Mycobacterium bovis BCG Vaccination Induces Divergent Proinflammatory or Regulatory T Cell Responses in Adults
ABSTRACT
Mycobacterium bovis bacillus Calmette-Guérin (BCG), the only currently available vaccine against tuberculosis, induces variable protection in adults. Immune correlates of protection are lacking, and analyses on cytokine-producing T cell subsets in protected versus unprotected cohorts have yielded inconsistent results. We studied the primary T cell response, both proinflammatory and regulatory T cell responses, induced by BCG vaccination in adults. Twelve healthy adult volunteers who were tuberculin skin test (TST) negative, QuantiFERON test (QFT) negative, and BCG naive were vaccinated with BCG and followed up prospectively. BCG vaccination induced an unexpectedly dichotomous immune response in this small, BCG-naive, young-adult cohort: BCG vaccination induced either gamma interferon-positive (IFN-γ+) interleukin 2-positive (IL-2+) tumor necrosis factor α-positive (TNF-α+) polyfunctional CD4+ T cells concurrent with CD4+ IL-17A+ and CD8+ IFN-γ+ T cells or, in contrast, virtually absent cytokine responses with induction of CD8+ regulatory T cells. Significant induction of polyfunctional CD4+ IFN-γ+ IL-2+ TNF-α+ T cells and IFN-γ production by peripheral blood mononuclear cells (PBMCs) was confined to individuals with strong immunization-induced local skin inflammation and increased serum C-reactive protein (CRP). Conversely, in individuals with mild inflammation, regulatory-like CD8+ T cells were uniquely induced. Thus, BCG vaccination either induced a broad proinflammatory T cell response with local inflammatory reactogenicity or, in contrast, a predominant CD8+ regulatory T cell response with mild local inflammation, poor cytokine induction, and absent polyfunctional CD4+ T cells. Further detailed fine mapping of the heterogeneous host response to BCG vaccination using classical and nonclassical immune markers will enhance our understanding of the mechanisms and determinants that underlie the induction of apparently opposite immune responses and how these impact the ability of BCG to induce protective immunity to TB.
INTRODUCTION
Tuberculosis (TB), caused by Mycobacterium tuberculosis, is the second greatest infectious cause of death worldwide after HIV, accounting for 1.3 million deaths in 2012 (1). The only available vaccine, Mycobacterium bovis bacillus Calmette-Guérin (BCG), protects infants from disseminated forms of TB but has insufficient and inconsistent efficacy in protecting adults from pulmonary TB (1, 2). A vaccine preventing active pulmonary TB, the contagious form of the disease, would greatly impact the epidemic (3), and a better understanding of vaccine-induced mechanisms of protection is essential in developing new surrogate endpoints (4).
Both CD4+ Th1 (gamma interferon-positive [IFN-γ+]) cells and CD8+ T cells are critical for protection against TB (5). Specifically, CD4+ IFN-γ+ interleukin 2-positive (IL-2+) tumor necrosis factor α-positive (TNF-α+) polyfunctional T cells have been proposed as a correlate of vaccine-induced protective immunity in murine infection models (6). In infants, BCG vaccination induced specific cytokine expression in CD4+ and CD8+ T cells (7–9), including IFN-γ+ IL-2+ TNF-α+ polyfunctional CD4+ T cells (10). However, there was no relation between the presence of such cells and the development of TB during follow-up (11).
In adults, BCG vaccination induced CD4+ IFN-γ+ responses (12–14) as well as IFN-γ- and TNF-α-secreting CD8+ T cells with cytotoxic activity (15). However, data on the induction of polyfunctional T cells by BCG vaccination in adults have been conflicting (16, 17). In one report, the induction of polyfunctional CD4+ T cells was similar in magnitude in BCG-vaccinated infants and adults; however, when induction was analyzed as the proportion of polyfunctional versus single-cytokine-producing T cells, the proportion of polyfunctional CD4+ T cells was larger in children than in adults (16). Further, studies on latent (controlled) versus active TB in adults yielded variable results on changes in mono- and triple-cytokine-producing T cell subsets (18, 19), such that it was suggested that polyfunctional T cells are also present in active TB disease and that these cells are not a surrogate marker of protection against TB in humans (19, 20).
Another explanation for the inconsistent protection induced by BCG against TB in adults is the induction of regulatory T cells (Tregs) by mycobacteria, which can dampen proinflammatory responses (21). In that context, we reported that live BCG triggers the specific activation of CD8+ (but not CD4+) Tregs from peripheral blood mononuclear cells (PBMCs) of mycobacterial purified protein derivative (PPD)-responsive adults (22), while others found that BCG vaccination induced CD4+ Tregs in newborns (23) and adults (24). Here, in a small, well-defined cohort of previously BCG-naive adults, we studied the induction of multiple-cytokine-producing and regulatory T cell subsets following BCG vaccination.
MATERIALS AND METHODS
Participants.
Dutch volunteers were recruited via posters in the university library. All volunteers were screened for tuberculosis by anamnesis (history of TB disease or treatment), by a tuberculin skin test (TST; negative, <5 mm), and by the QuantiFERON TB gold in-tube test according to the manufacturer's specifications. Included volunteers (n = 6 males, n = 6 females; median age, 24 years [interquartile range (IQR), 23 to 25 years]; median weight, 70 kg [IQR, 67 to 80 kg]; all Dutch, all Caucasian) had not been vaccinated with BCG at any time prior to entering the trial (anamnestic, presence of scar, or described on a vaccination card), were never treated for TB disease, and had negative TST and QuantiFERON test results. In addition, they did not receive any live vaccination at ≤4 weeks prior to BCG vaccination. Volunteers were excluded who were pregnant or not generally healthy, who had fever or received antibiotic treatment ≤2 weeks prior to enrollment, or who were treated with immune-modulating drugs ≤3 months prior to enrollment; all volunteers tested negative for HIV at screening.
Procedures.
