Viral control is required for intravenous BCG-mediated protection against tuberculosis in simian immunodeficiency virus-infected macaques

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), is a major global health challenge, especially for people living with HIV (PLWH) not on antiretroviral therapy. We previously showed that intravenous Bacillus Calmette-Guérin (IV-BCG) provides robust protection in SIV+ Mauritian cynomolgus macaques (MCM). Here, we evaluate whether the immunogenicity and efficacy of IV-BCG in SIV+ MCM was impaired by eliminating BCG with drugs. SIV+ macaques were treated with an anti-BCG regimen of isoniazid, rifampicin, and ethambutol (HRE), starting either 1- or 3-weeks post-BCG vaccination. Five months post-vaccination, macaques were challenged with Mtb for 12 weeks. Protection conferred by IV-BCG was significant regardless of HRE timing. However, non-controllers, characterized by high viremia and reduced CD4⁺ T cells in the lungs, were significantly less protected than controllers. Thus, the robust efficacy of IV-BCG in SIV+ MCM is retained regardless of anti-BCG drug treatment, indicating that BCG does not need to survive long to provide protection. In contrast, SIV control was necessary for full protection induced by IV-BCG. Thus, our results support that viral control is critical for vaccine-elicited protection from TB in PLWH.

Extended Data Figure 1:. Higher SIV viremia and systemic inflammation in SIV non-controllers.

a, Longitudinal SIV viremia by qPCR from all groups. SIV control is defined as viral load < 10⁵ cp/ml (Blue-shaded region); non-control as viral load ≥ 10⁵ cp/ml (red-shaded region). Limit of detection is 62 cp/ml. Controllers and non-controllers are identified 14 weeks post-SIV (pre-BCG) and are indicated by blue circles or maroon triangles, respectively. b, Longitudinal ESR of controllers (blue lines) and non-controllers (red lines). Lines are means and error bars represent standard deviation

Extended Data Figure 2:. Immune cells in airways and lung tissue of SIV-infected, IV-BCG-vaccinated, Mtb-challenged MCM.

a, The T, B, and NK cell count in airways, collected by BAL at indicated time points, indexed to time of vaccination. Thin blue and red lines represent individual controllers and non-controllers, respectively. Solid red and blue lines represent group median, and error bars represent range. Mixed-effect analysis with multiple comparisons was used. P values are from uncorrected Fisher’s LSD. Mixed-effect analysis with multiple comparisons of change over time in controllers or non-controllers are in Supplemental Data Table 2. The fixed effect (type III) are in Supplemental Data Table 3. b, Frequency of T, B, and NK cells in lung tissue at necropsy. Three lung lobes were collected per animal. Individual lung lobes are represented by light blue (controllers) or maroon (non-controllers); mean of all three lobes for each animal is represented by solid blue (controllers) and maroon (non-controllers) and lines represent medians. Mann-Whitney test was used to report the P values in (b).

Figure 1:. IV-BCG protection depends on SIV control, not on anti-BCG treatment.

(a) The study outline depicts vaccination groups and the timing of SIV infection, BCG vaccination, HRE treatment, Mtb challenge, PET/CT, blood draws, BAL, and necropsy. b, Mtb burden in lung granulomas, lung lobes, and thoracic lymph nodes at necropsy 12 weeks post-Mtb challenge. c, Mtb burden, comparing HRE-treated and No-HRE groups, irrespective of SIV viremia levels. d, SIV levels in plasma measured longitudinally in SIV controllers (blue circles; viral load < 10⁵ cp/ml) and non-controllers (maroon triangles; viral load ≥ 10⁵ cp/ml). The limit of detection is indicated at 62 cp/ml. e, Plasma viremia 14 weeks post-SIV (pre-BCG) was used to declare animals as either SIV controllers (blue circles) or non-controllers (maroon triangles). Mtb burden 12 weeks post-Mtb challenge, comparing (f) controllers and non-controllers and (g) HRE-treated and No-HRE groups, excluding non-controllers. P values reported in (b) are from Dunn’s multiple comparison adjusted test (after Kruskal-Wallis), comparing each IV-BCG group to the ID BCG group. Mann-Whitney test was used to report the P value in panels c, e, f, and g. Each dot represents an animal, and lines represent medians.

