Tuberculosis Susceptibility and Vaccine Protection Are Independently Controlled by Host Genotype

Abstract

The outcome of Mycobacterium tuberculosis infection and the immunological response to the bacillus Calmette-Guerin (BCG) vaccine are highly variable in humans. Deciphering the relative importance of host genetics, environment, and vaccine preparation for the efficacy of BCG has proven difficult in natural populations. We developed a model system that captures the breadth of immunological responses observed in outbred individual mice, which can be used to understand the contribution of host genetics to vaccine efficacy. This system employs a panel of highly diverse inbred mouse strains, consisting of the founders and recombinant progeny of the "Collaborative Cross" project. Unlike natural populations, the structure of this panel allows the serial evaluation of genetically identical individuals and the quantification of genotype-specific effects of interventions such as vaccination. When analyzed in the aggregate, our panel resembled natural populations in several important respects: the animals displayed a broad range of susceptibility to M. tuberculosis, differed in their immunological responses to infection, and were not durably protected by BCG vaccination. However, when analyzed at the genotype level, we found that these phenotypic differences were heritable. M. tuberculosis susceptibility varied between lines, from extreme sensitivity to progressive M. tuberculosis clearance. Similarly, only a minority of the genotypes was protected by vaccination. The efficacy of BCG was genetically separable from susceptibility to M. tuberculosis, and the lack of efficacy in the aggregate analysis was driven by nonresponsive lines that mounted a qualitatively distinct response to infection. These observations support an important role for host genetic diversity in determining BCG efficacy and provide a new resource to rationally develop more broadly efficacious vaccines.

Importance: Tuberculosis (TB) remains an urgent global health crisis, and the efficacy of the currently used TB vaccine, M. bovis BCG, is highly variable. The design of more broadly efficacious vaccines depends on understanding the factors that limit the protection imparted by BCG. While these complex factors are difficult to disentangle in natural populations, we used a model population of mice to understand the role of host genetic composition in BCG efficacy. We found that the ability of BCG to protect mice with different genotypes was remarkably variable. The efficacy of BCG did not depend on the intrinsic susceptibility of the animal but, instead, correlated with qualitative differences in the immune responses to the pathogen. These studies suggest that host genetic polymorphism is a critical determinant of vaccine efficacy and provide a model system to develop interventions that will be useful in genetically diverse populations.

FIG 1

M. tuberculosis disease phenotypes in diverse mice. (A to C) Lung CFU (A), spleen CFU (B), and weight change relative to initial weight (C) for individual mice, colored by genotype, at weeks 3 and 6 (for CC001, NOD, 129, A/J, B6, PWK, CAST, and NZO lines) or weeks 3 and 4 (for WSB and CC042 lines, which were moribund by week 4). (D to F) Average lung CFU (D), average spleen CFU (E), and average weight change (F) broken out by mouse genotype. (D to F) All data are the average value ± SD for 3 or 4 mice per strain at each time point. A statistically significant comparison within mouse strains (P < 0.05) using Student’s t test is indicated by a triangle (▵), and hash symbols (#) indicate statistical significance compared to the result for B6 mice via one-way ANOVA with Tukey’s multiple comparison test (#, P < 0.05; ##, P < 0.01; ###, P < 0.001)

FIG 2

Distinct cytokine environments and histopathology results in diverse mice. (A) Representative images from B6, PWK, CAST, and WSB mice at ×2 and ×20 magnification. (B) Proportional areas of granuloma occupied by necrosis, neutrophils, lymphocytes, and macrophages at 3 weeks postinfection (left) and 4 weeks or 6 weeks postinfection (right). Bar heights represent the average values from 10 randomly selected granulomas for each of 3 to 4 mice per mouse strain. (C) Percentage of lung area that was damaged, measured using ImageJ to trace lesion size relative to the whole lung section. Bar heights represent the average values from 3 or 4 mice per strain. Error bars show standard deviations. (D and E) IFN-γ (D) and TNF (E) cytokines measured in homogenates of infected lungs at 3 and 6 weeks (or 3 and 4 weeks for WSB and CC042 mice) postinfection. All data are the average value and SD from 3 or 4 mice per strain at each time point, assayed in technical duplicate. Limit of detection (LOD) was calculated as twofold the background (17 pg/ml).

FIG 3

The effect of genetic background on BCG protection in diverse mice. (A and B) Lung CFU (A) and spleen CFU (B) in naive and vaccinated mice for all genotypes. Each point represents the CFU for an individual mouse at the indicated time point. (C) Fold protection of each mouse strain was calculated from the average CFU of naive compared to BCG-vaccinated groups at 4 and 14 weeks postinfection in the lung and spleen. Coloring indicates CFU reduction (green) or exacerbation (red) in the BCG-vaccinated group. “Protection” is defined as a statistically significant 0.5 log₁₀ reduction in the BGC-vaccinated group. (D) Average percent weight change of each mouse strain relative to initial body weight for BCG-vaccinated and naive mice at 4 and 14 weeks. The percent weight changes were compared between naive and BCG-vaccinated mice within each mouse strain (unpaired t test). n = 6 mice per strain per time point for naive or BCG-vaccinated mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 4

Phenotypic relationships between M. tuberculosis susceptibility and BCG efficacy. (A) Hierarchical clustering of correlations between metrics of susceptibility at weeks 3, 4, 6, and 14 postinfection and response to vaccination at weeks 4 and 14 postinfection. Blue indicates positive Pearson’s correlations, and red indicates negative correlations. “Protection” indicates the relative reduction in CFU or weight loss in the BCG-vaccinated group. “Weight” indicates percent weight change relative to initial body weight. “L” and “S” indicate lung and spleen, respectively; “N” and “BCG” indicate naive and vaccinated groups. Green box indicates the trait cluster containing metrics of BCG-mediated protection. (B and C) Lack of correlation between TB susceptibility and BCG efficacy at early (B) or late (C) time points. (D and E) Positive correlation between IFN-γ production after M. tuberculosis infection and BCG efficacy in the lung (D) or spleen (E). Solid lines are correlations of results at the 4-week time point (circles), and dotted lines are correlations of results at the 14-week time point (squares); each data point in panels B to E is the average value from 3 to 6 mice per genotype. Note that NOD mice are not displayed on the correlation plots due to the highly variable response of this strain (see Fig. S2J in the supplemental material).

FIG 5

BCG protects M. tuberculosis-infected NZO mice from diabetes. (A) Blood glucose measurements obtained from naive and BCG-vaccinated NZO mice at 4 and 14 weeks postinfection. Mice were fasted for 6 h prior to obtaining blood glucose measurements. Mice were considered diabetic if they had a blood glucose concentration greater than 200 mg/dl. (B) Percent weight changes of naive and BCG-vaccinated mice relative to initial mouse weights at 4 and 14 weeks postinfection. Correlation of blood glucose versus lung CFU (C) and correlation of blood glucose versus weight change (D) at 14 weeks postinfection in naive and BCG-vaccinated mice. Each data point is the value for an individual NZO mouse at each specified time point. Naive and BCG-vaccinated groups were compared by Student’s t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001.