Participants were vaccinated with the live attenuated BCG Danish strain 1331 (Statens Serum Institut, Denmark) by intradermal injection in the upper arm and were followed up prospectively at 2 weeks prior to vaccination, at the day of vaccination, at 1, 3, and 7 days after vaccination, at 4, 8, and 12 weeks after vaccination, and at 1 year after vaccination. During follow-up, the injection site was inspected and photographed, and adverse events were recorded using a standardized case report form. Venous blood samples were collected in heparin-containing vacutainers for whole-blood stimulation assays and for PBMC isolation and cryopreservation according to standard operating procedures. Serum samples were collected from serum tubes after blood coagulation and stored at −80°C.
Calculation of skin inflammation score.
Signs of inflammation by visual inspection of the vaccination site and symptoms recorded in volunteer diaries were documented on standardized case report forms and photographed at 4, 8, and 12 weeks after vaccination. The local reaction was scored by two researchers (M.C.B. and C.P.) independently, with one point per sign of inflammation: redness of ≥1 cm, swelling of ≥1 cm, pus discharge and ulceration, pain, and regional lymph node enlargement (>90% consensus; disagreements were solved by mutual reexamination of case report forms, photographs, and volunteer diaries). The inflammation score was calculated as the cumulative scores of weeks 4, 8, and 12 after vaccination.
C-reactive protein enzyme-linked immunosorbent assay.
The serum samples of all prevaccination and postvaccination visits were thawed, and the C-reactive protein (CRP) concentration was measured using a standardized, highly sensitive, CRP human enzyme-linked immunosorbent assay (ELISA) according to the instructions of the manufacturer (Abnova, Heidelberg, Germany).
Whole-blood live BCG stimulation.
Bacillus Calmette-Guerin (Pasteur) was grown in 7H9 plus ADC, frozen in 25% glycerol, and stored at −80°C. Before use, the bacteria were thawed and washed in phosphate-buffered saline (PBS)–0.05% Tween 80 (Sigma-Aldrich). Then, 1 ml of heparinized blood was added within 1 h of blood collection to Sarstedt microtubes (Sarstedt B.V., Etten-Leur, The Netherlands) containing 0.9 × 10E6 CFU (calculated multiplicity of infection [MOI] of 3) and anti-CD28 and anti-CD49d antibodies as costimulants (1 μg/ml; BD Biosciences, Eerembodegem, Belgium) (25) and immediately incubated at 37°C. Staphylococcal enterotoxin B (SEB) (final concentration, 5 μg/ml; Toxin Technology, Sarasota, FL, USA) and unstimulated samples were used as controls. After 3 h, brefeldin A (10 μg/ml; Sigma-Aldrich, Zwijndrecht, The Netherlands) and monensin (1:1000; BD Biosciences) were added, and samples were transferred to a water bath set at 37°C and programmed to switch off after 12 h. Samples were harvested the next morning using EDTA (2 mM; Sigma-Aldrich), fixed and erythrocyte-lysed using a fluorescence-activated cell sorter (FACS) lysing solution (BD Biosciences), and cryopreserved in fetal calf serum with 10% dimethyl sulfoxide (DMSO) (25).
Cell cultures and BCG infection.
PBMCs were thawed, and cells were counted using the Casy cell counter (Roche, Woerden, The Netherlands). Infections were done at an MOI of 1.5. SEB (final concentration, 2 μg/ml; Toxin Technology) and unstimulated samples were used as controls. The PBMCs were cultured in 24-well plates (2 × 10E6/well) for 6 days in Iscove's modified Dulbecco's medium (Life Technologies-Invitrogen, Bleiswijk, The Netherlands) with 10% pooled human serum. For flow cytometric analysis, the PBMCs were incubated for the last 16 h with αCD3/28 beads (Invitrogen) and brefeldin A (3 μg/ml; Sigma-Aldrich).
Lymphocyte stimulation assays were performed using PBMCs (0.5 × 10E6/well in 48-well plates) and stimulation with 5 μg/ml PPD (Statens Serum Institut, Copenhagen, Denmark) at 37°C and 5% CO2. Phytohemagglutinin (PHA) (final concentration, 2 μg/ml; Remel Europe) and unstimulated samples were used as controls. After 7 days, the supernatants were collected and an IFN-γ ELISA (U-CyTech, Utrecht, The Netherlands) was performed.
Direct IFN-γ enzyme-linked immunosorbent spot (ELISpot) assays were performed at 1 year postvaccination: 250,000 freshly isolated PBMCs were added in an AIMV (synthetic nonhuman serum supplemented) medium (Invitrogen) to 96-well ELISpot plates (Millipore, Bedford, MA, USA) that were precoated with anti-IFN-γ antibody (1-D1K, 5 μg/ml; Mabtech, Stockholm, Sweden) and blocked with AIMV. The PBMCs were stimulated overnight with PHA (2 μg/ml), PPD (5 μg/ml), or an antigen 85B (Ag85B) peptide pool (1 μg/ml) in triplicate (26). For detection, biotinylated anti-IFN-γ antibody (0.5 μg/ml; Mabtech), streptavidin-alkalic phospatase conjugate (1:1,000 dilution in 1% bovine serum albumin [BSA]-PBS; Mabtech IFN-γ ELISpot kit reagent), and a SigmaFast NBT-BCIP substrate (Sigma-Aldrich) were used. Positivity for vaccine take (27) was defined as an increase of ≥100% of the average count in PPD-stimulated wells compared to unstimulated controls and at least 5 spots more than in unstimulated controls (28).
Flow cytometry.
