T-Cell Recognition of the HspX Protein of Mycobacterium tuberculosis Correlates with Latent M. tuberculosis Infection but Not with M. bovis BCG Vaccination

Abstract

During stationary growth or in vitro conditions mimicking relevant aspects of latency, the HspX protein (Rv2031c) is specifically upregulated by Mycobacterium tuberculosis. In this study we compared T-cell responses against HspX and the secreted M. tuberculosis protein Ag85B (Rv1886c) in tuberculosis (TB) patients, tuberculin skin test-positive individuals, M. bovis BCG-vaccinated individuals, and healthy negative controls. Gamma interferon responses to HspX were significantly higher in M. tuberculosis-exposed individuals than in M. tuberculosis-unexposed BCG vaccinees. In contrast, no such differences were found with respect to T-cell responses against Ag85B. Therefore, BCG-based vaccines containing relevant fragments of HspX may induce improved responses against this TB latency antigen. To identify relevant major histocompatibility complex class I- and class II-restricted HspX-specific T-cell epitopes, we immunized HLA-A2/K^b and HLA-DR3.Ab⁰ transgenic (tg) mice with HspX. Two new T-cell epitopes were identified, p91-105 and p31-50, restricted via HLA-A*0201 and HLA-DRB1*0301, respectively. These epitopes were recognized by human T cells as well, underlining the relevance of HspX T-cell recognition both in vivo and in vitro. In line with the data in humans, BCG immunization of both tg strains did not lead to T-cell responses against HspX-derived epitopes, whereas nonlatency antigens were efficiently recognized. These data support the notion that BCG vaccination per se does not induce T-cell responses against the latency antigen, HspX. Thus, we suggest that subunit vaccines incorporating HspX and/or other latency antigens, as well as recombinant BCG strains expressing latency antigens need to be considered as new vaccines against TB.

Tuberculosis (TB) is one of the leading causes of mortality (two million annually) from an infectious disease, particularly in the developing world (36). It is estimated that 2 billion people are latently infected with Mycobacterium tuberculosis, which causes active disease in 5 to 10% of infected individuals. Most individuals initially control the infection by mounting cell-mediated immunity. However, residual mycobacteria remain viable for many years in healthy immunocompetent hosts in the absence of disease (4), and the majority of contagious TB disease cases will arise from this enormous source of latent TB infection.

The currently available vaccine against TB, M. bovis BCG, is largely ineffective at protecting against pulmonary disease in adults (16, 23). The precise nature of the T-cell response needed for protection against adult pulmonary TB is incompletely defined, as is the cause for the lack of protection by BCG vaccination (3).

The 16-kDa heat shock protein HspX (Rv2031c) is required for mycobacterial persistence within the macrophage and is a dominant protein produced during static growth or under oxygen deprivation (37). Under these conditions, it can account for up to 25% of the total bacillary protein expression. It is proposed that HspX plays an active role in slowing the growth of M. tuberculosis in vivo immediately after infection, as M. tuberculosis mutants lacking the hspX gene exhibited increased growth both in mice and in macrophages (24). In addition to the presence of specific humoral responses against HspX in the sera of cavitary TB patients (29), both T-cell and B-cell responses to HspX were found to be associated with latent M. tuberculosis infection (13, 14), pointing to the importance of HspX as an antigenic target of immune responses during latent TB infection.

Since new vaccines containing relevant fragments of HspX, may induce improved responses against this TB latency antigen, we have generated and characterized HspX-specific, human CD8⁺ and CD4⁺ T cells, restricted by common human lymphocyte antigen (HLA) class I and class II alleles. In addition, peripheral blood mononuclear cells (PBMC) from M. tuberculosis-infected individuals (both active and latent infections) and BCG-vaccinated individuals with or without exposure to M. tuberculosis were examined for their in vitro response to HspX. Finally, the effect of BCG- or HspX immunization on induced immunity against HspX was analyzed in HLA-A2/K^b and HLA-DR3.Ab⁰ transgenic (tg) mice.

MATERIALS AND METHODS

Antigens.

