human lung). Furthermore, there are no data about whether T-Regs when cocultured with infected human macrophages can marginalize mycobacterial containment. To address these gaps in our knowledge, we evaluated compartmentspecific T-cell profiles and mycobacterial stasis in in vitro cultures of infected monocyte-derived and pulmonary macrophages from patients with TB and presumed latently infected control subjects (latent TB infection [LTBI]).

To determine cytokine profiles and T-helper cell phenotype, and the relative proportions of T-Regs in each compartment, PBMC and BAL cells were stimulated with purified protein derivative (PPD; Statens Serum Institut, Copenhagen, Denmark) or phytohemagglutinin (Murex, Fisher Scientific, Ottawa, ON, Canada), which served as a positive control. After 12 hours, the cells were stained with antibodies against CD3, CD4, CD8, CD25, and CCR6 (BD Biosciences, San Jose, CA); FoxP3, IL-10, IL-17, and IFN-g (eBiosciences, San Diego, CA); and tumor necrosis factor-a, IL-22, and IL-13 (BD Biosciences). T-Regs were isolated from PB and BAL using the Miltenyi Regulatory T cell isolation kit (Miltenyi Biotec, Surrey, UK). To confirm the purity of the obtained cell fractions, the positively selected cells were retrieved from the magnetic column and the T-Regs thus obtained, together with the flow-through cells (CD41/CD252) (i.e., non–T-Reg [NTR]), were stained with antibodies against CD3, CD4, CD25, and CD127 (BD Biosciences) and detected by flow cytometry. To confirm that CD4 1/CD25 1CD127 dim/2 cells were FoxP3 1, PBMC were stained with antibodies against CD3, CD4, CD25, and CD127/FoxP3 and acquired on a BD LSRII flow cytometer before analysis.

METHODS

Suppression of Cell Proliferation

Human Participation and Ethical Approval The participants comprised two groups (active TB and healthy control subjects with presumed LTBI) from who blood and bronchoalveolar lavage fluid (BAL) were obtained. The diagnostic reference standard for pulmonary TB was culture positivity for M.tb (MGIT 960; Becton and Dickinson, Franklin Lakes, NJ), together with a radiographic picture consistent with TB, and with a clinical response to anti-TB treatment. Healthy individuals with normal chest radiographs and who were positive for tuberculin skin test (10 mm cut-point) and/or QuantiFERONTB Gold (Cellestis, Valencia, CA) were recruited as presumed LTBI control subjects. All the participants were HIV uninfected. None of the patients were receiving anti-TB therapy at the time of enrolment and those with immunosuppressive, endocrine, renal, hepatic, or malignant disease were excluded from the study. Participants provided written informed consent and study approval was obtained from the Human Ethics Committee at the University of Cape Town.

Cell Isolation from PB and Alveolar Lavage Fluid PB mononuclear cells (PBMC) were isolated from PB by density centrifugation. Alveolar lavage cells were obtained by a pulmonologist using bronchoscopy and low-pressure suction after instilling approximately 300 ml sterile saline in an affected segment (preferentially from the right middle lobe, lower lobes, or upper lobes in that order). The cells obtained from blood and BAL fluid were used in downstream experiments including flow cytometry to determine cytokine profiles and T-cell phenotype. T-Regs were isolated from the blood and cocultured with other cells to determine the suppressive impact on cell proliferation and mycobacterial containment (killing/stasis assays).

To drive proliferation T-Regs and NTR (CD41/CD252) isolated from PBMC and BAL fluid were stimulated with purified anti-CD3 (BD Biosciences) and anti-CD28 for 3 days. To determine whether T-Regs suppressed the proliferative capacity of NTR in a dose-dependent manner, T-Regs were added to NTR at a concentration of one T-Reg to one NTR, one T-Reg to two NTR, and one T-Reg to four NTR (see Table 1 for cell numbers used). Three days later the cells were pulsed with titrated thymidine (3HTdr; Amersham, Aylesbury, UK) and harvested.

