Key hepatic metabolic pathways are altered in germ-free mice during pregnancy

Abstract

Pregnancy is associated with metabolic changes to accommodate the mother and her growing fetus. The microbiome has been shown to modulate host metabolism of endogenous and exogenous substances. However, the combined effects of pregnancy and the microbiome on host metabolism have not been investigated. The objective of this study was to investigate how the microbiome affects overall hepatic metabolic processes during pregnancy. We assessed these changes within 4 groups of C57BL/6 mice: conventional non-pregnant, conventional pregnant, germ-free non-pregnant, and germ-free pregnant mice. We performed RNA-seq analysis on liver tissues and LC-MS/MS analysis of the plasma to assess the effects of pregnancy and the microbiome on hepatic transcriptome and untargeted plasma metabolome to describe metabolic changes as results of both pregnancy and lack of microbiome. By integrating transcriptomics and metabolomics data, we identified eight metabolic pathways that were significantly enriched for differentially expressed genes associated with pregnancy in both conventional and germ-free mice. Notably, of the eight pathways, 4 pathways (retinol metabolism, arachidonic acid metabolism, linoleic acid metabolism, and steroid hormone biosynthesis) which are all critical for normal pregnancy and fetal development were affected by the germ-free status in pregnant mice, but not at all in non-pregnant mice, indicating that the alterations in these four pathways caused by the lack of microbiome are unique for pregnancy. These results provide novel insight into the role of the microbiome in modulating host metabolic processes critical for maternal health and fetal development during pregnancy.

Fig 1. The number of differentially expressed genes and differentially produced metabolites between various comparison mouse groups.

Differentially expressed hepatic genes (A) and differentially produced metabolites in maternal plasma (B) between CVP and CVNP, GFP and GFP, GFNP and CVNP, and GFP and CVP mice. The number of CVNP, CVP, GFNP and GFP mice used was 6, 5, 6, and 5, respectively. Inclusion criteria for genes and metabolites were FDR of 0.1 and a minimum 2-fold change in at least one comparison group.

Fig 2. Heatmap of hepatic genes with at least 2-fold change and FDR < 0.1.

Inclusion criteria for the genes presented in this heatmap were FDR of 0.1 or less and a minimum 2-fold change in at least one comparison group between CVP and CVNP, GFP and GFP, GFNP and CVNP, and GFP and CVP mice. CVNP, conventional non-pregnant mice; CVP, conventional pregnant mice; GFNP, germ-free non-pregnant mice; GFP, germ-free pregnant mice.

Fig 3. Heatmap of metabolites in maternal plasma with at least 2-fold change and FDR < 0.1.

Inclusion criteria for the metabolites presented in this heatmap were FDR of 0.1 of less and a minimum 2-fold change in at least one comparison group between CVP and CVNP, GFP and GFP, GFNP and CVNP, and GFP and CVP mice. CVNP, conventional non-pregnant mice; CVP, conventional pregnant mice; GFNP, germ-free non-pregnant mice; GFP, germ-free pregnant mice.

Fig 4. Heatmap of False Discovery Rates (FDRs) of top metabolic pathway hits among all comparison groups.

Filtering criterion was 1) genes and metabolites metabolites with FDR of 0.1 of less and a minimum 2-fold change in at least one comparison group between CVP and CVNP, GFP and GFP, GFNP and CVNP, and GFP and CVP mice; and 2) a minimum of 1 gene and 1 metabolite hit per pathway with FDR < 0.1 and a minimum 2-fold change in the comparison groups between CVP and CVNP, GFP and GFP, GFNP and CVNP, and GFP and CVP mice. Those that failed to meet this criterion was labeled as P = 1. CVNP, conventional non-pregnant mice; CVP, conventional pregnant mice; GFNP, germ-free non-pregnant mice; GFP, germ-free pregnant mice.