Fixed whole-blood samples were thawed and stained in batches. Surface staining included CD3-Brilliant Violet 570 (clone UCHT1), CD19-Pacific Blue (clone HIB19), and CD56-Brilliant Violet 421 (clone HCD56) (all Biolegend, London, United Kingdom); CD14-Pacific Blue (clone TüK4) and CD4-phycoerythrin (PE)-Texas Red (clone S3.5) (both Life Technologies-Invitrogen); and CD8-HorizonV500 (clone RPA-T8), CD45RA-allophycocyanin (APC)-H7 (clone HI100), and CD62L-Brilliant Violet 605 (clone DREG-56) (all BD Biosciences). For intracellular staining, IL-17A-fluorescein isothiocyanate (clone eBio64DEC17; eBioscience, Hatfield, United Kingdom); IFN-γ Alexa Fluor 700 (clone B27), TNF-α-APC (clone 6401.1111), IL-4-PE (clone 3010.211), and CD69-PeCy5 (clone FN50) (all BD Biosciences); and IL-2-Brilliant Violet 650 (clone MQ1-17H12), IL-10-Pe-Cy7 (clone JES3-9D7), and IL-13-PE (clone JES10-5A2) (all Biolegend) were used in a permeabilization solution (Fix&Perm cell permeabilization kit; An Der Grub BioResearch GMBH, Susteren, The Netherlands).
The stimulated PBMCs were labeled with violet LIVE/DEAD stain (Vivid, Invitrogen) and surface stained with CD3-Brilliant Violet 570 (clone UCHT1), CD19-Pacific Blue (clone HIB19), CD56-Brilliant Violet 421 (clone HCD56), and CD39-PE (clone A1) (all Biolegend, London, United Kingdom); CD14-Pacific Blue (clone TüK4) and CD4-PE-Texas Red (clone S3.5) (both Life Technologies-Invitrogen); and CD8-HorizonV500 (clone RPA-T8; BD Biosciences). Cells were fixed and permeabilized using the Fix&Perm cell permeabilization kit (An Der Grub BioResearch GMBH). For intracellular staining, the following antibodies were used: CC chemokine ligand 4 (CCL4)-fluorescein isothiocyanate (clone 24006; R&D Systems, Abingdon, United Kingdom), Foxp3-Alexa Fluor 700 (clone PCH101; eBioscience), lymphocyte activation gene (LAG)-3-atto 647N (clone 17B4; Enzo Life Sciences, Antwerp, Belgium), and CD25-allophycocyanin-H7 (clone M-A251; BD Biosciences).
Samples were acquired on a BD LSRFortessa using FACSDiva software (version 6.2; BD Biosciences) with compensated parameters. The analysis was performed using FlowJo software (version 9.5.3; Treestar, Ashland, OR, USA), and gates were synchronized per donor for all visits and for both CD4+ and CD8+ T cell subsets, using the comparison with unstimulated samples and SEB as controls.
Statistical analyses.
GraphPad Prism (version 6; GraphPad Software, La Jolla, CA, USA) and SPSS statistical software (version 20; SPSS IBM, Armonk, NY, USA) were used for the Wilcoxon signed-rank tests and Mann-Whitney tests. To correct for paired and unpaired multiple testing, Friedman tests followed by Dunn's multiple comparisons tests and Kruskal-Wallis tests followed by Dunn's multiple comparisons tests, respectively, were used. Only the values significant after multiple-testing correction are demonstrated.
Study approval.
Approval was obtained from the medical ethical committee (registration number P 12.87) of the Leiden University Medical Center, The Netherlands. Each participant signed written informed consent prior to inclusion.
RESULTS
Participants.
Twelve healthy adults (TST negative, QuantiFERON test [QFT] negative) were vaccinated with BCG (n = 6 males, n = 6 females; median age, 24 years [IQR, 23 to 25 years]; all Caucasian). Ex vivo IFN-γ ELISpot assays were performed 1 year postvaccination to determine the mycobacterium-specific immunity (29); all recipients tested positive except for one, who nevertheless demonstrated high BCG-induced IFN-γ responses in flow cytometric analysis (see Fig. S1 in the supplemental material).
Local reactogenicity is variable and corresponds with serum CRP and PPD-induced IFN-γ production.
T cell responses as well as BCG vaccine-induced local skin reactions varied strongly between individuals. At 4, 8, and 12 weeks postvaccination, inflammatory symptoms were scored for redness, swelling, ulceration and pus discharge, pain, and regional lymph node enlargement. A cumulative skin inflammation score was then calculated (further described in the Materials and Methods). Representative photographs of low- and high-inflammation skin lesions are shown in Fig. 1A. Participants were subdivided into a low (n = 6) and a high (n = 6) responder group (referred to as low and high skin inflammation responders, respectively) based on skin inflammation scores using the median cumulative skin inflammation score of 7 as a cutoff (Fig. 1B).
FIG 1
Since serum CRP has been used previously to study vaccine-induced inflammation (30–32), we determined CRP concentrations by a highly sensitive ELISA. Serum CRP concentrations 3 and 7 days postvaccination correlated with the cumulative skin inflammation score (R2 = 0.76 at both visits using nonlinear regression; day 7 results shown in Fig. 1C, left). In high skin inflammation responders, serum CRP was significantly higher 7 days postvaccination than in low skin inflammation responders (P = 0.03; Fig. 1C, right). CRP concentrations at baseline were not different between high and low inflammation responders.
We assessed the in vitro IFN-γ production by ELISA after PPD stimulation of PBMCs for 7 days. The PBMCs of high skin inflammation responders produced more IFN-γ than low skin inflammation responders 4 and 8 weeks postvaccination (P = 0.026 and 0.002, respectively). IFN-γ production was significantly induced at 8 and 12 weeks postvaccination compared to prevaccination in high skin inflammation responders (both P = 0.031) but not in low skin inflammation responders (Fig. 1D). Prior to BCG vaccination, IFN-γ production was not significantly different between high and low skin inflammation responders.
CD4+ IFN-γ+ IL-2+ TNF-α+ T cells and CD8+ IFN-γ+ T cells are induced in high inflammation responders.
Whole-blood samples from prevaccination and at 4, 8, and 12 weeks and 1 year after vaccination were stimulated directly ex vivo with live BCG for 16 h. The gating strategy is shown in Fig. S2A in the supplemental material, and the representative flow cytometric analyses of coexpression of the cytokines in CD4+ T cells with negative and positive controls are shown in Fig. S2B in the supplemental material, compliant with MIATA (minimal information about T cell assays) guidelines (33).