Mycobacterial Stasis Assays To evaluate their impact on mycobacterial containment T-Regs were cocultured with infected macrophages (either blood or BAL-derived) and effector cells. To generate effector T cells, the NTR fraction from PB (obtained using the T-Reg isolation kit) was stimulated with PPD for 6 days. In parallel, T-Regs alone and a combination of NTR and T-Regs (1:1 ratio) were similarly stimulated with PPD for 6 days. Five days after initial phlebotomy, bronchoscopy and repeat phlebotomy was undertaken to obtain further cells. Alveolar macrophages (AM) obtained from BAL fluid or derived from blood monocytes, guided by the relevant optimization experiments, were infected with H37Rv at a multiplicity of five H37Rv to one AM (MOI 5:1) or one H37Rv to one monocyte-derived macrophage (MOI 1:1). Organism counts were verified by randomly plating out standardized frozen aliquots of M.tb. After approximately 18 hours the cells were washed to remove extracellular mycobacteria. Thereafter, previously generated (as outlined previously after initial phlebotomy) PPD-stimulated autologous cells (T-Regs or NTR or a combination in a 1:1 ratio) were added to the infected AM or monocyte-derived macrophages (T-Regs or NTR or a combination in a 1:1,

TABLE 1. NUMBER AND TYPE OF CELLS USED IN THE INTERVENTION AND CONTROL WELLS, WHERE RELEVANT, WHEN EVALUATING THE SUPPRESSION OF T-REGS ON T-CELL PROLIFERATION, AND THEIR EFFECT ON MYCOBACTERIAL STASIS IN ALVEOLAR AND MONOCYTE-DERIVED MACROPHAGES Dose-Dependent Coculturing of T-Reg with Non–T-Reg Fraction 0.5 3 105 T-Reg 1 0.5 3 105 non–T-Reg*

1:1

0.33 3 105 T-Reg 1 0.66 3 105 non–T-Reg*

1:2

0.25 3 105 T-Reg 1 0.75 3 105 non–T-Reg*

1:4

H37Rv-infected Alveolar or Monocyte-derived Macrophages with 6 d PPD-stimulated Cells 1 3 105 infected macrophages with media only

1 3 105 infected macrophages with 1 3 105 T-Reg†

1 3 105 infected macrophages with 1 3 105 non–T-Reg†

1 3 105 infected macrophages with 0.5 3 105 T-Reg† 1 0.5 3 105 non–T-Reg†

Definition of abbreviations: non–T-Reg ¼ peripheral blood mononuclear cells depleted of T-Reg CD41/CD252 cells; PPD ¼ purified protein derivative; T-Reg ¼ CD41/ 251/FoxP31 cells. * Media supplemented with CD3/anti-CD28. y After 6 days of PPD stimulation.

Semple, Binder, Davids, et al.: FoxP31 Regulatory T Cells and Mycobacterial Stasis

2:1, and 4:1 ratio) for approximately 24 hours (details outlined in Table 1). CFUs were enumerated in all wells including a reference control containing infected alveolar or monocyte-derived macrophages only. This reference (arbitrarily designated 100% survival) served as control against which relative mycobacterial stasis was expressed. The term “decreased ability to restrict the growth of M. tb” was defined as reduction in antimycobacterial effect caused by either killing and/or driving a state of nonreplication.

Statistical Analysis The Mann-Whitney two-tailed test for nonparametric data or KruskalWallis test was used for group comparison. All statistical analysis was performed using GraphPad Prism software (version 5.0, GraphPad Software, La Jolla, CA).

RESULTS T-Regs with a CD41/251/CD127dim/2 Phenotype Express the Intracellular FoxP3 Marker and Are Significantly Increased at the Site of Infection

To determine whether T-Regs with a CD41/251/CD127dim/2 phenotype also express the FoxP3 intracellular marker PB and BAL T-Regs isolated from five subjects using the Miltenyi Regulatory Cell Kit were stained with antibodies to CD3, CD4, CD25, CD127, and FoxP3. There was an inverse correlation; more than 80% of the isolated cells that were CD127dim/2 were positive for FoxP3 (81.6 6 8.4%) (representative dot-plot is shown in Figure E1D in the online supplement). To determine the compartment-specific proportion of CD41 T cells that were T-Regs, the blood (from 16 patients with TB and 14 with LTBI) and BAL (from 16 patients with TB and 6 with LTBI) were subjected to flow cytometry and in parallel cell counts and differentials were determined before the isolation of T-Regs to determine the percentage of lymphocytes recovered from each subject. Figure 1A shows the percentage of isolated T-Regs (using the Miltenyi kit) expressed as a fraction of CD41 T cells. T-Regs were significantly higher in the blood (P ¼ 0.007) and BAL (P , 0.001) of patients with TB compared with LTBI. However, the frequency (% of CD41 lymphocytes) in the BAL of the patients with TB was significantly higher than in the blood of the same patients (P ¼ 0.0001). Figure 1B shows the percentage of T-Regs based on flow cytometric detection (rather than cell isolation) of CD41CD251 FoxP31 cells; there was a similar pattern including relative enrichment of the BAL compartment in patients with TB (P ¼ 0.003). T-Regs Isolated from PB and BAL Are Functionally Suppressive in Patients with Active TB