BCG (M. bovis bacillus Calmette-Guérin, Danish 1331) was purchased from the Statens Serum Institute (Copenhagen, Denmark), and killed M. tuberculosis H37Rv sonicate was obtained from D. van Soolingen (RIVM, The Netherlands). The antigen 85B gene (Rv1886c) and the hspX gene (Rv2031c) of M. tuberculosis were amplified by PCR and cloned by Gateway Technology (Invitrogen, San Diego, CA) in a bacterial expression vector containing an N-terminal histidine tag. The proteins were overexpressed in Escherichia coli BL21(DE3) and purified as described previously (17). Residual endotoxin levels were <50 IU/mg as determined by a Limulus amebocyte lysate assay (Cambrex). All proteins were tested to exclude antigen nonspecific T-cell stimulation and cellular toxicity in gamma interferon (IFN-γ) release assays using PBMC of M. tuberculosis-unexposed, BCG-negative, Mantoux skin test-negative healthy donors.

Study subjects.

The study included 63 individuals, among whom were 17 TB patients, 18 tuberculin skin test (TST)-positive individuals, 17 BCG-vaccinated individuals, and 11 non-BCG-vaccinated, TST-negative, healthy Dutch controls. The induration used to define TST positivity is 10 mm. All individuals gave informed consent before blood sampling. The study protocol was approved by the institutional review board of the Leiden University Medical Center.

Enzyme-linked immunospot (ELISPOT) assay for single-cell IFN-γ release.

Venous blood was obtained from study participants in heparinized tubes and PBMC isolated by Ficoll density centrifugation. PBMC (10⁶) were pulsed (16 h) in 48-well plates with antigen (10 μg/ml) in IMDM (Life Technologies, Rockville, MD) with 10% pooled human serum. Nonadherent antigen-pulsed cells (2.5 × 10⁵ cells/well; 150 μl) were transferred to polyvinylidene difluoride-backed 96-well plates (MAIPS45; Millipore, Bedford, MA) that had been precoated for 1 h at 37°C with 5 μg of anti-IFN-γ monoclonal antibody (MAb) 1-D1K (Mabtech, Stockholm, Sweden)/ml, washed six times with IMDM, and blocked (2 h) with IMDM containing 10% fetal calf serum (FCS).

After 16 h of incubation at 37°C and 5% CO₂, plates were washed (phosphate-buffered saline [PBS]-0.05% Tween 20) and incubated with 100 μl of biotinylated anti-IFN-γ MAb (0.3 μg/ml) for 3 h at room temperature, washed and incubated with streptavidin-alkaline phosphatase conjugate (1:1,000; Mabtech) for 2 h, and washed and incubated with 100 μl of nitroblue tetrazolium-BCIP (5-bromo-4-chloro-3-indolylphosphate) substrate (Sigma, St. Louis, MO). The reaction was stopped by the addition of water. Plates were dried and analyzed on a Zeiss Axioplan 2 microscope using KS ELISPOT software (Carl Zeiss Vision, Hallbergmoos, Germany). A positive response to antigen was taken as twice the background.

Synthetic peptides.

15-mer and 20-mer peptides were synthesized by simultaneous multiple peptide synthesis as described previously (30). Homogeneity was confirmed by analytical high-pressure liquid chromatography and by mass spectrometry. Fluorescence-labeled peptides were synthesized as described previously (19).

HLA-DR/peptide binding assay.

As a source of HLA-DR molecules Epstein-Barr virus-transformed B lymphoblastoid cell lines (EBV-BLCL) homozygous for HLA-DR were used. HLA-DR molecules were purified by affinity chromatography and peptide binding to purified HLA-DR3 molecules (60 to 600 nM) was determined as described previously (21). As standard fluorescent peptides influenza hemagglutinin p308-319 (PKYVKQNTLKLAT) was used for HLA-DR1 and HLA-DR2, and hsp65 p3-13 (KTIAYDEEARR) was used for HLA-DR3. Peptide binding affinity was defined as high (50% inhibitory concentration of <1 μM), intermediate (1 to 10 μM), weak (10 to 100 μM), or absent (>100 μM) (21).

Generation and epitope mapping of human, HspX-reactive CD8⁺ T-cell lines.

