Sex differences in susceptibility to influenza A virus infection depend on host genotype

Abstract

Infection with the respiratory pathogen influenza A virus (IAV) causes significant morbidity and mortality each year. While host genotype is thought to contribute to severity of disease, naturally occurring genetic determinants remain mostly unknown. Moreover, more severe disease is seen in women compared with men, but genetic mechanisms underlying this sex difference remain obscure. Here, using IAV infection in a mouse model of naturally selected genetic diversity, namely C57BL6/J (B6) mice carrying chromosomes (Chr) derived from the wild-derived and genetically divergent PWD/PhJ (PWD) mouse strain (B6.ChrPWD consomic mice), we examined the effects of genotype and sex on severity of IAV-induced disease. Compared with B6, parental PWD mice were completely protected from IAV-induced disease, a phenotype that was fully recapitulated in the B6.Chr16PWD strain carrying the PWD-derived allele of Mx1. In contrast, several other consomic strains, including B6.Chr3PWD and B6.Chr5PWD, demonstrated greatly increased susceptibility. Notably, B6.Chr5PWD and B6.ChrX.3PWD strains, the latter carrying the distal one-third of ChrX from PWD, exhibited increased morbidity and mortality specifically in male but not female mice. Follow up analyses focused on B6 and B6.ChrX.3PWD strains demonstrated moderately elevated viral load in B6.ChrX3PWD male, but not female mice. Transcriptional profiling demonstrated genotype- and sex-specific gene expression profiles in the infected lung, with male B6.ChrX.3 mice exhibiting the most significant changes, including upregulation of a proinflammatory gene expression program associated with myeloid cells, and altered sex-biased expression of several X-linked genes that represent positional candidates, including Tlr13 and Slc25a53. Taken together, our results identify novel loci on autosomes and the X chromosome regulating IAV susceptibility and demonstrate that sex differences in IAV susceptibility are genotype-dependent, suggesting that future genetic association studies need to consider sex as a covariate.