Because there are no compartment-specific comparative data about the functional characteristics of these cells we tested the ability of blood and BAL-derived CD4 1/25 1/FoxP3 1 T-Regs to suppress proliferative responses in CD3/anti-CD28–

1251

driven CD41/252 cells (NTR) obtained from the blood and BAL of five patients with TB (T-Regs were added to NTR fraction at a ratio of 1:1, 1:2, and 1:4). T-Regs were anergic (did not respond to T-cell receptor stimulation) but NTR had a high proliferative response as determined by counts per million (data not shown). When coculturing T-Regs and NTR at a ratio of 1:1 in the same wells, T-Regs reduced the proliferative capacity of NTR by a median of 71.0% in the PB (Figure 2A) and by a mean of 62.3% in the BAL (Figure 2B). At lower ratios of T-Reg to NTR the reduction was diminished in a dose-dependent manner in PB and BAL (Figures 2A and 2B). T-Regs Decrease the Restriction of Mycobacterial Growth in M.tb-infected Monocyte-derived Macrophages from PB

PB was obtained to assess whether T-Regs impacted growth restriction of M.tb (reduction in CFU) in monocyte-derived macrophages obtained from four patients with pulmonary TB (Figures 3C and 3D) and three control subjects with LTBI (Figures 3A and 3B). Summated data from the combined data set (TB and LTBI; n ¼ 7) are shown in Figures 3E and 3F. The leftside panels (Figures 3A, 3C, and 3E) show the data presented as absolute CFU counts, whereas the right-side panels (Figures 3B, 3D, and 3F) show the data presented as a % reduction in CFUs. Figure 3G shows a summary line plot of the % reduction in CFU in the NTR versus the NTR 1 T-Reg group. For example, compared with the infected monocyte-derived macrophages alone, there was a median (6 SD) reduction in mycobacterial growth (% CFU) of 36.1 6 3.9 by the effector NTR cells (Figure 3F). By contrast T-Regs alone (6.0 6 8.9%) and cocultured NTR/T-Reg (12.7 6 8.9%) had a minimal effect on mycobacterial restriction of growth (Figure 3F). Thus, there was a significant reduction in % CFU growth restriction in the NTR group compared with the cocultured NTR/T-Reg group (P ¼ 0.0002) (Figure 3G). T-Regs Decrease the Restriction of Mycobacterial Growth in M.tb-infected AM

To evaluate whether T-Regs decreased the ability of effector cells to restrict the growth of M.tb at the site of disease, AMs from four patients with pulmonary TB (Figures 4C and 4D) and five LTBI control subjects (Figures 4A and 4B) were obtained by bronchoscopy. Combined data (active TB and LTBI; n ¼ 9) are shown in Figures 4E and 4F. The same pattern was seen compared with PB: coculture with blood-derived T-Regs significantly reduced the restriction in mycobacterial growth by effector cells. For example, there was a median (6 SD) reduction (% CFU) of 42 6 13.3% by the effector NTR compared with the infected AM alone, 9.3 6 8.1% by T-Regs alone, and 11.2 6

Figure 1. (A) Percentage of peripheral blood and bronchoalveolar lavage (BAL)-derived CD41 lymphocytes that are regulatory T cells (T-Regs; CD41CD251Foxp31) after T-Reg isolation in patients with active tuberculosis (TB) (n ¼ 24 in blood, n ¼ 16 in BAL) and presumed latent TB infection (LTBI) (n ¼ 14 in blood; BAL is not shown because the low T-Reg count precluded reliable isolation). (B) The same data obtained through staining and flow cytometric detection (not isolation) in patients with active TB (n ¼ 19 in blood, n ¼ 6 in BAL) and presumed LTBI (n ¼ 10 in blood, n ¼ 6 in BAL).

1252


VOL 187

2013

Figure 2. Representative sample of the suppressive effects of regulatory T cells in (A) peripheral blood and (B) alveolar lavage cells of patients with tuberculosis. Purified regulatory T cells were added to anti-CD3/CD28–stimulated cells at a ratio of 1:1, 1:2, and 1:4.