Figure 2. IV-BCG restricts Mtb establishment and dissemination in SIV controllers.

a, Total number of unique Mtb barcodes isolated from each animal, reflecting the number of bacilli that established infection in controllers (blue circles) and non-controllers (maroon triangles). b, Total number of tissue samples sharing a given Mtb barcode, reflecting the extent of dissemination of each bacillus in controllers (blue circles) and non-controllers (maroon triangles). c, Number of tissue samples sharing individual barcodes in lung, thoracic lymph nodes (tLN), and extrapulmonary (EP) tissues (liver, spleen). d, Correlation between total number of unique barcodes and total Mtb CFU in each animal. e, Total number of Mtb in each granuloma from each animal is identified on x-axis, with controllers in blue circles and non-controllers in maroon triangles. NG = No Granuloma from indicated animal. (a) and (d) each dot represents an animal. (b and c) Each dot represents a barcode; lines represent medians. Mann-Whitney test was used to report the P values in (a) - (c). Spearman’s test was used to conduct correlation analysis in d.

Figure 3.. Lung inflammation, granuloma numbers, and pathology increased in SIV non-controllers.

Serial measurement by PET/CT following Mtb challenge of (a) lung inflammation (total FDG activity) and (b) number of granulomas in lungs. Each line represents an animal, with controllers in blue and non-controllers in maroon. TNTC: too numerous to count. c, Three dimensionally-rendered PET/CT images of the thoracic cavity, obtained 12 weeks post-Mtb challenge. Controllers are shown in the top two rows, non-controllers in the bottom row. d, Gross pathology scores assessed at necropsy, with controllers in blue circles and non-controllers in maroon triangles. Each dot represents an animal; lines represent medians. e, The presence of extra-pulmonary disease (EP) was determined by gross pathology and/or growth of Mtb. Percent of controllers and non-controllers exhibiting EP is shown. Mann-Whitney test was used to report P values in (d) and Fisher’s exact test was used to compare prevalence of extra-pulmonary disease (e).

Figure 4.. SIV controllers recruit more T cells and have more functional responses in airways.

a, T cell subsets in airways of SIV-infected, IV-BCG-vaccinated MCM, collected by BAL at indicated time points. b, CD4⁺ T cells producing Th1- or Th17-associated cytokines over time. c, CD8αα⁺ T cells, d, CD8αβ⁺ T cells, and (e) NK cells producing GzmB⁺ or GzmK⁺ at indicated time points. The thin blue and red lines represent individual controllers and non-controllers, respectively. The thicker red and blue lines represent median of controllers and non-controllers; error bars represent range. Mixed-effect analysis with multiple comparisons was used; P values are from uncorrected Fisher’s LSD tests. Red P value indicates changes over time in non-controllers; blue P value indicates changes over time in controllers.

Figure 5.. IV-BCG induces a more robust functional CD4⁺ T cell response in lungs from SIV controllers than from non-controllers.

a, Stacked bar representation of T cell proportions in lung tissue at necropsy. b, Th1- and Th17-associated cytokines produced by CD4⁺ T cells in lungs. c, CD4⁺ T cells producing FoxP3, RORγT, and T-bet in lungs. Three lung lobes were sampled per animal; light colors are individual lung lobes and solid colors are means of all three lung lobes per animal; lines represent medians Mann-Whitney test was used to report P values in (b) and (c).

Figure 6.. CD8αα⁺ T and NK cells express modestly more cytotoxic effectors in lung tissue of SIV controllers.

a, CD8αα and (b) CD8αβ T cells producing granulysin, GzmB, GzmK, and perforin in lungs at necropsy. c, NK cells producing granulysin, GzmB, GzmK, and perforin in lungs. Three lung lobes were collected per animal; light colors are individual lung lobes and solid colors are the mean of all three lung lobes per animal; lines represent medians. Mann-Whitney test was used to report the P values.