Metagenomic and metabolomic analyses reveal synergistic effects of fecal microbiota transplantation and anti-PD-1 therapy on treating colorectal cancer

Abstract

Anti-PD-1 immunotherapy has saved numerous lives of cancer patients; however, it only exerts efficacy in 10-15% of patients with colorectal cancer. Fecal microbiota transplantation (FMT) is a potential approach to improving the efficacy of anti-PD-1 therapy, whereas the detailed mechanisms and the applicability of this combination therapy remain unclear. In this study, we evaluated the synergistic effect of FMT with anti-PD-1 in curing colorectal tumor-bearing mice using a multi-omics approach. Mice treated with the combination therapy showed superior survival rate and tumor control, compared to the mice received anti-PD-1 therapy or FMT alone. Metagenomic analysis showed that composition of gut microbiota in tumor-bearing mice treated with anti-PD-1 therapy was remarkably altered through receiving FMT. Particularly, Bacteroides genus, including FMT-increased B. thetaiotaomicron, B. fragilis, and FMT-decreased B. ovatus might contribute to the enhanced efficacy of anti-PD-1 therapy. Furthermore, metabolomic analysis upon mouse plasma revealed several potential metabolites that upregulated after FMT, including punicic acid and aspirin, might promote the response to anti-PD-1 therapy via their immunomodulatory functions. This work broadens our understanding of the mechanism by which FMT improves the efficacy of anti-PD-1 therapy, which may contribute to the development of novel microbiota-based anti-cancer therapies.

Keywords: Bacteroides; anti-PD-1 therapy; colorectal cancer; fecal microbiota transplantation; immunotherapy.

Figure 1

FMT and PD-1 antibody exerted synergistic anti-tumor effect in the CT26 tumor-bearing mice. (A) Schematic diagram of this study. (B) Survival curve of the CT26 tumor-bearing mice treated with FMT, aPD-1 or the combination. Statistical differences among four groups were examined using log-rank (Mantel-Cox) tests. Post hoc pair-wise comparisons were performed; *, p-value < 0.05; **, p-value < 0.01. (C) Tumor growth curves of the CT26 tumor-bearing mice treated with FMT, aPD-1 or the combination. Data are represented as mean ± SD (n = 10). Statistical differences were examined using Dunnett’s test; *, p-value < 0.05.

Figure 2

FMT altered the composition of gut microbiota in CT-26 tumor-bearing mice receiving anti-PD-1 therapy. (A) Principal components analysis (PCA) plot of the gut microbiota from mice. (B) Relative abundance of top 15 bacterial families in different groups. (C) LEfSe analysis showing differentially abundant bacterial species between FMT and Combo groups. (D) Heatmap showing the correlations of species significantly different between FMT and Combo groups. (E) Abundance of specific species in different groups. Data are represented as mean ± SD. *, p-value < 0.05; **, p-value < 0.01; ***, p-value < 0.001.

Figure 3

The effect of FMT and PD-1 antibody administration on gut metagenomic gene pathways. (A) Volcano plot showing differentially expressed microbial gene pathways between Combo and aPD-1 groups. (B) Abundance of specific gene pathways in different groups. Data are represented as mean ± SD. *, p-value < 0.05; **, p-value < 0.01; ***, p-value < 0.001.

Figure 4

FMT altered plasma metabolites in CT-26 tumor-bearing mice receiving anti-PD-1 therapy. (A) Venn diagrams showing number of significantly changed metabolites in each group after treatment. (B) PCA plot of metabolomic results. (C) Heatmap of differentially abundant metabolites using one-way analysis of variance. (D) The correlations between metabolites and microorganism. (E) Abundance of specific metabolites in different groups. Data are represented as mean ± SD. *, p-value < 0.05; **, p-value < 0.01; ***, p-value < 0.001.