PBMC (10⁵c/well) derived from an HLA-A*0201⁺ donor were stimulated with peptide pools containing four 15-mer HspX peptides overlapping 10 amino acids (10 μg of each peptide/ml) for 7 days in IMDM (10% human serum) in the presence of rIL-7 (5 ng/ml; Biocarta) and rIL-15 (5 ng/ml; Biocarta) in 96-well round-bottom plates. After 2 days rIL-2 (25 U/ml; Cetus) was added to the cultures.

Intracellular IFN-γ staining.

Seven days later, cells were collected and cocultured for 6 h with the HLA-A*0201-positive EBV-BLCL JY that had been pulsed overnight with single HspX peptides in IMDM (10% FCS) and washed twice with IMDM (10% FCS). During the last 2 h of coculture, brefeldin A was added (10 μg/ml; Sigma). Cell surface staining was performed using CD19-fluorescein isothiocyanate, CD4-PerCP, and CD8-APC (all from Becton Dickinson), after which the cells were permeabilized with Perm buffer (Pharmingen) and stained with anti-IFN-γ-phycoerythrin (Becton Dickinson). Stimulation with phorbol myristate acetate-ionomycin was used as a positive control, and unstained cells were used as a negative control.

Generation and epitope mapping of human, HspX-reactive CD4⁺ T-cell lines.

PBMC (1.5 × 10⁶ cells/well) of healthy human donors were stimulated with HspX (10 μg/ml) in IMDM (10% human serum) in 24-well plates. After 6 days rIL-2 (25 U/ml; Chiron) was added, and cells were additionally cultured for 10 to 15 days, frozen and stored in liquid nitrogen until further use. Human CD4⁺ T-cell lines (1.5 × 10⁵/well) and HLA-DR-matched irradiated PBMC (5 × 10⁵/well) were incubated with either HspX (10, 1.0, 0.1, or 0.01 μg/ml) or HspX-derived peptides (10, 1.0, 0.1, or 0.01 μg/ml) in flat-bottom 96-well plates. After 72 h, 0.5 μCi of [³H]thymidine was added to each well. After 18 h, cells were collected on glass fiber filter strips, and the radioactivity incorporated into the DNA was determined by liquid scintillation counting. The results are the mean of triplicate cultures. The standard errors of the mean (SEM) were <20%.

HLA tg mice.

HLA-DRB1*0301/DRA tg, murine class II-deficient (HLA-DR3.Ab⁰) mice were generated as described previously (18). During breeding, PBMC were typed for expression and segregation of the transgene by flow cytometry and PCR (27). HLA-A2/K^b mice expressed, in addition to the murine class I alleles H2-K^b and H2-D^b, a chimeric HLA-A*0201/K^b gene encoding H2-K^b α3 domain and the HLA-A*0201 α1 and α2 domains (20). HLA-A2/K^b mice were bred under specific-pathogen-free conditions at TNO-PG (Leiden, The Netherlands). Surface expression of HLA-A*0201/K^b molecule was confirmed by flow cytometry.

Immunizations.

Emulsions comprising equal volumes of HspX recombinant protein in PBS and incomplete Freund adjuvant (Difco, Detroit, MI) were prepared and administered as subcutaneous injections into the flanks (in total, 10 μg of protein per mouse [n = 4]). Live M. bovis BCG (Montreal strain) was diluted with PBS. Each mouse was subcutaneously injected with 50 μl around each hind leg (in total, 10⁶ CFU/mouse). At 7 days postinjection, the spleens were removed, and cell suspensions were prepared for in vitro culture. For DNA immunizations mice were injected intramuscularly three times (at 3-week intervals) in both quadriceps (50 μl in each) with HspX plasmid (1 mg/ml) or control DNA (empty vector) in PBS. Splenocytes were harvested 3 weeks after the last DNA injection.

In vitro culture.

Splenocytes from each mouse (1.5 × 10⁵ cells/well) were stimulated in triplicate cultures with antigen in round-bottom 96-well plates in 200 μl of RPMI 1640 (Life Technologies) supplemented with 2 mM l-glutamine (Life Technologies), 100 U/100-μg/ml penicillin-streptomycin solution (Life Technologies), and 10% heat-inactivated FCS (Life Technologies). HspX and M. tuberculosis sonicate were tested at 5 μg/ml, and peptides were tested at 25 μg/ml. After 6 days, 0.5 μCi of [³H]thymidine was added to each well. After 18 h, cells were collected on glass fiber filter strips, and the radioactivity incorporated into the DNA was determined by liquid scintillation counting. The results are the mean of triplicate cultures. SEM were <20%.

