Influenza severity is determined by the interplay between the virus and the host response. Previously, we identified a three-pronged lung gene expression signature that predicted severe influenza outcomes in mice. We hypothesized that this signature resulted from differential activation or infiltration of immune cells in the lungs during mild versus severe influenza infection. In order to build a comprehensive model of lung immune dynamics in response to influenza virus infections, we employed tissue deconvolution and linear regression to model the lung immune cell changes that occurred during mild or severe influenza virus infections. Modeling accurately predicted known macrophage dynamics that occur during influenza infection in vivo, and further predicted a novel population that transcriptionally resembled small serosal macrophages. This macrophage population was predicted to be present in the lungs of animals that recovered from influenza infection and absent from the lungs of those that succumbed. Overall, we provide a multidimensional analysis that delineates how immune cell responses relate to immunopathological consequences of influenza infection. We also demonstrate the utility of bulk tissue gene expression data in generating hypotheses about immune cell dynamics.
Learning Objectives:
1. Understand how different lung immune cell populations impact influenza severity
2. Understand how gene expression data can be used to generate hypotheses about immune cell dynamics