9.9% by the cocultured NTR/T-Reg (Figure 4F). Thus, there was a significant reduction in % CFU growth restriction in the NTR group compared with the cocultured NTR/T-Reg group (P ¼ 0.004) (Figure 4G). CD41 Helper T-Cell Profile and Cytokine Expression, and Polyfunctional T-Reg Frequency (CD81 T-Reg and IL-171 T-Reg) in the Blood and BAL of TB versus LTBI

Because there are hardly any available human data we interrogated the comparative T-helper cell profiles (Th1, Th2, Th17, and T-Reg) in the blood and BAL of patients with TB versus LTBI control subjects. There was a significantly higher frequency of Th2 (IL-13), Th17 (IL-17), and T-Regs (CD41CD25 1 FOXP31) in the PB compartment (gated on CD31CD41 cells) of active patients with TB compared with LTBI (P ¼ 0.03, P ¼ 0.04, P ¼ 0.04, respectively) (Figure 5A). There was no difference in Th1 (IFN-g) expression (data not shown). By contrast, although CD4 1CD25 1FOXP3 1 T-Regs were significantly increased (P ¼ 0.04) (Figure 5B), no significant differences were found in the expression of IL-13, IL-17, or IFN-g in the BAL compartment of those with TB versus LTBI (the median frequency of IL-17 expression, however, was higher in the TB group). No differences in level of polyfunctional T-Regs, including CD81 T-Reg and IL-171 T-Reg, were observed between TB and LTBI in the blood or BAL compartment (Figure 5C shows the blood compartment; the BAL compartment is not shown because all the same biomarkers were virtually undetectable in the LTBI group). Within the TB Subgroup CD41, Th Cell Phenotype and Cytokine Responses Are Compartmentalized

To discern the degree of compartmentalization, we interrogated the comparative T-helper and cytokine profiles within the TB group. In BAL, compared with blood, T-Regs and CD41 T cells expressing IFN-g (P ¼ 0.005), tumor necrosis factor-a (P ¼ 0.02), IL-17 (P ¼ 0.02), and IL-22 (P ¼ 0.01) were all significantly increased (Figure 6A). No differences in levels of polyfunctional T-Regs were observed between blood and BAL within the TB group (Figure 6B). The Comparative Th Profile (Th1/Th2/Th17/T-Reg) in BAL and Blood of Patients with TB Is Heterogeneous

The comparative intrapatient and interpatient frequency of T-helper cell subtypes in patients with TB has been poorly studied. As shown in Figure 7, the comparative representation of the various T cell lineages (Th1, Th2, Th17, and T-Regs) is heterogeneous in each patient (i.e., each patient expressed a variable proportion of each T cell lineage). However, when comparing the intergroup summation data, the expression of T-Regs and Th1-, Th2-, and Th17-like cytokines tended to be increased in BAL compared with the blood compartment.

DISCUSSION Our data indicate that the alveolar lung compartment, compared with the blood compartment, in patients with TB is relatively enriched for T-Regs. Furthermore, in vitro coculture of T-Regs with monocyte-derived macrophages and AMs infected with H37Rv decreased their ability to restrict the growth of M.tb. We also showed that within the TB group responses were notably compartmentalized and that the within-person Th cell profiles (Th1 vs. Th2 vs. Th17 vs. T-Reg) in blood and BAL were heterogeneous. There are hardly any data about T-Regs at the site of disease (i.e., in the human lung), the dominant form of TB worldwide. Recently, Herzmann and colleagues (13) demonstrated an increased frequency of T-Regs in the alveolar lavage fluid of persons with presumed LTBI (defined by a positive QuantiFERON Gold-in-Tube test). In the context of active TB, Sharma and coworkers (14) found a higher frequency of T-Regs in alveolar lavage fluid from persons with miliary TB, a relatively uncommon disseminated form of TB. We have confirmed that these observations can be extended to persons with the more common nonmiliary form of pulmonary TB, and that the lung compartment is relatively enriched for T-Regs, which can suppress T-cell proliferative responses. In contrast to Sharma and coworkers (14), however, who found lower levels of activated CD41251 cells in BAL than in PB suggesting failure to home into the lung, we found an increase in activated (CD41251) cells in BAL. Given that these T-Regs do not proliferate but have suppressive properties suggests that they are likely to be natural and not inducible T-Regs. Collectively, these data indicate that the lung compartment is relatively enriched for T-Regs that have T-cell suppressive properties. Interestingly, in the alveolar compartment of patients with TB heparin-binding hemagglutinin antigen-driven IFN-g responses were not attenuated by T-Regs as they were in PB (2), suggesting that attenuation of effector responses may be compartment-specific. In humans T-Regs isolated from PB have been shown to attenuate TB antigen-specific Th1 PB T-cell responses (15). However, the impact of T-Regs on mycobacterial stasis in humans has not been investigated. Data from animals are conflicting. Murine T-Regs delayed the priming of T cells in the mediastinal lymph nodes (10), a combination of soluble transforming growth factor-b receptor, cyclooxygenase-2 inhibitor, and antimicrobials resulted in accelerated mycobacterial clearance from the lung (16), and T-Regs blunted the mycobactericidal effect of levofloxacin (17). By contrast, other murine data show that although T-Regs attenuate antigen-specific cytokine responses they had no effect on pathogen load (18). Our data support the hypothesis that in humans, rather than being an epiphenomenon or attenuating cytokine production in response to chronic inflammation, T-Regs subvert antimycobacterial immunity and a sterilizing host immune response. Historically, demonstrating significant mycobacterial kill using in vitro human cell culture models has been challenging. However, we have consistently been able to show mycobacterial stasis in a model where PBMC are preprimed