ELISA murine IFN-γ.

Splenocytes were resuspended in RPMI 1640 medium (Life Technologies) supplemented with 10% heat-inactivated FCS (INTEGRA Biosciences AG, Switzerland) were added to 96-well U-bottom plates (Corning B.V. Life Sciences, The Netherlands) and stimulated in triplicates with antigens. After 72 h, culture supernatants were taken and evaluated for their IFN-γ content by using a murine IFN-γ CytoSet enzyme-linked immunosorbent assay (Biosource, Camarillo, CA). The assay was performed according to the manufacturer guidelines.

Cytotoxicity assays.

The human EBV-BLCL JY (HLA-A*0201, -B7, -Cw7) was incubated at 37°C for 1 h with 0.1 mCi Na₂⁵¹CrO₄ (Amersham, United Kingdom), washed, and plated with pooled splenocytes from immunized mice (n = 4) in triplicates in 96-well round-bottom plates (2500 cells/well) together with medium, peptide (2 μg), or 5% Triton X-100 (20). After 6 h the supernatants were harvested, and the percent specific lysis was calculated as follows: [(release − spontaneous release)/(maximum release − spontaneous release)] × 100%.

Statistical analysis.

Differences in IFN-γ responses to HspX between different test groups were analyzed with the two-tailed Mann-Whitney U test for nonparametric distribution by using GraphPad Prism (version 4). P values were corrected for multiple comparisons. The statistical significance level used was a P value of <0.05.

RESULTS

T-cell responses against HspX do not correspond to BCG vaccination per se.

Individuals with different stages of M. tuberculosis infection were previously reported to respond differently to HspX (14). Since the effect on T-cell responses in BCG vaccination in areas where TB is and is not endemic may vary (5), we analyzed HspX-responses in IFN-γ ELISPOT assays using PBMC from BCG-vaccinated individuals and from non-BCG vaccinated, healthy controls in addition to TB patients, TST⁺ individuals (Fig. 1). Based on their in vitro IFN-γ response to the M. tuberculosis-specific proteins ESAT-6 and/or CFP-10 (32), we divided BCG vaccinees into M. tuberculosis-unexposed individuals (<10 spot-forming cells [SFC]/10⁶ PBMC; n = 8) and individuals with likely exposure to M. tuberculosis (>10 SFC/10⁶ PBMC; n = 9; ranging from 16 to 116 SFC/10⁶ PBMC).

FIG. 1.

IFN-γ production in response to HspX and Ag85B of M. tuberculosis by PBMC of 17 TB patients (TB), 18 TST⁺ (TST⁺) individuals, 17 BCG-vaccinated individuals (nine in vitro positive for ESAT-6 and/or CFP-10 and eight in vitro negative for ESAT-6 and/or CFP-10), and 11 non-BCG-vaccinated, healthy controls (HC). SFC per 10⁶ PBMC corrected for medium values are given on the y axis. Median values are indicated by horizontal lines. P values are calculated by the nonparametric Mann-Whitney U test. T-cell responses were considered positive if the medium-corrected number of SFC was >10/10⁶ cells.

HspX induced high numbers of IFN-γ-producing T cells in three of the five groups (TB, TST⁺, and BCG vaccinees responding to ESAT-6 and/or CFP-10). These T-cell responses differed significantly (P ≤ 0.0081) from BCG-vaccinated individuals without a response to ESAT-6 and/or CFP-10 and from healthy controls. As a control, the number of IFN-γ-producing T cells induced by the secreted Ag85B was analyzed in the same individuals. Interestingly, no significant differences between the five groups were detected in the number of SFC induced by Ag85B. This indicates that the lack of IFN-γ production in response to HspX in BCG vaccinees and controls is not caused by the absence of antimycobacterial responses in these individuals.

