Tuberculosis (TB) remains a global health problem. The solution involves development of an effective vaccine, but has been limited by incomplete understanding of what constitutes protective immunity during natural infection with Mycobacterium tuberculosis. In this study, M. tuberculosis-specific responses following an overnight whole-blood assay were assessed by intracellular cytokine staining and luminex, and compared between TB cases and exposed household contacts. TB cases had significantly higher levels of IFN-γ+TNF-α+IL-2+CD4+T cells compared with contacts. TB cases also had a significantly higher proportion of cells single-positive for TNF-α, but lower proportion of cells producing IL-2 alone and these differences were seen for both CD4+and CD8+ T cells. Cytokine profiles from culture supernatants were significantly biased toward a Th1 phenotype (IFN-γ and IL-12(p40)) together with a complete abrogation of IL-17 secretion in TB cases. Our data indicate that despite a robust response to TB antigens in active TB disease, changes in the pattern of cytokine production between TB infection and disease clearly contribute to disease progression.
It is estimated that 1.6 million deaths resulted from tuberculosis (TB) in 2005 with both the highest number of deaths and the highest mortality per capita in the African region 1. The global incidence of TB is rising by approximately 1% per annum despite the widespread coverage of BCG vaccination, which is unable to protect against reactivation of latent TB as seen with concomitant HIV infection. In order to develop new therapeutic strategies, it is important to determine what constitutes protective immunity during natural infection with Mycobacterium tuberculosis (MTb).
It is well established that Th1 cytokines, in particular IFN-γ and TNF-α, are important in the control of MTb infection 2–4. However, it is becoming clear that IFN-γ alone is not sufficient to confer protection against TB 5–7, thus warranting the assessment of the role of other cytokines in MTb immunity. For example, IL-18 and IL-12(p40) correlate with IFN-γ production and are associated with protection against TB 8, 9, whereas lower mRNA expression of IL-8 and IL-12(p40) and high FOXP3 expression could distinguish TB cases from those individuals latently infected with MTb 10. Recently IL-17, which is produced by Th17 cells, has also been associated with protection against TB 11, 12.
Most studies assessing possible protective cytokines to MTb infection have measured only single cytokines, or multiple cytokines in the context of post-vaccination responses 13–15 or in children 14, 16. Given that both CD4+ and CD8+ T lymphocytes contribute to protection against TB 17, we assessed multiple cytokine production from both subsets in response to MTb antigens in 36 adults with active TB disease and compared these with responses from 36 skin-test-positive (latently infected) and 34 skin-test-negative household contacts. Surprisingly, and in contrast to what has been assumed to be “protective” against disease progression in HIV, we found a significantly higher proportion of IFN-γ+IL-2+TNF-α+CD4+ T cells in subjects with active TB disease compared with their household contacts.
Overnight stimulation with MTb-specific antigens induces distinct cytokine patterns
Whole, fresh blood was stimulated overnight with αCD3/CD28 (positive control), 6 kDa early secreted antigenic target/culture filtrate protein-10 fusion protein (EC) or purified protein derivative-tuberculin (PPD-T) and assessed for intracellular cytokine content. In subjects that responded (both cases and contacts), we consistently found a distinct cluster of cells within the CD4+ T-cell subset, double-positive for both TNF-α and IFN-γ after stimulation with EC or PPD (Fig. 1A) and these were generally IL-2+as well (data not shown). Surprisingly, while the majority of responses were from the CD4+T cells (median [interquartile range (IQR)]%CD4+TNF-α+ cells=0.5 [0.2–1.5]), we saw a broad range of responses from the CD8+subset as well (median [IQR]%CD8+TNF-α+ cells=0.3 [0.03–2.8] (Fig. 1A and B).
Increased proportion of cytokine-producing cells in TB cases
Blood was stimulated overnight with αCD3/CD28 to determine if T cells from confirmed TB cases were impaired in their ability to respond to non-specific stimulation compared with both tuberculin skin tests (TST+) (latently infected) or TST− (uninfected) household contacts. Results are presented as the proportion of IFN-γ or TNF-α that was produced by each cell population (with background subtracted). The proportion of CD4+IFN-γ+ cells was significantly higher in TST+ contacts compared with TST− contacts (p=0.006; Fig. 2A) but not between TB cases and TST+ or TST− contacts. However, there was a significant increase in the proportion of CD8+TNF-α+ cells in TB cases compared with both TST+and TST− contacts (p=0.018 and 0.031 respectively; Fig. 2B). No difference in the proportion of CD4+TNF-α+ cells was seen between the groups.
