Host genetic background is a barrier to broadly effective vaccine–mediated protection against tuberculosis
Rocky Lai, Diana N. Gong, Travis Williams, Abiola F. Ogunsola, Kelly Cavallo, Cecilia S. Lindestam Arlehamn, Sarah Acolatse, Gillian L. Beamer, Martin T. Ferris, Christopher M. Sassetti, Douglas A. Lauffenburger, Samuel M. Behar
Rocky Lai, Diana N. Gong, Travis Williams, Abiola F. Ogunsola, Kelly Cavallo, Cecilia S. Lindestam Arlehamn, Sarah Acolatse, Gillian L. Beamer, Martin T. Ferris, Christopher M. Sassetti, Douglas A. Lauffenburger, Samuel M. Behar
View: Text | PDF
Research Article Infectious disease

Host genetic background is a barrier to broadly effective vaccine–mediated protection against tuberculosis

Abstract

Heterogeneity in human immune responses is difficult to model in standard laboratory mice. To understand how host variation affects Bacillus Calmette Guerin–induced (BCG-induced) immunity against Mycobacterium tuberculosis, we studied 24 unique collaborative cross (CC) mouse strains, which differ primarily in the genes and alleles they inherit from founder strains. The CC strains were vaccinated with or without BCG and challenged with aerosolized M. tuberculosis. Since BCG protects only half of the CC strains tested, we concluded that host genetics has a major influence on BCG-induced immunity against M. tuberculosis infection, making it an important barrier to vaccine-mediated protection. Importantly, BCG efficacy is dissociable from inherent susceptibility to tuberculosis (TB). T cell immunity was extensively characterized to identify components associated with protection that were stimulated by BCG and recalled after M. tuberculosis infection. Although considerable diversity is observed, BCG has little impact on the composition of T cells in the lung after infection. Instead, variability is largely shaped by host genetics. BCG-elicited protection against TB correlated with changes in immune function. Thus, CC mice can be used to define correlates of protection and to identify vaccine strategies that protect a larger fraction of genetically diverse individuals instead of optimizing protection for a single genotype.

Authors

Rocky Lai, Diana N. Gong, Travis Williams, Abiola F. Ogunsola, Kelly Cavallo, Cecilia S. Lindestam Arlehamn, Sarah Acolatse, Gillian L. Beamer, Martin T. Ferris, Christopher M. Sassetti, Douglas A. Lauffenburger, Samuel M. Behar

×

Figure 6

Multivariate correlates of protection across all 24 CC strains.

Options: View larger image (or click on image) Download as PowerPoint
Multivariate correlates of protection across all 24 CC strains. (A) A PL...
(A) A PLSR model was performed to associate BCG-induced changes in lung immune features with BCG-induced changes in lung CFU. Each data point represents a vaccinated mouse that has the mean of the unvaccinated mice from the same strain subtracted from it. The scores of the top 2 latent variables of the model are shown. The color of each data point represents the change in lung CFU induced by BCG vaccination. The percentages of the variance in the lung features captured by each latent variable are shown as percentages in the axes. (B) The bar graph depicts the loadings of the first latent variable of the PLSR model (A). Positive loadings (i.e., on the positive x-axis in the scores plot) indicate immune features that increase in BCG-vaccinated mice, which generally indicates an association with worse protection. Negative loadings indicate immune features that increase in mice protected by BCG. (C) The heatmap illustrates all features selected included in the model, indicating their importance in associating CC strains with the lung protection continuum. Colors are representations of z-scored features, with a different color representing each 10% quantile of the data. Columns are averages of the CC strains (n = 5 for most cases). Feature names are colored based on the data set from which they come. (B and C), features from the ICS data set indicate production of the cytokines listed and no production of any measured cytokines not listed.