The median values of the HspX/Ag85B response ratio were 0.060 and 0.0062 in BCG-vaccinated, M. tuberculosis-unexposed individuals and the healthy control group, respectively, whereas M. tuberculosis-infected or likely exposed groups had 10-fold-higher HspX/Ag85B response ratios ranging between 0.55 and 0.33 (Table 1).

TABLE 1.

HspX/Ag85B ratios

Test group	Median HspX/Ag85B value
TB patients	0.55
TST+	0.33
BCG, ESAT-6/CFP-10 positivea	0.35
BCG, ESAT-6/CFP-10 negativeb	0.060
Controls, M. tuberculosis unexposed	0.0062

^aIn vitro responsive to ESAT-6 and/or CFP-10.

^bIn vitro unresponsive to ESAT-6 and to CFP-10.

HLA-A*0201-restricted T-cell epitopes.

Since mounting evidence from human studies and murine models of TB points toward a role of CD8⁺ T cells in controlling (latent) M. tuberculosis infection (22, 28, 31), we set out to identify HLA-class I-restricted T-cell responses to the HspX M. tuberculosis latency antigen. PBMC derived from an HspX responding HLA-A*0201⁺ TB-exposed donor were stimulated with seven peptide pools, each containing four HspX-derived 15-mer peptides with an overlap of 10 amino acids. After 7 days, cells were collected and cocultured for 6 h with HLA-A*0201-positive EBV-BLCL that had been pulsed overnight with single HspX peptides, and the induction of intracellular IFN-γ was assessed by fluorescence-activated cell sorting analysis (Fig. 2). Intracellular IFN-γ-producing CD8⁺ T cells were detected in the T-cell line generated with pool 5. In this pool only one peptide, HspX p91-105, induced significant levels of intracellular IFN-γ production (Fig. 2).

FIG. 2.

Intracellular IFN-γ staining of HLA-A*0201⁺, CD8⁺ human T-cell lines generated by stimulation of PBMC with seven peptide pools, each containing four HspX 15-mers. T-cell lines were cocultured for 6 h with the HLA-A*0201-positive EBV-BLCL JY that had been pulsed overnight with one single HspX peptide (indicated on the left). Cells were incubated with phycoerythrin-labeled anti-IFN-γ and APC-labeled anti-CD8 MAb. For the analysis, only lymphocytes were used with the exclusion of CD19-positive cells to avoid contamination of the JY cells.

Murine HLA-A*0201-restricted epitopes.

HLA-A2/K^b mice represent a suitable model for the induction and detection of high-affinity HLA-A*0201-restricted CD8⁺ CTL responses in vivo (11, 20) Thus, these mice were immunized three times with plasmid DNA encoding hspX or empty vector DNA. Three weeks after the last DNA injection, the splenocytes were incubated with peptide-pulsed, ⁵¹Cr-labeled HLA-A*0201-positive JY cells (which express HLA-A*0201 but not H2-K^b or H2-D^b) in order to analyze their cytolytic potential. JY cells pulsed with HspX p91-105 were lysed by splenocytes of immunized mice (Fig. 3). Splenocytes derived from mice immunized similarly with empty vector DNA did not show any lysis (data not shown). These data show that the same HLA-A*0201-restricted cytotoxic-T-lymphocyte (CTL) epitope is identified in HLA-A*0201 humans and in HLA-A2/K^b. The relatively low level of lysis by the CTL raised in HLA-A2/K^b mice (maximal 25%) could be due to the presence of the chimeric HLA-A*0201/K^b gene in the tg mice. This expresses the H2-K^b α3- and the HLA-A*0201 α1 and α2 molecules, allowing efficient CD8 interaction of murine CD8⁺ T cells, and thus interacts less efficiently with the human HLA-A*0201 molecule expressed on the JY target cells.

FIG. 3.

Cytotoxic activity of splenocytes derived from HLA-A2/K^b mice, immunized with HspX plasmid DNA. M. tuberculosis-derived peptides used to pulse HLA-A*0201⁺ target cells (JY) are indicated on the x axis. The percent specific lysis is indicated on the y axis. E:T, effector/target ratio.

HLA-DR binding of HspX-derived peptides and HLA-DR-restricted T-cell epitopes.