The proportion of CD4+IFN-γ and CD4+TNF-α cells was markedly increased in cases compared with TST− contacts following stimulation with PPD (p=0.005 and 0.025 respectively; Fig. 2C and D). Similarly, following stimulation with EC, TB cases and TST+contacts had a significantly higher proportion of CD4+IFN-γ+ cells compared with TST− contacts (p=0.04 and 0.03 respectively; Fig. 2E). TNF-α+ cells followed a similar pattern with the proportion of CD4+TNF-α+ cells being significantly higher in TB cases and TST+ contacts compared with TST− contacts (p=0.002 for both; Fig. 2F). Interestingly, both PPD and EC induced CD8 responses, though these were of lesser frequency to that seen with CD4+ T cells in regards to IFN-γ production and there were no differences seen between the three study groups.
When the data were expressed as the proportion of cells producing any of the three cytokines (TNF-α, IFN-γ, IL-2), a similar distinction between TB cases and contacts was seen (Table 1). Following CD3/CD28 stimulation, TB cases had a significantly higher proportion of CD8+ T cells responding compared with both TST+and TST− contacts (p<0.0001 and p=0.0002 respectively; Table 1) with no difference seen for CD4+ T cells. Both PPD and EC stimulation induced significantly higher proportions of responding cells from both CD4+and CD8+ T cells in TB cases compared with their household contacts (p≤0.01 for all; Table 1). EC stimulation resulted in the best discrimination between cases and TST+ contacts for both CD4+ and CD8+ total cytokine production (p=0.0056 and 0.0153, respectively; Table 1).
Table 1. The proportion of CD4+ or CD8+ cells positive for any of the three cytokines (IFN-γ, IL-2 or TNF-α) following stimulation with various antigensa)
a) Data expressed as median [IQR]+ cells per subset. *p≤0.01 compared with cases; Kruskal–Wallis ANOVA with Dunn's post-test comparisons.
TB cases show a more diverse cytokine profile than household contacts
Many studies suggest that cells producing IFN-γ, TNF-α and IL-2 concurrently are associated with good clinical outcome, at least with HIV infection 18, 19; thus we analysed CD4+and CD8+ T cells for different combinations of IL-2, TNF-α and IFN-γ production in TB cases and compared these with their Mycobacterium-exposed household contacts. Following stimulation with αCD3/CD28, the proportion of cells producing the three cytokines simultaneously was similar between cases and contacts (Fig. 3A, pie graphs). However, TB cases had a significantly higher proportion of TNF-α+ cells within both the CD4+and CD8+subsets (p=0.01 for both; Fig. 3A).
There were a number of differences between cases and contacts following stimulation with PPD (Fig. 3B). We saw a significantly higher proportion of CD4+ T cells producing TNF-α, either alone or in combination with IL-2 and/or IFN-γ in cases compared with contacts (Fig. 3B). In contrast, contacts had a higher proportion of CD4+ cells that produced IL-2 alone (p=0.0001; Fig. 3B). CD8+ T cells from TB cases also had a higher proportion of TNF-α+and lower proportion of IL-2+ cells compared with household contacts (p=0.0001 for both; Fig. 3B) with no difference in production of multiple cytokines seen.
EC induced a pattern of response similar to that observed following PPD stimulation (Fig. 3C). Within the CD4+ subset, there was a significantly higher proportion of TNF-α+IL-2+, TNF-α+IFN-γ+ and TNF-α+IFN-γ+IL-2+ cells in cases compared with contacts (p=0.0001, 0.02 and 0.04, respectively; Fig. 3C). Cases had a higher proportion of cells producing only TNF-α, but a decreased proportion producing IL-2 alone compared with contacts (p<0.0001 for both; Fig. 3C). This pattern was also similar for the CD8+T-cell subset (Fig. 3C).
Taken together, the proportion of cells simultaneously producing IFN-γ, TNF-α and IL-2 in TB cases and contacts were similar following non-specific stimulation via the TCR (αCD3/CD28), but were significantly higher in cases than contacts following stimulation with MTb antigens.
TB cases show a Th1 bias and loss of IL-17 production in their secreted cytokine profile
Culture supernatants obtained after overnight stimulation of PBMC with PPD and EC were assessed for secreted cytokines using a 7-plex bead array that measured IL-10, IL-12(p40), IL-13, IL-17, IL-18, TNF-α and IFN-γ (Fig. 4). Following PPD stimulation, TB cases produced significantly lower levels of IL-13, IL-10 and IL-17 (p=0.0005, 0.01 and 0.008, respectively; Fig. 4) compared with contacts but no difference in IFN-γ or IL-12(p40). EC induced a significantly lower production of IL-18 and also IL-13 and IL-17 in TB cases compared with contacts (p=0.005, 0.002 and 0.01; Fig. 4). In contrast, EC stimulation resulted in significantly higher amounts of IFN-γ and IL-12(p40) production in cases compared with contacts (p=0.03 and 0.001, respectively) (Fig. 4).