BCG vaccination induced significant IFN-γ expression in CD4+ T cells but only in high inflammation responders (P = 0.031; 4 and 12 weeks postvaccination) and not in low inflammation responders (Fig. 2A, left graphs). Polyfunctional CD4+ IFN-γ+ IL-2+ TNF-α+ T cells were significantly induced in BCG recipients, but a division of the high versus low inflammation responders revealed that the proportion of CD4+ IFN-γ+ IL-2+ TNF-α+ T cells was significantly increased in the high versus low inflammation responders (P = 0.015, 0.048, and 0.041 at 4, 8, and 12 weeks after vaccination, respectively), but it was almost absent in the latter group. Thus, significant induction of CD4+ IFN-γ+ IL-2+ TNF-α+ T cells was confined to high inflammation responders (Fig. 2A, right graphs). The majority of IFN-γ-, TNF-α-, and/or IL-2-expressing CD4+ T cell subsets of high inflammation responders peaked at 8 weeks postvaccination (see Fig. S3A in the supplemental material). Single-, double- and triple-cytokine-producing CD4+ T cell subsets consisted predominantly of effector cells at 4, 8, and 12 weeks postvaccination (see Fig. S3B in the supplemental material).
FIG 2
CD4+ IL-17A+ T cells were induced significantly postvaccination but only in high inflammation responders (Fig. 2B). Similarly, significant induction of IFN-γ expression in CD8+ T cells was confined to high inflammation responders (Fig. 2B). No polyfunctional CD8+ T cell responses could be detected (data not shown). CD4+ T cells did not coexpress IL-17A and IFN-γ (see representative graph in Fig. S2B in the supplemental material). IL-17A expression correlated with IFN-γ expression in CD4+ T cells at 4, 8, and 12 weeks postvaccination (Fig. 2C, upper graphs; R2 of 0.86, 0.94, and 0.83, respectively, using nonlinear regression). IFN-γ expression in CD8+ T cells highly correlated with IFN-γ expression in CD4+ T cells (Fig. 2C, lower graphs; R2 of 0.98, 0.85, and 0.99 at 4, 8, and 12 weeks postvaccination, respectively, using nonlinear regression), thus revealing a broad proinflammatory response induced by BCG vaccination in high inflammation responders. No induction of IL-10 or IL-4/IL-13 expression was observed in CD4+ T cells in low or in high inflammation responders (Fig. 2D).
The association between skin inflammation score and induction of CD4+ cytokine coexpression was further substantiated by segregating vaccinees based on induction versus no induction of polyfunctional CD4+ IFN-γ+ IL-2+ TNF-α+ T cells instead of local inflammation scores. This revealed an increased total skin inflammation score in vaccinees with induction of CD4+ polyfunctional T cells compared to vaccinees with no polyfunctional CD4+ T cell induction (n = 7 cytokine responders versus n = 5 nonresponders, including in the latter group a low skin inflammation responder with a skin inflammation score of 6; P = 0.037, Mann-Whitney test; see also Fig. S4A in the supplemental material). The differences in the CRP concentration 7 days after vaccination did not reach statistical significance in cytokine responders versus nonresponders (see Fig. S4B in the supplemental material; P = 0.078, Mann-Whitney test).
CD8+, but not CD4+, T cells express regulatory markers after live BCG stimulation in vitro.
PBMCs were stimulated for 6 days with live BCG to assess the expression of the regulatory markers CD25, Foxp3, CD39, lymphocyte activation gene-3 (LAG-3), and CCL4 (macrophage inflammatory protein-1β) by FACS analysis (22, 34, 35) (see gating strategy in Fig. S2A in the supplemental material). BCG-induced expression of Treg markers on CD4+ and CD8+ T cells was then compared to that on unstimulated control samples at 8 weeks postvaccination, the peak response (Fig. 3A). Expression of the single Treg markers CD25, Foxp3, CD39, LAG-3, and CCL4 was significantly induced on CD8+ T cells by in vitro live BCG stimulation 8 weeks postvaccination (P < 0.01 for CD25, Foxp3, and CCL4; P < 0.001 for CD39 and LAG-3). In contrast, on CD4+ T cells, only CD25 was significantly induced by BCG stimulation (P < 0.05) (Fig. 3B).
FIG 3
We then analyzed coexpression of Treg markers using Boolean gating with synchronized gates on a per-donor basis for both CD4+ and CD8+ T cells (22). Coexpression of Treg markers was significantly induced on CD8+ but not CD4+ T cells by in vitro live BCG stimulation 8 weeks postvaccination (Fig. 3C; P < 0.001). On CD4+ T cells, no expression of CD25+ Foxp3+ CD39+ was induced in either high or low inflammation responders (see Fig. S5 in the supplemental material).
CD8+ CD25+ CD39+ Foxp3+ T cells increase postvaccination only in low inflammation responders.
We then compared Treg markers on CD8+ T cells post- and prevaccination; Fig. 4A displays a representative flow cytometric analysis of induction of CD25+ CD39+ coexpression and expression of Foxp3 on CD25+ CD39+ CD8+ T cells from a low skin inflammation responder. Compared to prevaccination, BCG vaccination significantly induced CD8+ CD25+ Foxp3+ CD39+ T cells as well as CD8+ CD25+ Foxp3+ CD39+ LAG-3+ CCL4+ T cells, but only in low inflammation responders (Fig. 4B; P = 0.031 at 8 weeks postvaccination; P = 0.031 at 4 and 8 weeks postvaccination, respectively). In contrast, in high inflammation responders, the expression of CD25+ Foxp3+ CD39+ or CD25+ Foxp3+ CD39+ LAG-3+ CCL4+ on CD8+ T cells postvaccination did not differ from the patterns prevaccination (Fig. 4B). Of interest, CD8+ CD25+ Foxp3+ CD39+ T cells and CD8+ CD25+ Foxp3+ CD39+ LAG-3+ CCL4+ T cells were still significantly increased 1 year postvaccination in low inflammation responders (for both, P = 0.031), suggesting that BCG vaccination can induce long-term imprinting of CD8+ Treg phenotypes in individuals with low inflammation following vaccination.