Previously, several studies have shown that different regions of the M. tuberculosis HspX latency antigen can provoke human CD4⁺ T-cell responses. In particular, the immunodominant region p91-105 is permissively recognized, since peptide epitopes were defined in the context of HLA-DRB1*0101, HLA-DRB1*1101, HLA-DRB1*1301, HLA-DRB1*1501, and HLA-DRB1*0401 (1, 2). Other peptide sequences recognized by human T cells include p21-40 (HLA-DRB1*0401) (1), p21-29, and p120-128 (both HLA-A*0201) (10).

To study CD4⁺ T-cell responses to HspX, we first assessed the HLA-binding affinity of HspX-derived peptides. Ten amino acids overlapping 20-mer peptides, covering the entire sequence of the HspX protein of M. tuberculosis were used to analyze peptide binding to purified HLA molecules (Table 2). HLA-DR1 and HLA-DR2 showed similar binding affinities for the HspX peptides, particularly p11-30 and p91-105 bound with high affinity to these alleles. HLA-DR3, on the other hand, bound especially well to p31-50, which was confirmed by the presence of the HLA-DR3 peptide binding motif (21) with L (leucine) at position n, E (glutamate) at position n + 2, and K (lysine) at position n + 4 (Table 2).

TABLE 2.

HLA-DR binding of overlapping HspX peptides^a

Amino acids	Sequence	IC50 (μM)
		HLA−:DR*0101	DR*1501	DR*0301
1-20	MATTLPVQRHPRSLFPEFSE	15	22	>100
11-30	PRSLFPEFSELFAAFPSFAG	<0.7	<0.7	42
21-40	LFAAFPSFAGLRPTFDTRLM	5	0.9	68
31-50	LRPTFDTRLMRLEDEMKEGR	24	35	0.5
41-60	RLEDEMKEGRYEVRAELPGV	8	13	41
51-70	YEVRAELPGVDPDKDVDIMV	12	17	35
61-80	DPDKDVDIMVRDGQLTIKAE	5	3	4
71-90	RDGQLTIKAERTEQKDFDGR	14	9	>100
81-100	RTEQKDFDGRSEFAYGSFVR	6	1	>100
91-110	SEFAYGSFVRTVSLPVGADE	<0.7	<0.7	45
101-120	TVSLPVGADEDDIKATYDKG	>100	>100	>100
111-130	DDIKATYDKGILTVSVAVSE	>100	70	>100
121-140	ILTVSVAVSEGKPTEKHIQI	>100	>100	>100
125-144	SVAVSEGKPTEKHIQIRSTN	>100	>100	>100

^aT-cell epitopes and associated values are depicted in boldface.

HspX 20-mers were used for epitope mapping in three human, HspX-specific T-cell lines from individuals homozygous for HLA-DRB1*0101, HLA-DRB1*01501, or HLA-DRB1*0301 (Fig. 4). The HLA-DRB1*0101⁺ T-cell line recognized p91-105 (Fig. 4A), which also binds with high affinity to both HLA-DR1 and HLA-DR2. Another peptide that bound with high affinity to these alleles, p21-40 (Table 2), induced high proliferative responses, whereas p71-90, which bound with lower affinity, induced moderate responses in the HLA-DRB1*01501⁺ T-cell line (Fig. 4B). The HLA-DR3-restricted epitope identified here, p31-50 (Fig. 4C), corresponded well with the high binding affinity for HLA-DR3 as described for other HLA-DR3-restricted, mycobacterial peptide epitopes (18). T-cell responses were HLA-DR-restricted as MAbs directed against HLA-DR (B8.11.2) decreased proliferation of all three T-cell lines (data not shown).

FIG. 4.

T-cell proliferation of HLA-DR1⁺ (A), HLA-DR2⁺ (B), or HLA-DR3⁺ (C) human T-cell lines responsive to the HspX protein or of splenocytes from HLA-DR3.Ab⁰ tg mice immunized with recombinant protein of HspX (D). T-cell proliferation (in counts per minute [cpm]) corrected for medium values is shown on the y axis. The SEM were <12%. The amino acid numbering of the overlapping HspX peptides from Table 2 is given on the x axis.

Murine HLA-DR3-restricted epitopes.

