Non-tuberculous mycobacteria have diverse effects on BCG efficacy against Mycobacterium tuberculosis

Abstract

The efficacy of Bacillus Calmette-Guerin (BCG) vaccination in protection against pulmonary tuberculosis (TB) is highly variable between populations. One possible explanation for this variability is increased exposure of certain populations to non-tuberculous mycobacteria (NTM). This study used a murine model to determine the effect that exposure to NTM after BCG vaccination had on the efficacy of BCG against aerosol Mycobacterium tuberculosis challenge. The effects of administering live Mycobacterium avium (MA) by an oral route and killed MA by a systemic route on BCG-induced protection were evaluated. CD4+ and CD8+ T cell responses were profiled to define the immunological mechanisms underlying any effect on BCG efficacy. BCG efficacy was enhanced by exposure to killed MA administered by a systemic route; T helper 1 and T helper 17 responses were associated with increased protection. BCG efficacy was reduced by exposure to live MA administered by the oral route; T helper 2 cells were associated with reduced protection. These findings demonstrate that exposure to NTM can induce opposite effects on BCG efficacy depending on route of exposure and viability of NTM. A reproducible model of NTM exposure would be valuable in the evaluation of novel TB vaccine candidates.

Keywords: Bacillus Calmette-Geurin; Mouse model; Mycobacterium avium; Non-tuberculous mycobacteria; Tuberculosis.

Figure 1

Cellular responses in BCG vaccinated mice that were or were not sensitised with MA 724 IP after BCG Responses were assayed at the 12 month time point; 4 weeks after the last dose of MA 724. Cells from inguinal lymph nodes (LN) and spleens were stimulated with PPD-T. (A) Experiment plan (B and C) IFNγ, TNFα and IL-2 production from CD4+ or CD8+ T cells in the LN, (D) IL-17 production from CD4+ T cells in the LN and spleen, (E) IL-4 production from CD4+ T cells in the LN and spleen, (F) expression of CD25, FoxP3 and CD39 on CD4+ T cells from the LN and spleen, (G) IL-10 production from CD4+ T cells in the LN and spleen. B = BCG vaccinated only, BM = BCG vaccinated mice which received MA 724. Results are expressed as a percentage of the CD4+ or CD8+ T cell population and are stimulation specific; media only control well values are subtracted. Each data point represents one mouse and the median is displayed, n = 5. The Mann–Whitney U test was used to determine statistical significance and is shown where differences were significant. Subcutaneous (Sub Cut.), Intraperitoneal (IP), M. avium (MA), Colony Forming Units (CFU), Day 0 (D0), Regulatory T cells (Tregs).

Figure 2

Burden of M. tb CFU in the lungs and spleens 35 days after challenge Groups included naïve, BCG vaccinated (BCG) and BCG followed by MA 724 IP exposure (BCG + MA IP). Data are Log (10) of the colony forming unit (CFU) count of the whole organ. Each data point represents one mouse and the mean is displayed. A One-way ANOVA with Bonferroni post test was used to determine statistical significance; the p value for each comparison is displayed.

Figure 3

Cellular response to M. tb infection 13 days after challenge Thirteen and sixteen days after M. tb challenge cells were isolated from lungs and pulmonary LN (LN) of naïve mice, BCG-vaccinated mice (BCG) and mice vaccinated with BCG followed by MA 724 IP exposure (BM IP). Cells were stimulated with PPD-T and then stained for analysis by flow cytometry. (A) CD44 + CD4+ T cells in the lung, (B) IFNγ, TNFα and IL-2 production from CD4+ T cells in the lung, (C) CD44 + CD8+ T cells in the lung, (D) IFNγ, TNFα and IL-2 production from CD8+ T cells in the LN, (E) IL-17 + CD4+ T cells in the lung and LN, (F) IL-4+ CD4+ T cells in the lung and LN, (G) IL-10 + CD4+ T cells in the lung and LN, (H) expression of CD25, FoxP3 and CD39 on CD4+ T cells from the lung and LN. Cytokine responses are antigen-specific; media control values are subtracted. Regulatory T cell (Treg) frequency is not antigen-specific. Each data point represents one mouse and the median is displayed. The One-way ANOVA Kruskal–Wallis with Dunns post test was used to determine statistical significance; the p value is displayed where significance was calculated.

Figure 4

Responses induced by oral exposure to live MA 104 after BCG vaccination (A) Experimental plan; mice were orally exposed to live MA 104 8 times at 2 week intervals after BCG vaccination (BM) or left un-manipulated after BCG (B) and then challenged with M. tb. Immune responses were recorded at week 28, six weeks after the last dose of MA in the exposed group and directly before challenge with M. tb. Cells from mesenteric lymph nodes (LN) and spleens were stimulated with PPD-T then stained for analysis by flow cytometry. (B and C) IFNγ, TNFα and IL-2 production from CD4+ and CD8+ T cells in the spleen, (D) IL-4+ CD4+ T cells in the LN and spleen, (E) IL-10 + CD4+ T cells in the LN and spleen, (F) expression of CD25, FoxP3 and CD39 on CD4+ T cells from the LN and spleen. Cytokine responses are antigen-specific; media control values are subtracted. Regulatory T cell (Treg) frequency is not antigen-specific. Each data point represents one mouse and the median is displayed. The Mann–Whitney U test was used to test for statistical significance and is shown where differences were significant. Subcutaneous (Sub Cut.), M. avium (MA), Colony Forming Units (CFU), Day 0 (D0), Regulatory T cells (Tregs), Intracellular staining (ICS).