FIG 4
DISCUSSION
In this study, we describe high interindividual variability in T cell cytokine and regulatory responses following BCG vaccination of BCG-naive healthy young adults in a setting where TB is not endemic. The unexpectedly dichotomous T cell response consisted of either concurrent induction of IL-2-, TNF-α-, and IFN-γ-coexpressing polyfunctional CD4+ T cell subsets, CD4+ IL-17A+ T cells, and CD8+ IFN-γ+ T cells in high inflammation responders or an almost absent cytokine response accompanied by the induction of CD8+ regulatory T cells in low inflammation responders. We quantified local reactivity by classical clinical symptoms of inflammation and found that the total skin inflammation score correlated with serum CRP concentration early postvaccination. Significant induction of IFN-γ+ IL-2+ TNF-α+ polyfunctional CD4+ T cells was confined to high inflammation responders, while the induction of regulatory-phenotype CD8+ CD25+ CD39+ Foxp3+ and CD8+ CD25+ Foxp3+ CD39+ LAG-3+ CCL4+ T cells was confined to low inflammation responders.
In theory, this study could have been limited by the description of T cell responses based on the skin inflammation score, since dividing high and low skin inflammation groups using the median as a cutoff dichotomizes the described response. However, the dichotomy was also based on the induction of polyfunctional CD4+ IFN-γ+ IL-2+ TNF-α+ T cells, and this revealed a significantly increased total skin inflammation score in vaccinees with CD4+ polyfunctional T cells compared to vaccinees with no polyfunctional CD4+ T cell induction. Thus, it is unlikely that the described variability in responses is caused by a dichotomized representation, and this further affirms the relation between skin reactivity and cytokine responses. Further, the opposing immune responses and phenotypes were observed within a relatively small cohort. Variability in BCG immunogenicity has been ascribed to differences in preexisting antimycobacterial responses in settings where disease is endemic versus nonendemic (36, 37), the presence of helminth infections (38), variations in the BCG vaccine strain (39, 40), and host genetic factors (41). In addition, timing of sampling and technical variability may influence the detection of cytokines (42, 43). This cohort, though small, was uniform in terms of age, genetic background, BCG vaccine strain (Danish strain 1331), sampling, and testing in a setting not endemic for TB or helminth infections. This excludes the above-mentioned possible confounders and points to an unexpectedly large variation in adult human primary BCG vaccine-induced immune responses.
Importantly, we confirmed vaccine take at 1 year postvaccination by IFN-γ ELISpot, which was positive for both high and low inflammation responders. IFN-γ-ELISpot has been described as the most sensitive assay for detecting long-term vaccine responses (29) and is used in TB vaccine trials to describe the magnitude of vaccine-induced immunity. However, a sole reliance on IFN-γ ELISpot would disregard variability in other assays, thereby not fully capturing possible correlations between variation of the human immune response and vaccine-induced protection. The etiology of this variation remains unknown, but its unraveling could contribute significantly to a better understanding of BCG and related TB vaccine-induced immunity.
The height of the in vitro cytokine response in BCG-vaccinated infants was associated with scarring of the BCG vaccination site but only in response to mycobacterial antigens, not unrelated antigens (44). Also, cell-mediated immunity, as assessed by a leukocyte migration inhibition test, correlated with infant local skin reactivity 8 weeks after BCG vaccination, but not with TST conversion after vaccination (45). The absence of an association between BCG-induced TST conversion and immunity against TB has been confirmed in various populations (46). Here, induction of cytokine responses was confined to recipients with high skin reactivity, suggesting that a simple phenotype like vaccine-induced skin inflammation might be used as a marker of strong proinflammatory T cell induction in adults. The skin inflammation score was associated with the serum CRP concentration 7 days postvaccination; thus, the absence of an increase in CRP early postvaccination might be used as an indicator of absent proinflammatory T cell responses at later time points.
Interference of CD4+ Tregs with effector immunity has been described in active TB (47, 48). Following MVA-85A vaccination, circulating CD4+ CD25+ Foxp3+ T cells were increased in recipients with low antigen 85A-specific IFN-γ responses compared to high IFN-γ responses (28), and MVA-85A-induced CD4+ Tregs inhibited IL-17A production in vitro (49). Interestingly, IL-10-producing CD8+ Tregs were described in TB patients anergic to intradermally injected PPD (50). Thus, Tregs can interfere with inflammatory and specific antigen-induced cytokine responses. We previously reported in vitro activation of CD8+ (but not CD4+) Tregs by live BCG, both phenotypically and functionally, in mycobacterially sensitized but not PPD-unresponsive donors (22, 34, 35). These BCG-activated CD8+ Tregs expressed CD25 and LAG-3 and inhibited Th1 responses through secretion of CCL4 (35); in addition, we reported CD8+ CD39+ Tregs which utilized CD39 to suppress Th1 proliferation (34). Here, we found that CD8+ CD25+ CD39+ Foxp3+ and CD8+ CD25+ Foxp3+ CD39+ LAG-3+ CCL4+ T cells were induced following BCG vaccination. Interestingly, the frequency of CD8+ T cells with these Treg phenotypes was significantly increased only in comparison to that at prevaccination in low inflammation responders with low to absent cytokine responses, suggesting an inverse relation between the induction of CD8+ Tregs and BCG-induced skin inflammation with T cell cytokine production.
In murine leishmaniasis, cytokine-producing polyfunctional T cells were inversely correlated with lesion size after (dermal) challenge (6). In dermal BCG challenge models in humans, vaccination-induced IFN-γ ELISpot responses were inversely correlated with PCR quantification of the BCG load in biopsy specimens of the challenge site (51). The PCR quantification method was suggested as a measure of pathogen clearance, possibly reflecting some degree of protective immunity, which might be used in human TB vaccine trials. Based on the current study, it will also be relevant to assess the presence of proinflammatory versus regulatory T cells in skin vaccine or challenge lesions and to further validate the modulation of skin inflammation and/or pathogen clearance by CD8+ Tregs in relevant models. Of note, in low inflammation responders, CD8+ CD25+ Foxp3+ CD39+ T cells were still significantly increased at 1 year postvaccination, suggesting that BCG vaccination can induce long-term imprinting of a CD8+ Treg phenotype with a significant memory component. Further work is needed to assess their precise longevity.
