Enhanced priming of adaptive immunity by a proapoptotic mutant of Mycobacterium tuberculosis

Abstract

The inhibition of apoptosis of infected host cells is a well-known but poorly understood function of pathogenic mycobacteria. We show that inactivation of the secA2 gene in Mycobacterium tuberculosis, which encodes a component of a virulence-associated protein secretion system, enhanced the apoptosis of infected macrophages by diminishing secretion of mycobacterial superoxide dismutase. Deletion of secA2 markedly increased priming of antigen-specific CD8(+) T cells in vivo, and vaccination of mice and guinea pigs with a secA2 mutant significantly increased resistance to M. tuberculosis challenge compared with standard M. bovis bacille Calmette-Guérin vaccination. Our results define a mechanism for a key immune evasion strategy of M. tuberculosis and provide what we believe to be a novel approach for improving mycobacterial vaccines.

Figure 1. Induction of caspase and SodA-dependent apoptosis in macrophages by infection with M. tuberculosis ΔsecA2.

(A) Human THP-1 cells were infected with M. tuberculosis (H37Rv), ΔsecA2, and complemented ΔsecA2 (ΔsecA2-C) strains at an MOI of 10. Cells were stained after 48 and 72 hours for TUNEL or for caspase activation using FLICA probes and analyzed by FACS. Representative histograms of staining with TUNEL or a polycaspase-specific probe (fluorescein-VAD-FMK) at 48 hours are shown (open histogram, uninfected; filled histogram, infected with H37Rv or ΔsecA2 as indicated). Quantitation by FACS of cells positive for TUNEL or for active caspase-8, caspase-9, or polycaspase is summarized. Data shown are from the time points for peak signal detection (48 hours for TUNEL and polycaspase, 72 hours for caspase-8 and 9). (B) TUNEL staining of mouse BMMs 72 hours after infection with indicated bacterial strains. Staining was quantitated by FACS as in A. (C) TUNEL staining of THP-1 cells infected with the indicated bacteria for 48 hours in the absence or presence of caspase-3–specific inhibitor (Z-DEVD-FMK). *P < 0.001 compared to no inhibitor (1-way ANOVA, Bonferroni post-test). (D) Filtrates from cultures of H37Rv, ΔsecA2 mutant, or ΔsecA2-αsodA were analyzed for SodA activity (left). Graph shows mean and range for duplicate samples, with SodA activity normalized to the level of WT (H37Rv). ^†Below detection level. The ΔsecA2-αsodA showed restoration of SodA activity in the culture filtrate and also reversed the TUNEL⁺ cell death induced by ΔsecA2 in THP-1 cells infected for 48 hours at an MOI of 10:1 (right).

Figure 2. Enhanced CD8⁺ T cell activation in vivo by ΔsecA2.

(A) Mice were infected i.v. with the indicated bacterial strains, and spleens were harvested after 7 days for quantitation of SIINFEKL-specific T cells by IFN-γ ELISPOT. SFC, spot-forming cells. (B) CTL activity was measured in vivo by transfer of a mixture of SIINFEKL-pulsed CFSE^low and unpulsed CFSE^high splenocytes into recipient mice 14 days after infection with the indicated bacteria. Relative proportions of CFSE^low and CFSE^high were quantitated by FACS to determine percent specific lysis. (C) CD4⁺ T cell responses were unaffected by deletion of secA2. Seven days after infection with indicated bacteria, responses of splenocytes to purified protein derivative (PPD) and peptide-25 (Pep25) were determined by ELISPOT. (D) Thy1.1⁺ B6.PL mice were injected with CFSE-labeled Thy1.2⁺ OT-I splenocytes, and infected with the indicated bacteria. CD8⁺ T cell activation was assessed 6 days after infection by CFSE dilution. Dot plots show representative mice, and bar graphs show means and SDs for percentages of undivided cells in the transferred population for 2 or 3 mice per bacterial strain. (E) The enhanced CD8⁺ T cell proliferation induced by ΔsecA2 was reversed by reexpression of SodA secretion in the ΔsecA2-αsodA strain. Dot plots show CFSE dilution in the transferred population 6 days after infection with the indicated strains. The bar graph on the right shows means and SDs of the percentages of undivided cells for groups of 4 mice infected with each bacterial strain. All results are representative of at least 2 independent experiments.