1253

Figure 3. Mycobacterial growth restriction (stasis) assay using H37Rv-infected blood monocyte-derived macrophages (MDM) either cocultured with purified protein derivative preprimed peripheral blood mononuclear cells (PBMC), purified protein derivative preprimed PBMCs that were depleted of regulatory T cells (non–T-Regs), or PBMCs cocultured with blood-derived CD41CD25hiCD127lo (T-Regs) at different ratios. The median values expressed as CFU are shown in the left panels and the % reduction in CFU relative to the MDM reference control are shown in the right panels. (A and B) Data in those with latent tuberculosis (TB) infection (n ¼ 3). (C and D) Data in those with active TB (n ¼ 4). (E and F) Combined data of both groups (latent TB infection and active TB; n ¼ 7). (G) Patient-specific line plots summarizing the impact of T-Reg on the % CFU reduction (n ¼ 7). *P values of less than 0.05 were considered significant (Wilcoxon matched-paired test) and error bars represent the interquartile range.

with PPD for 6 days to generate effector cells (19). When these cells are cocultured with infected monocyte-derived macrophages or AMs, we were able to demonstrate significant mycobacterial containment (stasis/kill), but coaddition of T-Regs decreased their ability to restrict the growth of M.tb. Further studies are required to determine the mechanism by which T-Regs subverts the containment of mycobacterial growth. The first logical step is to determine

whether this is by a cell-to-cell contact mechanism or by some humoral factor (e.g., IL-10, transforming growth factor-b, granzyme B, galectin 10, and/or inducible T-cell costimulator) (20, 21). The T-Regs isolated in this study were defined by the CD127 low phenotype (CD127dim/2). CD127dim/2 correlates well with FoxP3 healthy volunteers (22, 23) and we show that this

1254


VOL 187

2013

Figure 4. Mycobacterial stasis assay using H37Rv-infected alveolar macrophages alone or cocultured with purified protein derivative preprimed CD41CD252 T cells (non–T-Regs), blood-derived CD41CD25hiCD127lo T-Regs alone, or a combination of non–T-Regs and T-Regs using a 1:1 ratio. The median values expressed as CFU are shown in the left panel and the % reduction in CFU relative to the monocyte-derived macrophages reference control are shown in the right panel. (A and B) Data in those with latent tuberculosis (TB) infection (n ¼ 4). (C and D) Data in those with active TB (n ¼ 5). (E and F) Combined data of both groups (latent TB infection and active TB; n ¼ 9). (G) Patient-specific line plots summarizing the impact of T-Reg on the % CFU reduction (n ¼ 9). *P values of less than 0.05 were considered significant (Wilcoxon matched-paired test) and error bars represent the interquartile range.


1255

Figure 5. Frequency of purified protein derivative–stimulated CD41 T cells expressing Th2 (IL-13), Th17 (IL-17), and CD25/FOXP3 (regulatory T cells [T-Regs]) in the (A) peripheral blood of patients with tuberculosis (TB) (n ¼ 20) versus latent TB infection (LTBI) (n ¼ 19), and in the (B) bronchoalveolar lavage (BAL) fluid of patients with TB (n ¼ 6) and LTBI (n ¼ 5). Comparisons for IFN-g, tumor necrosis factor-a, IL-17, IL-22, and IL-10 were not significant in blood or BAL (data not shown). (C) Frequency of polyfunctional T-Regs in TB versus LTBI using peripheral blood. The frequency of polyfunctional T-Regs cells in BAL is not shown because they were virtually undetectable in this compartment.

relationship holds true in the context of TB, and these cells have suppressive capability. There are no data about the frequency of CD81 T-Regs (24) and the recently described IL-17–producing T-Regs in TB (25). However, our data suggest that these cells

are unlikely to be involved in the pathogenesis of TB. To correct for person-specific differences in absolute T-Reg number caused by several factors, including sex and age, we expressed quantified T-Regs as a percentage of the CD4 T cells.