To address the in vivo immunogenicity and epitope specificity of the HspX protein, we used HLA-DR3.Ab⁰ tg mice. After immunization of the mice with HspX protein in incomplete Freund adjuvant, splenocytes were restimulated in vitro with the overlapping HspX 20-mer peptides. Since the HLA-DR3.Ab⁰ mice are devoid of any murine class II molecules expressed at the cell surface, all CD4⁺ T cells are restricted by HLA-DR3. Figure 4D shows that murine HLA-DR3-restricted T cells recognize the same immunodominant HspX peptide (p31-50) as human HLA-DR3-restricted T cells, indicating that p31-50 is an in vivo-processed and -presented HLA-DR3-restricted T-cell epitope.

BCG immunization of HLA-A2/K^b and HLA-DR3.Ab⁰ does not induce T-cell responses against HspX.

Since BCG-vaccinated, M. tuberculosis-unexposed individuals in a cross-sectional comparison did not show T-cell responses directed against HspX (Fig. 1) (33), we decided to evaluate this in HLA-tg mice. Splenocytes were harvested 10 days after BCG vaccination and analyzed for their ability to lyse the human target cell JY, which expresses HLA-A*0201. Target cells were pulsed with one of seven known HLA-A*0201-epitopes (10 μg/ml per peptide): hsp65 p369-377 (11), Ag85B p143-152 and Ag85B p199-207(20), HspX p91-105 (the present study), HspX p21-29, and HspX p120-128 (10) or the HLA-A2-binding influenza A matrix peptide: Flu p58-66 (34).

The hsp65 and Ag85 epitopes were strongly recognized by CTL in a dose-dependent fashion (Fig. 5A). However, none of the target cells pulsed with HspX peptides were lysed. No lysis was observed for control cells: splenocytes pulsed with HLA-A*0201-binding influenza virus p58-66 (Fig. 5A) or splenocytes derived from unimmunized A2/K^b mice (data not shown).

FIG. 5.

(A) Cytotoxic activity of splenocytes derived from HLA-A2/K^b mice immunized with BCG. Ten days after immunization, splenocytes were harvested and analyzed for their ability to lyse human HLA-A*0201-positive target cells (JY). HLA-A*0201-restricted peptides used to pulse target cells are indicated on the x axis. The HLA-A*0201-binding peptide of influenza A matrix (F58-66) was used as a control. E:T, effector/target ratio. The sequences of peptides used in the CTL assay were as follows: hsp65p369-377, KLAGGVAVI; Ag85B p143-152, FIYAGSLSAL; Ag85B p199-207, KLVANNTRL; HspX p91-105, SEFAYGSFVRTVSLP; HspX p21-29, LFAAFPSFA; HspX p120-128, GILTVSVAV; and influenza A matrix p58-66, GILGFVFTL. (B) IFN-γ production of splenocytes of HLA-DR3.Ab⁰ mice immunized with BCG in PBS (BCG) or with PBS only (mock). The antigens used for in vitro challenge are indicated on the y axis.

Similarly, we investigated the T-cell responses in HLA-DR3.Ab⁰ tg mice after immunization with BCG. Splenocytes of BCG-immunized mice were restimulated in vitro with the HspX protein, its HLA-DR3-restricted epitope p31-50, M. tuberculosis hsp65, its HLA-DR3-restricted epitope p3-13, Ag85B, or its HLA-DR3-restricted epitope p51-70 (Fig. 5B). Again, no IFN-γ responses were detected in response to HspX or its HLA-DR3-restricted epitope. In contrast, both other proteins and their HLA-DR3-restricted epitopes induced significant levels of IFN-γ. No T-cell responses could be detected against any of the antigen in mock-immunized HLA-DR3.Ab⁰ mice.

DISCUSSION

Since BCG vaccination does not protect against reactivation of TB, mycobacterial antigens that induce differential T-cell responses in individuals latently infected with M. tuberculosis compared to unexposed BCG-vaccinated individuals may help to identify protective antigens and immune responses against TB. Such antigens may be applied in the development of postexposure vaccines, as well as for new specific diagnostic tools.

By means of whole-genome DNA microarray expression profiling and proteomic analysis, 48 so-called latency antigens have been identified (35). In a previous study, our group analyzed the immunogenicity of 25 of these latency antigens and observed strong IFN-γ responses, particularly in latently infected individuals (25).