In conclusion, our results show an unexpectedly dichotomous host response to BCG vaccination in a cohort of BCG-naive adults. It will be important to assess these divergent outcomes in settings where TB is endemic in order to determine the impact of these highly variable outcomes on protective efficacy against TB. The use of classical inflammation markers as nonclassical indicators of vaccine-induced proinflammatory responses might be a simple means to assist in assessing BCG-induced phenotypes, even in small cohorts. Further detailed fine mapping of the heterogeneous host response to BCG vaccination using classical and nonclassical immune markers will enhance our understanding of the mechanisms and determinants that underlie the induction of apparently opposite immune responses and how these impact the ability of BCG to induce protective immunity to TB.
ACKNOWLEDGMENTS
This work was supported by EC FP7 NEWTBVAC (contract HEALTH.F3.2009 241745), EC FP7 ADITEC (contract HEALTH.2011.1.4-4 280873), EC FP7 IDEA (grant agreement 241642), and TBVAC2020 Horizon2020 (contract 643381), and by The Netherlands Organization for Scientific Research (VENI grant 916.86.115), the Gisela Thier Foundation of the Leiden University Medical Center, and the Netherlands Leprosy Foundation.
The text represents the authors' views and does not necessarily represent a position of the Commission, which will not be liable for the use made of such information. The funders had no role in study design, data collection, or analysis, the decision to publish, or preparation of the manuscript.
We thank all the volunteers for participating in this study, and we thank Louis Wilson for providing BCG cultures.
Supplemental Material
File (zcd999095216so1.pdf)
- Download
- 510.98 KB
REFERENCES
1.
World Health Organization. 2013. Global tuberculosis report 2013. World Health Organization, Geneva, Switzerland.
2.
Ottenhoff TH, Kaufmann SH. 2012. Vaccines against tuberculosis: where are we and where do we need to go? PLoS Pathog 8:e1002607.
3.
Abu-Raddad LJ, Sabatelli L, Achterberg JT, Sugimoto JD, Longini IM, Jr, Dye C, Halloran ME. 2009. Epidemiological benefits of more-effective tuberculosis vaccines, drugs, and diagnostics. Proc Natl Acad Sci U S A 106:13980–13985.
4.
Ottenhoff TH, Ellner JJ, Kaufmann SH. 2012. Ten challenges for TB biomarkers. Tuberculosis (Edinb) 92(Suppl 1):S17–S20.
5.
Ottenhoff TH, Lewinsohn DA, Lewinsohn DM. 2008. Human CD4 and CD8 T cell responses to Mycobacterium tuberculosis: antigen specificity, function, implications and applications, p 119–156. In Kaufmann SH, Britton WJ (ed), Handbook of tuberculosis: immunology and cell biology. Wiley-VCH Verlag GmbH & Co, Weinheim, Germany.
6.
Darrah PA, Patel DT, De Luca PM, Lindsay RW, Davey DF, Flynn BJ, Hoff ST, Andersen P, Reed SG, Morris SL, Roederer M, Seder RA. 2007. Multifunctional TH1 cells define a correlate of vaccine-mediated protection against Leishmania major. Nat Med 13:843–850.
7.
Marchant A, Goetghebuer T, Ota MO, Wolfe I, Ceesay SJ, de Groote D, Corrah T, Bennett S, Wheeler J, Huygen K, Aaby P, McAdam KP, Newport MJ. 1999. Newborns develop a Th1-type immune response to Mycobacterium bovis bacillus Calmette-Guerin vaccination. J Immunol 163:2249–2255.
8.
Hussey GD, Watkins ML, Goddard EA, Gottschalk S, Hughes EJ, Iloni K, Kibel MA, Ress SR. 2002. Neonatal mycobacterial-specific cytotoxic T-lymphocyte and cytokine profiles in response to distinct BCG vaccination strategies. Immunology 105:314–324.
9.
Murray RA, Mansoor N, Harbacheuski R, Soler J, Davids V, Soares A, Hawkridge A, Hussey GD, Maecker H, Kaplan G, Hanekom WA. 2006. Bacillus Calmette Guerin vaccination of human newborns induces a specific, functional CD8+ T cell response. J Immunol 177:5647–5651.
10.
Soares AP, Kwong Chung CK, Choice T, Hughes EJ, Jacobs G, van Rensburg EJ, Khomba G, de KM, Lerumo L, Makhethe L, Maneli MH, Pienaar B, Smit E, Tena-Coki NG, van WL, Boom WH, Kaplan G, Scriba TJ, Hanekom WA. 2013. Longitudinal changes in CD4(+) T-cell memory responses induced by BCG vaccination of newborns. J Infect Dis 207:1084–1094.
11.
Kagina BM, Abel B, Scriba TJ, Hughes EJ, Keyser A, Soares A, Gamieldien H, Sidibana M, Hatherill M, Gelderbloem S, Mahomed H, Hawkridge A, Hussey G, Kaplan G, Hanekom WA. 2010. Specific T cell frequency and cytokine expression profile do not correlate with protection against tuberculosis after bacillus Calmette-Guerin vaccination of newborns. Am J Respir Crit Care Med 182:1073–1079.
12.
Fjällbrant H, Ridell M, Larsson LO. 2007. Primary vaccination and revaccination of young adults with BCG: a study using immunological markers. Scand J Infect Dis 39:792–798.
13.
Ravn P, Boesen H, Pedersen BK, Andersen P. 1997. Human T cell responses induced by vaccination with Mycobacterium bovis bacillus Calmette-Guerin. J Immunol 158:1949–1955.
14.
Hoft DF, Kemp EB, Marinaro M, Cruz O, Kiyono H, McGhee JR, Belisle JT, Milligan TW, Miller JP, Belshe RB. 1999. A double-blind, placebo-controlled study of Mycobacterium-specific human immune responses induced by intradermal bacille Calmette-Guerin vaccination. J Lab Clin Med 134:244–252.