1256


VOL 187

2013

Figure 6. (A) Frequency of purified protein derivative–stimulated CD41 cells expressing IFN-g, tumor necrosis factor (TNF)-a, IL-17, IL-22, and CD25/FOXP3 (regulatory T cells [T-Regs]) in bronchoalveolar lavage (BAL) fluid (n ¼ 7) and blood (n ¼ 19) in patients with active tuberculosis. Comparison for IL-10 and IL-13 (data not shown) were not significant. (B) Frequency of polyfunctional T-Regs in blood versus BAL of patients with tuberculosis.

What insights do our data provide about cytokine responses in BAL fluid versus PB? To support the hypothesis that Th2-like cytokines may subvert protective Th1 host immune responses (26, 27) we showed that IL-13 levels were relatively raised in TB versus LTBI, albeit only in the blood compartment. There are limited data about IL-17 in the alveolar compartment and, to the best of our knowledge, this is the first report evaluating intracellular IL-17 protein expression in alveolar CD4 T cells from patients with active TB. Matthews and colleagues (28) found that IL-17 was not up-regulated in PPD-driven alveolar CD4 T cells from healthy control subjects with presumed LTBI, and Scriba and coworkers (29) found that IL-17 protein was not detectable in alveolar lavage fluid from patients with TB (suggesting that IL-17 may have been suppressed by Th1 cells or this may have been caused by a dilutional effect of the instilled BAL fluid). We have previously shown that IL-17 mRNA levels were up-regulated in the alveolar lavage cells of patients with TB from the United Kingdom compared with healthy control subjects (30). Here we show that IL-17 expression in CD4 T cells was up-regulated in the PB of patients with TB compared with those with presumed LTBI; although there was a similar trend in BAL fluid this was not significant and may have been caused by the small sample size. Collectively, our data suggest a possible role for these cells in the pathogenesis of pulmonary TB.

There are several limitations of our work. The term “restriction of mycobacterial growth” was defined as antimycobacterial activity caused by killing and/or driving a state of nonreplication, and we were not able to distinguish between these processes. Plated mycobacterial cultures were only incubated for approximately 30 days because most were heavily overgrown by then with M.tb (thus precluding accurate counting), but it is possible that further growth could have occurred if organisms were in a relatively nonreplicating state. However, our extensive optimization experiments failed to detect growth even at 60 days. We cocultured effector cells with T-Regs obtained from the PB and not the alveolar lavage compartment. However, only the former was logistically and technically feasible given the low absolute numbers of T-Regs in the BAL fluid, which comprises more than 90% AMs (only a total z100,000 T-Regs could be isolated from z100 ml of alveolar lavage fluid). Thus, our conclusions are limited within this context. How biologically meaningful is an approximately 50% reduction in mycobacterial stasis after coculture for approximately 24 hours? The threshold for mycobacterial stasis that is biologically meaningful in an in vitro culture model is controversial and remains unclear, but we believe that an almost 50% reduction over a period of 24 hours is significant given the relatively short timeframe and that M.tb has a doubling time of 18–20 hours. We elected to evaluate mycobacterial kill after a period of only 18 hours to minimize the


1257

Figure 7. Comparative paired blood and bronchoalveolar lavage (BAL)-derived Th cell profiles for patients with active tuberculosis (n ¼ 6) expressed as the percentage of CD41 T cells expressing IFN-g (Th1), IL-13 (Th2), IL-17 (Th17), and % CD4 T cells that were CD251Foxp3 1 (regulatory T cell [T-Regs]) and negative for all markers.

confounding effect of mycobacterial proliferation in vitro, particularly with respect to mycobacteria not taken up by the macrophages (although we washed infected cells after z18 h with an z30–40% uptake). We are unsure whether the ratios of T-Regs to effecter cells in our experimental protocols are representative from a physiologic perspective. However, our preliminary unpublished data from histologic tissue sections from explanted human lungs from drug-resistant patients with TB showed a high frequency of Foxp31 T-Regs around granulomas with a ratio of approximately one T-Reg to four effector T cells, a ratio (1:4) at which mitigation of mycobacterial growth restriction was still seen in our experimental models (Figure 3). Thus, further studies are required to evaluate this compartment using explanted lungs from patients with active TB. In

some subgroups sample sizes were relatively small; however, the necessity to obtain BAL from relatively ill patients, difficulty in recruiting consenting but nonsmoking patients with TB, the complexity of the experiments, the inability to obtain high enough T-Reg numbers because of the limited amount of sample that could be taken for ethical reasons, and our limited resources precluded obtaining larger group-specific sample sizes. That it took almost 4 years to carry out the work reflects the difficulty in performing such studies in humans. In conclusion, we show that the alveolar lavage compartment in patients with TB, relative to the blood compartment, is enriched for T-Regs, and in vitro T-Regs subvert the containment of mycobacterial growth in alveolar and blood-derived