Figure 3. Augmented memory T cell populations following immunization with ΔsecA2.

C57BL/6 mice were adoptively transferred with 5 × 10⁵ OT-I splenocytes 24 hours prior to vaccination by subcutaneous infection with either H37Rv-OVA or ΔsecA2-OVA. The animals were sacrificed 2, 4, 8, and 20 weeks after vaccination, and splenocytes were stained for 5-color fluorescence analysis using antibodies specific to Thy1.2, B220, CD62L, and CD44 plus SIINFEKL-loaded H-2K^b tetramer. (A) FACS analysis of total B220^– events among 1.5 × 10⁶ total lymphocytes, showing proportions of Thy1.2⁺ cells staining with tetramer. (B) Dot plots show expression of CD44 and CD62L on cells gated for tetramer staining as in A from representative mice sacrificed 2 weeks after infection with the indicated bacterial strains. The graphs show percentages of total B220^– lymphocytes staining with tetramer (Tet) and expressing either high or low levels of CD62L at each time point after vaccination. Filled squares represent ΔsecA2-vaccinated mice and open squares, H37Rv-OVA–vaccinated mice. Each symbol represents mean and SD for groups of 2 or 3 mice.

Figure 4. Protective immunity against virulent M. tuberculosis challenge in mice following vaccination with ΔsecA2.

(A) C57BL/6 mice were vaccinated subcutaneously with saline (naive) or with 1 × 10⁶ BCG or ΔsecA2 and challenged by aerosol 2 months later with 50–100 CFU of virulent M. tuberculosis Beijing/W strain (HN878). Graphs show means and SDs of CFU of M. tuberculosis in lung and spleen at 1, 3, and 5 months after challenge for groups of 5 mice that were either naive (white bars), BCG vaccinated (gray bars), or ΔsecA2 vaccinated (black bars). *P < 0.05 and ^†P < 0.001, compared with unvaccinated group or between bracketed groups; 1-way ANOVA with Tukey’s post-hoc test. (B) Lungs of mice vaccinated and challenged with virulent M. tuberculosis as in A were examined histologically at 1 and 3 months after challenge. More severe, spreading lung lesions with extensive granulomatous pneumonia and consolidation were observed in unvaccinated mice as compared with mice vaccinated with either BCG or ΔsecA2. Original magnification, ×20. (C) C57BL/6 mice were vaccinated subcutaneously with saline or with 1 × 10⁶ BCG or ΔsecA2 and challenged by aerosol 2 months later with 50–100 CFU of virulent M. tuberculosis Erdman strain. Mice were observed daily for survival for 1 year after challenge. Moderate extension of survival was observed in BCG-vaccinated animals (P = 0.006 compared with naive control group; log rank test), and much more pronounced extension of survival was observed in ΔsecA2-vaccinated animals (P = 0.0001 compared with naive; P = 0.0062 compared with BCG).

Figure 5. Protective immunity against virulent M. tuberculosis challenge in guinea pigs following vaccination with ΔsecA2.

(A) Outbred Hartley guinea pigs were vaccinated intradermally with saline, BCG, or ΔsecA2 (n = 5 animals per group) and challenged by aerosol 2 months later with 10–30 CFU of virulent M. tuberculosis H37Rv. Graphs show SEM of CFU of M. tuberculosis in lung, spleen, and mediastinal lymph nodes at 1 and 2 months after challenge for naive (white bars), BCG-vaccinated (gray bars), or ΔsecA2-vaccinated (black bars) animals (CFU counts obtained from a minimum of 3 and in most instances 5 animals; an exception was 2-month spleen CFU for BCG-vaccinated animals, in which counts were obtained from only 1 animal for technical reasons). (B) Quantitative scoring of histopathology graphed as the SEM of the rank scores for the groups of guinea pigs vaccinated and challenged as described in A. Images show representative low-power (×25) views of H&E-stained sections of mediastinal lymph nodes harvested from each group at 1 or 2 months after challenge. Scale bars: 1 mm. *P < 0.05, **P < 0.01, ^†P < 0.001 compared with unvaccinated group or between bracketed groups; 1-way ANOVA with Tukey’s post-hoc test.