15.
Smith SM, Malin AS, Pauline T, Lukey Atkinson SE, Content J, Huygen K, Dockrell HM. 1999. Characterization of human Mycobacterium bovis bacille Calmette-Guerin-reactive CD8+ T cells. Infect Immun 67:5223–5230.
16.
Ritz N, Strach M, Yau C, Dutta B, Tebruegge M, Connell TG, Hanekom WA, Britton WJ, Robins-Browne R, Curtis N. 2012. A comparative analysis of polyfunctional T cells and secreted cytokines induced by bacille Calmette-Guerin immunisation in children and adults. PLoS One 7:e37535.
17.
Smith SG, Lalor MK, Gorak-Stolinska P, Blitz R, Beveridge NE, Worth A, McShane H, Dockrell HM. 2010. Mycobacterium tuberculosis PPD-induced immune biomarkers measurable in vitro following BCG vaccination of United Kingdom adolescents by multiplex bead array and intracellular cytokine staining. BMC Immunol 11:35.
18.
Harari A, Rozot V, Enders FB, Perreau M, Stalder JM, Nicod LP, Cavassini M, Calandra T, Blanchet CL, Jaton K, Faouzi M, Day CL, Hanekom WA, Bart PA, Pantaleo G. 2011. Dominant TNFα+ Mycobacterium tuberculosis-specific CD4+ T cell responses discriminate between latent infection and active disease. Nat Med 17:372–376.
19.
Caccamo N, Guggino G, Joosten SA, Gelsomino G, Di Carlo P, Titone L, Galati D, Bocchino M, Matarese A, Salerno A, Sanduzzi A, Franken WP, Ottenhoff TH, Dieli F. 2010. Multifunctional CD4(+) T cells correlate with active Mycobacterium tuberculosis infection. Eur J Immunol 40:2211–2220.
20.
Prezzemolo T, Guggino G, La Manna MP, Di LD, Dieli F, Caccamo N. 2014. Functional signatures of human CD4 and CD8 T cell responses to Mycobacterium tuberculosis. Front Immunol 5:180.
21.
Joosten SA, Ottenhoff TH. 2008. Human CD4 and CD8 regulatory T cells in infectious diseases and vaccination. Hum Immunol 69:760–770.
22.
Boer MC, van Meijgaarden KE, Joosten SA, Ottenhoff TH. 2014. CD8+ regulatory T cells, and not CD4+ T cells, dominate suppressive phenotype and function after in vitro live Mycobacterium bovis BCG activation of human cells. PLoS One 9:e94192.
23.
Akkoc T, Aydogan M, Yildiz A, Karakoc-Aydiner E, Eifan A, Keles S, Akin M, Kavuncuoglu S, Bahceciler NN, Barlan IB. 2010. Neonatal BCG vaccination induces IL-10 production by CD4+ CD25+ T cells. Pediatr Allergy Immunol 21:1059–1063.
24.
Li L, Qiao D, Zhang X, Liu Z, Wu C. 2011. The immune responses of central and effector memory BCG-specific CD4+ T cells in BCG-vaccinated PPD+ donors were modulated by Treg cells. Immunobiology 216:477–484.
25.
Hanekom WA, Hughes J, Mavinkurve M, Mendillo M, Watkins M, Gamieldien H, Gelderbloem SJ, Sidibana M, Mansoor N, Davids V, Murray RA, Hawkridge A, Haslett PA, Ress S, Hussey GD, Kaplan G. 2004. Novel application of a whole blood intracellular cytokine detection assay to quantitate specific T-cell frequency in field studies. J Immunol Methods 291:185–195.
26.
Smith SG, Joosten SA, Verscheure V, Pathan AA, McShane H, Ottenhoff TH, Dockrell HM, Mascart F. 2009. Identification of major factors influencing ELISpot-based monitoring of cellular responses to antigens from Mycobacterium tuberculosis. PLoS One 4:e7972.
27.
McShane H, Pathan AA, Sander CR, Keating SM, Gilbert SC, Huygen K, Fletcher HA, Hill AV. 2004. Recombinant modified vaccinia virus Ankara expressing antigen 85A boosts BCG-primed and naturally acquired antimycobacterial immunity in humans. Nat Med 10:1240–1244.
28.
Matsumiya M, Stylianou E, Griffiths K, Lang Z, Meyer J, Harris SA, Rowland R, Minassian AM, Pathan AA, Fletcher H, McShane H. 2013. Roles for Treg expansion and HMGB1 signaling through the TLR1-2-6 axis in determining the magnitude of the antigen-specific immune response to MVA85A. PLoS One 8:e67922.
29.
Beveridge NE, Fletcher HA, Hughes J, Pathan AA, Scriba TJ, Minassian A, Sander CR, Whelan KT, Dockrell HM, Hill AV, Hanekom WA, McShane H. 2008. A comparison of IFNgamma detection methods used in tuberculosis vaccine trials. Tuberculosis (Edinb) 88:631–640.
30.
Posthouwer D, Voorbij HA, Grobbee DE, Numans ME, van der Bom JG. 2004. Influenza and pneumococcal vaccination as a model to assess C-reactive protein response to mild inflammation. Vaccine 23:362–365.
31.
Paine NJ, Ring C, Bosch JA, Drayson MT, Veldhuijzen van Zanten JJ. 2013. The time course of the inflammatory response to the Salmonella typhi vaccination. Brain Behav Immun 30:73–79.
32.
van der Beek MT, Visser LG, de Maat MP. 2002. Yellow fever vaccination as a model to study the response to stimulation of the inflammation system. Vascul Pharmacol 39:117–121.
33.
Britten CM, Janetzki S, Butterfield LH, Ferrari G, Gouttefangeas C, Huber C, Kalos M, Levitsky HI, Maecker HT, Melief CJ, O'Donnell-Tormey J, Odunsi K, Old LJ, Ottenhoff TH, Ottensmeier C, Pawelec G, Roederer M, Roep BO, Romero P, van der Burg SH, Walter S, Hoos A, Davis MM. 2012. T cell assays and MIATA: the essential minimum for maximum impact. Immunity 37:1–2.