1258


macrophages infected with pathogenic strains of M.tb. Collectively, these data support the hypothesis that T-Regs subvert antimycobacterial immunity and likely has a deleterious effect in the context of active TB. These data have implications for novel immunotherapeutic interventions for TB (16, 17, 31) and further studies are now required to clarify the mechanisms by which T-Regs exert their effects in this context. Author disclosures are available with the text of this article at www.atsjournals.org. Acknowledgment: The authors thank Graham Rook, Arne Akbar, and Michael Ehrenstein from University College London for their mentorship and helpful suggestions when this work was initiated.

References 1. World Health Organization. WHO global tuberculosis report 2011 [accessed 2012 Jul 17]. Available from: http://whqlibdoc.who.int/publications/2011/ 9789241564380_eng.pdf 2. Hougardy JM, Place S, Hildebrand M, Drowart A, Debrie AS, Locht C, Mascart F. Regulatory T cells depress immune responses to protective antigens in active tuberculosis. Am J Respir Crit Care Med 2007;176: 409–416. 3. Ribeiro-Rodrigues R, Resende Co T, Johnson JL, Ribeiro F, Palaci M, Sá RT, Maciel EL, Pereira Lima FE, Dettoni V, Toossi Z, et al. Sputum cytokine levels in patients with pulmonary tuberculosis as early markers of mycobacterial clearance. Clin Diagn Lab Immunol 2002;9:818–823. 4. Guyot-Revol V, Innes JA, Hackforth S, Hinks T, Lalvani A. Regulatory T cells are expanded in blood and disease sites in patients with tuberculosis. Am J Respir Crit Care Med 2006;173:803–810. 5. Kursar M, Koch M, Mittrücker HW, Nouailles G, Bonhagen K, Kamradt T, Kaufmann SH. Cutting edge: regulatory T cells prevent efficient clearance of Mycobacterium tuberculosis. J Immunol 2007;178:2661–2665. 6. Sakaguchi S, Wing K, Onishi Y, Prieto-Martin P, Yamaguchi T. Regulatory T cells: how do they suppress immune responses? Int Immunol 2009;21: 1105–1111. 7. Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity 2005;22:329–341. 8. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development by the transcription factor Foxp3. Science 2003;299:1057–1061. 9. Qin XJ, Shi HZ, Liang QL, Huang LY, Yang HB. CD41CD251 regulatory T lymphocytes in tuberculous pleural effusion. Chin Med J (Engl) 2008;121:581–586. 10. Shafiani S, Tucker-Heard G, Kariyone A, Takatsu K, Urdahl KB. Pathogenspecific regulatory T cells delay the arrival of effector T cells in the lung during early tuberculosis. J Exp Med 2010;207:1409–1420. 11. Li L, Lao SH, Wu CY. Increased frequency of CD4(1)CD25(high) Treg cells inhibit BCG-specific induction of IFN-gamma by CD4(1) T cells from TB patients. Tuberculosis (Edinb) 2007;87:526–534. 12. Wergeland I, Assmus J, Dyrhol-Riise AM. T regulatory cells and immune activation in Mycobacterium tuberculosis infection and the effect of preventive therapy. Scand J Immunol 2011;73:234–242. 13. Herzmann C, Ernst M, Ehlers S, Stenger S, Maertzdorf J, Sotgiu G, Lange C. Increased frequencies of pulmonary regulatory T-cells in latent Mycobacterium tuberculosis infection. Eur Respir J 2012;40: 1450–1457. 14. Sharma PK, Saha PK, Singh A, Sharma SK, Ghosh B, Mitra DK. FoxP31 regulatory T cells suppress effector T-cell function at pathologic site in miliary tuberculosis. Am J Respir Crit Care Med 2009;179:1061–1070.