Figure 5

Burden of M. tb CFU in the lungs and spleens 35 days after challenge Groups in the challenge included naïve mice, BCG vaccinated mice (BCG) and BCG vaccinated followed by oral MA 104 exposure mice (BCG + MA 104). Data is Log (10) of the CFU count of the whole organ. Each data point represents one mouse and the mean is displayed. A One-way ANOVA with Bonferroni post test was used to determine statistical significance; the p value for each comparison is displayed. Colony forming units (CFU), 6 months since BCG (6 m).

Figure 6

Cellular response to M. tb infection 16 days after challenge Thirteen and sixteen days after M. tb challenge cells were isolated from lungs and pulmonary lymph nodes (LN) of naïve mice, BCG vaccinated mice (BCG) and BCG vaccinated mice followed by oral MA 104 exposure (BM oral). Cells were stimulated with PPD-T and then stained for analysis by flow cytometry. (A and B) CD44 + CD4+ T cells in the lung at day 13 and 16 after infection, (C) IFNγ, TNFα and IL-2 production from CD4+ T cells in the lung at day 16, (D) IL-17 + CD4+ T cells in the lung and LN at day 16, (E) CD44 + CD8+ T cells in the lung and LN at day 16, (F) IFNγ, TNFα and IL-2 production from CD8+ T cells in the LN at day 16, (G) IL-4+ CD4+ T cells in the lung and LN at day 16, (H) expression of CD25, FoxP3 and CD39 on CD4+ T cells from the lung and LN at day 16. Cytokine responses are antigen-specific; media control well values are subtracted. Regulatory T cell frequency is not antigen-specific. Each data point represents one mouse and the median is displayed. The One-way ANOVA Kruskal–Wallis with Dunns post test was used to determine statistical significance; the p value is displayed where significance was calculated.

Figure S1

Representative flow cytometry data gating strategy Lymphocytes were selected on forward scatter area vs side scatter area then doublets and cell clumps excluded on forward scatter area vs height. Dead cells and B cells were excluded before selecting CD3+ cells. CD4+ T cells were selected from the CD8− population and vice versa for CD8+ T cells. Cytokines were gated on CD4+ or CD8+ T cells. For regulatory T cell phenotyping CD25 expression was gated on CD4+ T cells. On CD25 + cells FoxP3 expression was gated and then CD39 expression was gated on those cells.

Figure S2

Th1 responses in BCG vaccinated mice and mice vaccinated then exposed to killed MA 2-151 IP Mice were vaccinated with BCG subcutaneously then 12 weeks later half the mice received killed MA 2-151 once a month for 7 months. Cells from inguinal lymph nodes and spleens were stimulated with PPD-T and then stained for analysis by flow cytometry. (A) IFNγ, TNFα and IL-2 production from CD4+ T cells in the LN 4 weeks after 1 dose of MA 2-151 was administered to the group receiving MA 2-151 after BCG vaccination; this was 4 months after BCG was administered to both groups, (B) IFNγ, TNFα and IL-2 production from CD4+ T cells in the LN 4 weeks after 7 doses of MA; this was 10 months after BCG, (C) expression of CD25, FoxP3 and CD39 on CD4+ T cells from the LN and spleen. Results are expressed as a percentage of the CD4+ T cell population. BCG with MA 2-151 exposure (BM), BCG alone (B). Cytokine responses are antigen-specific; media control well values are subtracted. Regulatory T cell (Treg) frequency is not antigen-specific. Each data point represents one mouse and the median is displayed. The Mann–Whitney U test was used to test for statistical significance and is shown where differences were significant. 4 months since BCG (4m), 10 months since BCG (10m).

Figure S3

Cellular responses in BCG vaccinated mice that were or were not sensitised with MA 724 IP after BCG Responses were assayed at 12 months; 4 weeks after the last dose of MA. Cells from inguinal lymph nodes were stimulated with PPD-T. The secretion of one or more of IFNγ, TNFα and IL-2 from CD4+ T cells is shown. Results are expressed as a percentage of the CD4+ T cell population and are antigen-specific; media only control well values are subtracted. Each data point represents one mouse and the median is displayed, n = 5. PPD-T (T-PPD).

Figure S4

Cellular responses in mice infected with BCG, MA dead or MA live Mice were injected IP with either 1 × 10⁵ CFU of live MA or 1 × 10⁸ CFU of dead MA or SC with 1 × 10⁵ CFU of BCG. Responses were assayed at 3 months after infection. Splenocytes were stimulated with PPD-T. (A) IFNγ, TNFα and IL-2 production from CD4+ T cells, (B) expression of CD25, FoxP3 and CD39 on CD4+ T cells. Results are expressed as a percentage of the CD4+ T cell population. Cytokine responses are antigen-specific; media only control well values are subtracted. Each data point represents one mouse and the median is displayed, n = 5. PPD-T (T-PPD). The One-way ANOVA Kruskal–Wallis with Dunns post test was used to determine statistical significance; no significant differences were calculated.