Concurrent infection with Mycobacterium tuberculosis confers robust protection against secondary infection in macaques

Abstract

For many pathogens, including most targets of effective vaccines, infection elicits an immune response that confers significant protection against reinfection. There has been significant debate as to whether natural Mycobacterium tuberculosis (Mtb) infection confers protection against reinfection. Here we experimentally assessed the protection conferred by concurrent Mtb infection in macaques, a robust experimental model of human tuberculosis (TB), using a combination of serial imaging and Mtb challenge strains differentiated by DNA identifiers. Strikingly, ongoing Mtb infection provided complete protection against establishment of secondary infection in over half of the macaques and allowed near sterilizing bacterial control for those in which a secondary infection was established. By contrast, boosted BCG vaccination reduced granuloma inflammation but had no impact on early granuloma bacterial burden. These findings are evidence of highly effective concomitant mycobacterial immunity in the lung, which may inform TB vaccine design and development.

Fig 1. Concurrent Mtb infection limits the establishment of new granulomas.

(a) Experimental schema of reinfection: Macaques (N = 8) were infected for 16 weeks with low dose (<15 CFU) Mtb library A and then rechallenged with low dose Mtb library B for 4–5 more weeks. Naïve (N = 6) macaques were challenged in parallel with only Mtb library B. (b) Low dose library A (primary) infection in the reinfection cohort resulted in a spectrum of host outcomes at 16 weeks as assessed by [¹⁸F]-FDG-PET/CT; total thoracic bacterial burden was estimated from the lung FDG activity and error bars denote 95% confidence intervals. (c) Imaging was used to discriminate primary strain infection and dissemination (green arrows), new granulomas after re-infection (yellow arrows); three different macaques are shown at 4 months post-primary infection, and 4 weeks post-reinfection. Monkey 19815 (far right panels) had no new granulomas detected after secondary infection. (d) Macaques with a primary infection had fewer newly established secondary granulomas seen by imaging and confirmed as containing Mtb library B DNA at 4 weeks vs. naïve control monkeys (p = 0.0083, Mann-Whitney test).

Fig 2. Reinfection promotes sterilizing immunity in secondary strain granulomas and restricts thoracic lymph node dissemination.

(a) The bacterial burdens (CFU) of all granulomas in naïve macaques (blue), or reinfection macaques (library A: gray; library B: red; library A and B: red/gray circles) 4 weeks post-Mtb library B challenge. The CFU for all granulomas in the reinfected macaques were significantly reduced relative to the naïve monkeys. N = 42 ranging from 2 to 20 per animal (naïve) and N = 94 ranging from 5 to 32 per animal (reinfection). Each symbol is a granuloma, p < .0001, Mann-Whitney test. (b) Total bacterial genome counts (CEQ) and viable bacterial burden counts (CFU) were determined from the same Library B granulomas from naïve or reinfection macaques and bacterial killing (CFU/CEQ) for each granuloma shown in (c), p = 0.0002, Mann-Whitney test. Each symbol for (b) and (c) is a granuloma. CFU was significantly lower than CEQ in both naïve and reinfection granulomas (b), paired t-tests, p-values reported in figure. Both CEQ and CFU were reduced in reinfection granulomas compared with granulomas from naïve animals (b), Mann-Whitney tests, p-values reported in figure. (d) There was no correlation between extent of host lung inflammation (total lung FDG activity) and bacterial killing (CFU/CEQ) of library B granulomas (Spearman r = 0.38, p = 0.2454). In b, N = 12 (naïve) and N = 21 (reinfection). In b and c, N = 12 (naïve) and N = 11 (reinfection).

Fig 3. Primary Mtb infection reduces bacterial burden in secondary challenge granulomas, while boosted BCG has no effect on early granuloma bacterial burden.

(a, b) Size (mm, by CT) and inflammation (as measured by [¹⁸F]-FDG avidity via standard uptake value, SUV) for library B (for naïve and reinfection macaques) or Mtb (for BCG+H56 and unvaccinated controls) granulomas. (c) CFU of Mtb library B granulomas from naïve (blue) and reinfection (red) macaques and 5–6 week granuloma CFU in boosted BCG (green) and unvaccinated control (purple) macaques challenged with Mtb Erdman. For a, b, and c, each symbol is a granuloma, p < 0.0001, Kruskal-Wallis tests, Dunn’s multiple comparison adjustment was used and adjusted p-values are reported in the figure above corresponding comparisons. All granulomas from N = 6 macaques (naïve); N = 7 macaques (reinfection); N = 6 BCG+H56 macaques; N = 6 unvaccinated macaques are represented. (d) Left panel: Number of thoracic lymph nodes from each macaque which were CFU+ for Mtb library B. Right panel: Number of thoracic lymph nodes from unvaccinated or BCG+H56 vaccinated CFU+ positive for Mtb at 5–6 weeks post-infection. N = 6 (naïve), N = 8 (reinfection), and N = 6 (unvaccinated). Each symbol is a macaque, Mann-Whitney test p-values reported in figure.

Fig 4. Reinfection challenge strain granulomas have elevated IL-10 T cell responses.

Granuloma T cell responses by intracellular cytokine staining and flow cytometry from naïve and reinfection animals. N = 26 (naïve), N = 34 (library A, reinfection), and N = 7 (library B, reinfection). Each dot is a granuloma, Kruskal-Wallis test, Dunn’s multiple comparison adjustment was used and adjusted p-values are reported.

Fig 5. Reinfection challenge strain granulomas have reduced levels of pro-inflammatory cytokines and chemokines.

Supernatants from library B granulomas from naïve or reinfection macaques were assessed for cytokine and chemokines. N = 27 (naïve) and N = 12 (reinfection). Each dot is a granuloma, p < 0.05–0.0001, Mann-Whitney test.