Description

Bistable individual variability in the immune response to seasonal influenza vaccination

Naotoshi Nakamura¹, Daigo Okada¹, Kazuya Setoh², Takahisa Kawaguchi², Maiko Narahara³, Koichiro Higasa², Yasuharu Tabara², Fumihiko Matsuda², Ryo Yamada¹*

¹Statistical Genetics, Kyoto University, Kyoto, Japan; ²Human Disease Genomics, Kyoto University, Kyoto, Japan; ³Department of Human Genetics, McGill University, Montreal, Canada *Corresponding Author

The human immune system is known to be highly variable between individuals, but it is not well understood how the variability changes over time, especially when faced with external perturbations. Here we analyzed individual variability in the immune system in a cohort of 301 Japanese volunteers who received the same trivalent inactivated influenza vaccine. Application of LAVENDER (latent axes discovery by nonparametric density estimation and multidimensional scaling reconstruction) to multidimensional flow cytometry datasets of B and T lymphocytes (taken before and 1, 7, 90 days after vaccination) uncovered two polarized immunological states in the population, which corresponded to Th1-dominant and Th2-dominant states. In the first axis of the LAVENDER space, individuals frequently switched their states with time, but in the second axis, they changed their states consistently in response to vaccination. Metabolome analysis identified lipid markers correlated with either of the above axes; GWAS analysis identified several genetic markers associated with this variability. These results suggest that LAVENDER is a useful data-driven method to extract important variability axes among multidimensional single-cell datasets.