34.
Boer MC, van Meijgaarden KE, Bastid J, Ottenhoff TH, Joosten SA. 2013. CD39 is involved in mediating suppression by Mycobacterium bovis BCG-activated human CD8(+) CD39(+) regulatory T cells. Eur J Immunol 43:1925–1932.
35.
Joosten SA, van Meijgaarden KE, Savage ND, de Boer T, Triebel F, van der Wal A, de Heer E, Klein MR, Geluk A, Ottenhoff TH. 2007. Identification of a human CD8+ regulatory T cell subset that mediates suppression through the chemokine CC chemokine ligand 4. Proc Natl Acad Sci U S A 104:8029–8034.
36.
Ota MO, Brookes RH, Hill PC, Owiafe PK, Ibanga HB, Donkor S, Awine T, McShane H, Adegbola RA. 2007. The effect of tuberculin skin test and BCG vaccination on the expansion of PPD-specific IFN-gamma producing cells ex vivo. Vaccine 25:8861–8867.
37.
Black GF, Weir RE, Floyd S, Bliss L, Warndorff DK, Crampin AC, Ngwira B, Sichali L, Nazareth B, Blackwell JM, Branson K, Chaguluka SD, Donovan L, Jarman E, King E, Fine PE, Dockrell HM. 2002. BCG-induced increase in interferon gamma response to mycobacterial antigens and efficacy of BCG vaccination in Malawi and the United Kingdom: two randomised controlled studies. Lancet 359:1393–1401.
38.
Elias D, Britton S, Aseffa A, Engers H, Akuffo H. 2008. Poor immunogenicity of BCG in helminth-infected population is associated with increased in vitro TGF-beta production. Vaccine 26:3897–3902.
39.
Comstock GW. 1994. Efficacy of BCG vaccine. JAMA 272:766.
40.
Ritz N, Dutta B, Donath S, Casalaz D, Connell TG, Tebruegge M, Robins-Browne R, Hanekom WA, Britton WJ, Curtis N. 2012. The influence of bacille Calmette-Guerin vaccine strain on the immune response against tuberculosis: a randomized trial. Am J Respir Crit Care Med 185:213–222.
41.
Ottenhoff TH, Verreck FA, Lichtenauer-Kaligis EG, Hoeve MA, Sanal O, van Dissel JT. 2002. Genetics, cytokines and human infectious disease: lessons from weakly pathogenic mycobacteria and salmonellae. Nat Genet 32:97–105.
42.
Kaveh DA, Whelan AO, Hogarth PJ. 2012. The duration of antigen-stimulation significantly alters the diversity of multifunctional CD4 T cells measured by intracellular cytokine staining. PLoS One 7:e38926.
43.
Haining WN. 2012. Travels in time: assessing the functional complexity of T cells. Proc Natl Acad Sci U S A 109:1359–1360.
44.
Anderson EJ, Webb EL, Mawa PA, Kizza M, Lyadda N, Nampijja M, Elliott AM. 2012. The influence of BCG vaccine strain on mycobacteria-specific and nonspecific immune responses in a prospective cohort of infants in Uganda. Vaccine 30:2083–2089.
45.
Faridi M, Kaur S, Krishnamurthy S, Kumari P. 2009. Tuberculin conversion and leukocyte migration inhibition test after BCG vaccination in newborn infants. Hum Vaccin 5:690–695.
46.
Comstock GW. 1988. Identification of an effective vaccine against tuberculosis. Am Rev Respir Dis 138:479–480.
47.
Hougardy JM, Place S, Hildebrand M, Drowart A, Debrie AS, Locht C, Mascart F. 2007. Regulatory T cells depress immune responses to protective antigens in active tuberculosis. Am J Respir Crit Care Med 176:409–416.
48.
Chiacchio T, Casetti R, Butera O, Vanini V, Carrara S, Girardi E, di Mitri D, Battistini L, Martini F, Borsellino G, Goletti D. 2009. Characterization of regulatory T cells identified as CD4(+)CD25(high)CD39(+) in patients with active tuberculosis. Clin Exp Immunol 156:463–470.
49.
de Cassan SC, Pathan AA, Sander CR, Minassian A, Rowland R, Hill AV, McShane H, Fletcher HA. 2010. Investigating the induction of vaccine-induced Th17 and regulatory T cells in healthy, Mycobacterium bovis BCG-immunized adults vaccinated with a new tuberculosis vaccine, MVA85A. Clin Vaccine Immunol 17:1066–1073.
50.
Ranjbar S, Ly N, Thim S, Reynes JM, Goldfeld AE. 2004. Mycobacterium tuberculosis recall antigens suppress HIV-1 replication in anergic donor cells via CD8+ T cell expansion and increased IL-10 levels. J Immunol 172:1953–1959.
51.
Harris SA, Meyer J, Satti I, Marsay L, Poulton ID, Tanner R, Minassian AM, Fletcher HA, McShane H. 2014. Evaluation of a human BCG challenge model to assess antimycobacterial immunity induced by BCG and a candidate tuberculosis vaccine, MVA85A, alone and in combination. J Infect Dis 209:1259–1268.
Information & Contributors
Information
Published In
Clinical and Vaccine Immunology
Volume 22 • Number 7 • July 2015
Pages: 778 - 788
Editor: D. W. Pascual
PubMed: 25947145
Copyright
© 2015 American Society for Microbiology.
History
Received: 18 March 2015
Returned for modification: 7 April 2015
Accepted: 30 April 2015
Published online: 23 June 2015
Contributors
Editor
D. W. Pascual
Editor
Metrics & Citations
Metrics
Note:
- For recently published articles, the TOTAL download count will appear as zero until a new month starts.
- There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
- Citation counts come from the Crossref Cited by service.
Citations
Citation
2015.
Mycobacterium bovis BCG Vaccination Induces Divergent Proinflammatory or Regulatory T Cell Responses in Adults. Clin Vaccine Immunol
22:.
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.