VOL 187

2013

15. Chen X, Zhou B, Li M, Deng Q, Wu X, Le X, Wu C, Larmonier N, Zhang W, Zhang H, et al. CD4(1)CD25(1)FoxP3(1) regulatory T cells suppress Mycobacterium tuberculosis immunity in patients with active disease. Clin Immunol 2007;123:50–59. 16. Hernandez-Pando R, De La Luz Streber M, Orozco H, Arriaga K, Pavon L, Al-Nakhli SA, Rook GA. The effects of androstenediol and dehydroepiandrosterone on the course and cytokine profile of tuberculosis in BALB/c mice. Immunology 1998;95:234–241. 17. Tong XD, Li MZ, Zhou BP, Chen XC, Peng YZ, Yue XH, Gou JZ, Tang ZJ. The effect of CD41 CD251 regulatory T cell inactivation combined with levofloxacin on murine tuberculosis [in Chinese]. Zhonghua Jie He He Hu Xi Za Zhi 2009;32:685–688. 18. Quinn KM, McHugh RS, Rich FJ, Goldsack LM, de Lisle GW, Buddle BM, Delahunt B, Kirman JR. Inactivation of CD41 CD251 regulatory T cells during early mycobacterial infection increases cytokine production but does not affect pathogen load. Immunol Cell Biol 2006;84: 467–474. 19. Semple PL, Watkins M, Davids V, Krensky AM, Hanekom WA, Kaplan G, Ress S. Induction of granulysin and perforin cytolytic mediator expression in 10-week-old infants vaccinated with BCG at birth. Clin Dev Immunol 2011;2011:438–463. 20. Collison LW, Pillai MR, Chaturvedi V, Vignali DA. Regulatory T cell suppression is potentiated by target T cells in a cell contact, IL-35and IL-10-dependent manner. J Immunol 2009;182:6121–6128. 21. Flores-Borja F, Jury EC, Mauri C, Ehrenstein MR. Defects in CTLA-4 are associated with abnormal regulatory T cell function in rheumatoid arthritis. Proc Natl Acad Sci USA 2008;105:19396–19401. 22. Liu W, Putnam AL, Xu-Yu Z, Szot GL, Lee MR, Zhu S, Gottlieb PA, Kapranov P, Gingeras TR, Fazekas de St Groth B, et al. CD127 expression inversely correlates with FoxP3 and suppressive function of human CD41 T reg cells. J Exp Med 2006;203:1701–1711. 23. Yu N, Li X, Song W, Li D, Yu D, Zeng X, Li M, Leng X, Li X. CD4(1) CD25 (1)CD127 (low/-) T cells: a more specific Treg population in human peripheral blood. Inflammation 2012;35:1773–1780. 24. Suzuki M, Jagger AL, Konya C, Shimojima Y, Pryshchep S, Goronzy JJ, Weyand CM. CD81CD45RA1CCR71FOXP31 T cells with immunosuppressive properties: a novel subset of inducible human regulatory T cells. J Immunol 2012;189:2118–2130. 25. Ayyoub M, Deknuydt F, Raimbaud I, Dousset C, Leveque L, Bioley G, Valmori D. Human memory FOXP31 Tregs secrete IL-17 ex vivo and constitutively express the T(H)17 lineage-specific transcription factor RORgamma t. Proc Natl Acad Sci USA 2009;106:8635–8640. 26. Rook GA, Dheda K, Zumla A. Immune responses to tuberculosis in developing countries: implications for new vaccines. Nat Rev Immunol 2005;5:661–667. 27. Rook GA, Dheda K, Zumla A. Do successful tuberculosis vaccines need to be immunoregulatory rather than merely Th1-boosting? Vaccine 2005;23:2115–2120. 28. Matthews K, Wilkinson KA, Kalsdorf B, Roberts T, Diacon A, Walzl G, Wolske J, Ntsekhe M, Syed F, Russell J, et al. Predominance of interleukin-22 over interleukin-17 at the site of disease in human tuberculosis. Tuberculosis (Edinb) 2011;91:587–593. 29. Scriba TJ, Kalsdorf B, Abrahams DA, Isaacs F, Hofmeister J, Black G, Hassan HY, Wilkinson RJ, Walzl G, Gelderbloem SJ, et al. Distinct, specific IL-17and IL-22-producing CD41 T cell subsets contribute to the human antimycobacterial immune response. J Immunol 2008;180:1962–1970. 30. Dheda K, Chang JS, Lala S, Huggett JF, Zumla A, Rook GA. Gene expression of IL17 and IL23 in the lungs of patients with active tuberculosis. Thorax 2008;63:566–568. 31. Taams LS, Palmer DB, Akbar AN, Robinson DS, Brown Z, Hawrylowicz CM. Regulatory T cells in human disease and their potential for therapeutic manipulation. Immunology 2006;118:1–9.