In the present study we focused on the latency antigen HspX, since it is required for bacterial growth within the macrophage and is predominantly present during stationary growth of M. tuberculosis (37). Moreover, strong T-cell responses to HspX in African populations were observed that were mostly restricted to latently infected individuals (14). In addition, T cells from TB patients recognizing the HspX protein showed a switch from Th0 toward Th1 after chemotherapy, indicating their potential to induce protective T-cell responses (9, 15). These characteristics of HspX make it an interesting target for postexposure TB vaccination, as well as for the possible diagnosis of preclinical infection.

We describe here HspX-specific T-cell responses in M. tuberculosis (likely)-exposed individuals (TB patients, TST⁺ asymptomatic individuals, and BCG vaccinees with positive ESAT-6 and/or CFP-10 T-cell responses) compared to M. tuberculosis-unexposed individuals (BCG vaccinees and healthy controls lacking a T-cell response to ESAT-6 and/or CFP-10). Our results show that most (24 of 34 [71%]) M. tuberculosis-infected or -exposed individuals responded well in the ELISPOT assay to HspX, whereas sporadic and significantly lower responses were observed in M. tuberculosis-unexposed individuals, including BCG-vaccinated individuals without any known exposure to M. tuberculosis. This suggests that BCG vaccination alone does not induce T-cell responses against the HspX antigen. Similar findings come from a study in The Gambia where neonatal BCG immunization did not lead to IFN-γ responses to HspX or CFP-10, whereas these proteins were well recognized in M. tuberculosis-exposed household contacts and healthcare workers (33).

Several studies have indicated that BCG can express the HspX homologue during oxygen depletion (6, 12, 26); however, our data suggest that vaccination with BCG in humans does not induce immune responses against HspX. We hypothesize that the expression of HspX by BCG in vivo after vaccination is limited and insufficient to induce an immune response.

To confirm this assumption in vivo, we applied BCG immunization to HLA-class I and HLA-class II tg mouse models—HLA-A2/K^b and HLA-DR3.Ab⁰, respectively—which provide powerful models to help characterize the in vivo T-cell responses to mycobacterial antigens in the context of HLA polymorphism (18, 20). Furthermore, the HLA-A*0201 and HLA-DRB1*0301 alleles are major HLA alleles since their overall frequencies are 42 and 24%, respectively, in Caucasian, African/Afro-Caribbean, and Oriental populations (8).

Splenocytes derived from BCG-immunized HLA-A*0201.K^b mice failed to lyse target cells pulsed with HLA-A*0201-restricted HspX peptides, whereas target cells pulsed with HLA-A*0201-restricted epitopes derived from hsp65 or Ag85B of M. tuberculosis were lysed up to 54% (Fig. 5A). Similarly splenocytes from BCG-immunized DR3.Ab⁰ mice were challenged in vitro with several mycobacterial antigen and their HLA-DR3-restricted peptides. Whereas immunization of the HLA-DR3.Ab⁰ mice with the HspX protein induced responses against HspX p31-50 and the HspX protein (Fig. 4D), such responses were not detected after BCG immunization of HLA-DR3.Ab⁰ mice (Fig. 5B). In contrast, both the cytosolic hsp65 and the secreted Ag85B of M. tuberculosis and their respective HLA-DR3-restricted epitopes, p3-13 and p51-70, induced significant levels of IFN-γ in these mice, a finding consistent with our previous findings (18). These data indicate that BCG immunization does not provoke T-cell responses against the HspX latency antigen, while it is possible to induce CD4⁺- and CD8⁺ T-cell responses to HspX using protein- or DNA vaccination. Thus, these data suggest that expression of HspX by BCG after in vivo vaccination is probably low compared to the expression of its homologue in M. tuberculosis under latent conditions and thus will not lead to a significant immune response directed against HspX.

This study shows that BCG vaccination alone does not induce T-cell responses against the HspX antigen but that HspX is an immunogenic antigen that harbors several T-cell epitopes. Thus, we anticipate that improved (BCG) vaccines, expressing relevant fragments of M. tuberculosis latency antigens (7, 25), may have potential as vaccines against latent TB.