This study provides a detailed analysis of the phenotype and frequency of cytokine-producing cells in TB disease in a TB-endemic setting. Importantly, we have shown that the frequencies of MTb-specific lymphocytes that produce multiple cytokines (TNF-α and IFN-γ and/or IL-2) or TNF-α alone are significantly higher in subjects with active TB disease, whereas those that produce IL-2 alone are lower compared with their Mycobacterium-exposed household contacts. This cellular profile was associated with secretion of a predominant Th1-type cytokine response and complete abrogation of IL-17 production in TB cases compared with their household contacts.
While the CD8+ T cells were fully capable of responding to non-specific stimulation through the TCR (αCD3/CD28) with higher TNF-α and IFN-γ production than from the CD4+ T cells, it was convincingly the CD4+ T cells that responded to the MTb-specific stimulation, including the quantity and diversity of cytokines from the responding population. However, in contrast to the perceived notion of MTb responses, we did see production of cytokines from the CD8+ cells in some subjects (mainly TB cases). Whether the quality of responses generated by these cells are equivalent to the CD4+ cells remains to be seen but this may be a potential factor in disease progression.
Data from the HIV-field associate CD4+ and/or CD8+ T cells producing IFN-γ, TNF-α and IL-2 simultaneously, with non-disease progression 19–21. Such a “polyfunctional” cell profile has also been used to define correlates of vaccine-mediated protection against Leishmania 22. Thus it was surprising that we saw a significantly increased proportion of these so-called “polyfunctional” CD4+ T cells in TB cases compared with Mycobacterium-exposed household contacts. Our data suggest that this particular cytokine profile is either not protective against TB or that there are other factors that militate against the efficacy of these cytokines on the effector cells that eliminate MTb.
Strikingly, the frequencies of CD4+and CD8+ T cells that produced IL-2 alone were significantly reduced in TB cases compared with their household contacts, following stimulation with MTb- but not non-MTb-specific antigens (αCD3/CD28). IL-2 is a T-cell growth factor that promotes proliferation and differentiation of antigen-specific T cells in an attempt to respond to pathogenic infections swiftly and efficiently 23. In the face of a higher frequency of MTb-specific cells producing multiple cytokines in TB cases compared with contacts, the selective defect in IL-2-secreting cells might be one of the critical determinants of immune dysfunction during MTb infection. Indeed, a diminished frequency of IL-2-secreting MTb-specific CD4+ T cells in patients with advanced HIV disease that bear the major brunt of TB 24, supports our findings and suggests that a defective IL-2 response to MTb could be playing a major role in reactivation and TB disease progression.
In contrast to IL-2, the proportion of TNF-α+ cells within both the CD4+and CD8+subsets was significantly increased in TB cases compared with contacts (both TST+and TST−). The activities of TNF-α are broad and include both beneficial and detrimental effects 25. Similar to IFN-γ, TNF-α helps recruit cells to the site of infection and promotes anti-microbial activity of macrophages 4, 26. In addition, it can lead to TNF-mediated apoptosis of infected macrophages, thus aiding in the elimination of the pathogen 25. It has been postulated that increasing disease severity in TB is partly due to the release of TNF-R2 by the pathogen, which neutralises the TNF-α 27. This may explain why the significant increase in production of TNF-α following stimulation of cells from TB cases that we observed does not protect against disease progression and should be evaluated further.
Luminex analysis demonstrated that MTb-specific antigens induced a Th1-biased cytokine profile (IFN-γ and IL-12(p40)) with less Th2 (IL-10 and IL-13). Consequently, the ratio of IFN-γ to IL-10 was increased in TB cases compared with contacts, a pro-inflammatory profile that has been associated with TB disease severity 28. Moreover, our data showing the complete abrogation of IL-17 production following stimulation with MTb antigens in TB cases supports recent data where lower frequency of IL-17 producing cells induced by MTb antigens was associated with TB disease progression through a yet-to-be understood mechanism 12.
In conclusion, this study shows that TB disease is associated with an increase in the proportion of T lymphocytes capable of producing TNF-α, either alone or in combination with IFN-γ and/or IL-2, but a paucity of those that produce IL-2 alone, and reduced levels of secreted IL-17 following MTb-specific antigen stimulation. Our data indicate that despite a robust response to TB antigens in active TB disease, changes in the pattern of cytokine production between TB infection and disease clearly contribute to disease progression.
Materials and methods
Thirty-six HIV-negative sputum smear-positive TB cases and 70 of their Mycobacterium-exposed household contacts were consecutively recruited as part of ongoing TB case contact studies at the Medical Research Council (UK) unit in The Gambia. All were ≥15 years of age and ethical approval was obtained from the Gambia Government/Medical Research Council joint Ethics committee. TST (two tuberculin units, PPD RT23, SSI, Denmark) were performed on the contacts in order to further classify them into TST+(36) and TST− (34). Subjects with skin test induration of ≥10 mm diameter were categorised as TST positive. Blood samples were taken from both TB cases and their contacts at recruitment prior to skin testing.
