Abstract We recently identified immunotherapy responder and nonresponder human microbiota compositions in the murine model of glioma. In this study, we sought to understand how human microbiota in mice influence the immune landscape in glioma. To test this, we utilized our novel humanized microbiome (HuM) mouse model (human microbiota transplanted into gnotobiotic mice) to determine the effect of immunotherapy in a mouse model of glioma utilizing single-cell RNA sequencing and 16S rRNA microbiome sequencing. We focused on two HuM mouse lines in this study: HuM1 (nonresponder) and HuM2 (responder). To confirm microbiota differences,16S rRNA sequencing revealed significant differences between alpha and beta diversity between HuM1 and HuM2 groups. HuM2 responder mice also displayed a significant increase in the relative abundances of Odoribacter, Muribaculaceae, and Ruminococcus genera compared to HuM1 mice. HuM1 and HuM2 mice were then injected with GL261 glioma cells into the brains, treated with one dose of anti-PD-1 or isotype control on day 13, and all mice were euthanized on day 14 for scRNA-sequencing of CD45+ cells in the tumors. We identified 21 clusters of immune cells in the tumors. Furthermore, we found significant differences between HuM1 and HuM2 mice in isotype control conditions including higher percentages of immature B cells, neutrophils, and microglia in HuM2 mice. In anti-PD-1 treated tumors, there was a robust increase in the percentage of innate immune cells in the tumor following, specifically the monocyte/macrophage (MM1; Cxcl10+) population in both HuM1 and HuM2 mice. However, the HuM2 MM1 cluster showed an increase in gene expression of Nos2 and decrease in Arg1, indicating that the MM1 population in HuM2 mice are more inflammatory (anti-tumor) compared to HuM1. Ultimately, our data demonstrate that the gut microbiome influences the tumor-immune landscape, and harnessing positive microbial compositions remains a viable translational avenue to pursue.
Read full abstract