Overnight antigenic stimulation
Stimulation with MTb and control antigens was done using 200 μL of heparinised whole blood per stimulation in a polypropylene tube (BDPharmingen). Each patient had five different stimulation set-ups: negative (media alone), positive (αCD3/CD28 – each at 5 μg/mL), PPD-T (10 μg/mL) and EC (10 μg/mL). Tubes were vortexed, covered and placed at 37°C, 5% CO2 overnight (16 h). Brefeldin A (Sigma) was added after 2 h (10 μg/mL).
Whole-blood intracellular cytokine staining
Following overnight stimulation, 20 μL of previously titrated surface marker cocktail (CD4-PercP, CD8-Pacific Blue and CD3-PECy7; BDPharmingen) was added to each tube. Tubes were then vortexed and incubated for 30 min at room temperature (RT). An aliquot of 2 mL of FACSlysing solution (BDPharmingen) was added, tubes vortexed and incubated for 9 min, at RT in the dark followed by centrifugation for 5 min at 600gmax. An aliquot of 500 μL of 1× FACS Perm2 solution (BDpharmingen) was then added, tubes vortexed and incubated for 20 min at RT in the dark. Following centrifugation, the supernatant was carefully removed and 20 μL of cytokine cocktail was added (IFN-γ-APC, TNF-α-FITC and IL-2-PE). Tubes were again vortexed and incubated for 30 min at RT in the dark, then washed, and cells were resuspended in 1% paraformaldehyde prior to acquisition.
Flow cytometry analysis
At least 200 000 lymphocytes were acquired with a CyAn ADP™ (Beckman Coulter, USA) flow cytometer following gating according to 90° forward and side scatter plots. FACS plots were analysed using FlowJo software (version 6.1.1; Treestar, OR, USA). Combinatorial cytokine data were analysed with PESTLE (version 1.5.4) and SPICE (version 4.1.5) software obtained from M. Roederer (National Institutes of Health, Bethesda, MD, USA). Percent frequencies of the different combinations of IL-2, TNF-α and IFN-γ+ cells following antigenic stimulation were calculated within the total population of CD4+or CD8+ T cells and background subtracted (as determined from the medium alone control). Non-specific background was extremely low when more than one cytokine was examined. A cut-off of 0.01% was used as described previously 13; values below this were set to zero.
Bio-plex assay for multiple cytokines
Overnight stimulation of PBMC
PBMC were separated, counted and plated out at 2×105 cells/well in RPMI+10% human AB serum (Sigma). Cells were stimulated with EC, PPD and a positive and negative control (PHA and culture medium alone, respectively). Following overnight incubation at 37°C, 5% CO2, supernatants were removed and stored at −20°C until needed.
Multi-plex analysis for cytokine production
Culture supernatants were analysed using a Bio-Rad custom made 7-plex kit according to the manufacturer's instructions. Cytokines assessed were IL-10, IL-12(p40), IL-13, IL-17, IL-18, IFN-γ and TNF-α. Following pre-wetting of the filter plate, 50 μL of bead suspension was added to each well and washed twice. Fifty microlitres of samples and standards were then added, plate sealed and shaken for 30 s at 1100 rpm and then incubated for 1 h at 300 rpm. The plate was washed three times, and then 25 μL of pre-diluted detection antibody was added. Following shaking, the plate was incubated for 30 min at 300 rpm in the dark. After washing, 50 μL of 1× streptavidin-PE was added to each well and incubated for 10 min. The plate was again washed and resuspended in 125 μL of the assay buffer, sealed, mixed and immediately read on the Bioplex analyser using Bioplex manager software (version 4.0; Bio-Rad, USA) and a low photomultiplier tube setting. All standards were run in duplicate.
Group medians and distributions for TB cases, TST+ and TST− contacts were compared by the Kruskal–Wallis ANOVA with Dunn's post-test comparison using GraphPad Prism software version 5 (Software MacKiev). Comparison of cytokine profiles was assessed using Student's t-tests available in-built with the SPICE analysis program. Comparison of secreted cytokine profiles of TB cases and contacts following luminex was done with a Mann–Whitney U-test using GraphPad Prism software version 5.
We thank all study participants, Mr. Simon Donkor for data management, field workers for sample collection, the TB Immunology laboratory staff, and the Gambian National Tuberculosis and Leprosy Control programme for continuing collaboration. We are grateful to Professor Sarah Rowland-Jones for critical review of the paper. This work was funded by the Bill and Melinda Gates Foundation and the Medical Research Council (UK), The Gambia.
Conflict of interest: The authors declare no financial or commercial conflict of interest.