Gut Microbiota and Metabolome Description of Antibiotic-Treated Neonates From Parturients With Intrauterine Infection

Abstract

Intrauterine infection is linked to adverse pregnancy outcomes in pregnant women. Neonates from parturients with intrauterine infection are usually treated with antibiotics, but their gut microbiota and metabolome are seldom studied. In this study, we collected fecal samples from antibiotic-treated neonates of parturients with intrauterine infection (intrauterine infection group), parturients with non-intrauterine infection (antibiotic group), and untreated neonates of healthy parturients (control group). 16S rRNA gene sequencing and untargeted metabolomics analyses were performed. Our results revealed that the α-diversity of intrauterine infection group differed from that of control group. There were significant differences in β-diversity between intrauterine infection group and control group, between antibiotic group and the control group, but there was no difference between the intrauterine infection and antibiotic groups, implying that antibiotic use has an obvious effect on β-diversity and that the effects of intrauterine infection on β-diversity cannot be identified. Enterococcus was more abundant in intrauterine infection and antibiotic groups than in control group. Gut metabolite differences in intrauterine infection group and antibiotic group (only in negative ion mode) from control group were observed, but no difference between intrauterine infection group and antibiotic group was observed. N-formyl-L-methionine was the most discriminant metabolite between intrauterine infection group and control group. Primary and secondary bile acid biosynthesis, bile secretion, and cholesterol metabolism pathways were altered, and the abundances of bile acids and bile salts were altered in intrauterine infection group compared with control group. Alterations in cholesterol metabolism, arginine biosynthesis and bile secretion pathways were observed both in intrauterine infection and antibiotic groups, which might be caused by the use of antibiotics. In conclusion, we provided a preliminary description of the gut microbiota and gut metabolites in antibiotics-treated neonates from intrauterine infection parturients. Our findings did not show intrauterine infection has a separate role in neonatal gut microbiota dysbiosis, while supporting the idea that antibiotics should be used with caution during neonatal therapy.

Keywords: antibiotic; gut microbiota; intrauterine infection; metabolome; neonate.

Figure 1

(A) Gut microbiota α-diversity of three groups. PCoA of gut microbiota of three groups based on the Bray–Curtis distance (B) and Euclidean distance (C). PCoA of the gut microbiota based on the Bray–Curtis distance between the intrauterine infection group and antibiotic group (D), between the intrauterine infection group and control group (E), and between the antibiotic group and control group (F). PCoA of the gut microbiota based on the Euclidean distance between the intrauterine infection group and antibiotic group (G), between the intrauterine infection group and control group (H), and between the antibiotic group and control group (I). “*” means P value smaller than 0.05.

Figure 2

(A) Relative abundance of the top five phyla among the three groups, and the significantly different phyla among the three groups. (B) Relative abundance of the top ten genera among the three groups, and the log₁₀ relative abundance of significantly different genera among the three groups. “*” means P or FDR-adjusted P values smaller than 0.05; “**” means P or FDR-adjusted P values smaller than 0.01; “***” means P or FDR-adjusted P values smaller than 0.001.

Figure 3

Heatmap of the correlation coefficients between the top ten genera and neonatal white blood cell and neutrophil counts. Spearman’s correlation coefficients are given in the heatmap, and significant correlations are represented by ‘***’ with FDR corrected P < 0.001.

Figure 4

OPLS-DA models of the metabolites between the intrauterine infection group and control group in positive ion mode (A) and negative ion mode (B). The variable importance in the projection generated in the OPLS-DA analysis between the intrauterine infection group and control group in positive and negative ion modes (C). (D) Log₁₀ fold-change of the metabolites involved in vitamin digestion and absorption, retinol metabolism, primary bile acid biosynthesis, secondary bile acid biosynthesis, cholesterol metabolism, bile secretion, cysteine and methionine metabolism, arginine biosynthesis, and the alanine, aspartate, and glutamate metabolism pathways. (E) Enriched pathways between the intrauterine infection group and control group in positive and negative ion modes. (F) Abundance of dehydrocholic acid, glycodeoxycholic acid, and taurodeoxycholate between the intrauterine infection group and control group. “*” meansP values smaller than 0.05; “**” means P values smaller than 0.01; “***” means P values smaller than 0.001.

Figure 5

OPLS-DA models of the metabolites between the antibiotic group and control group in positive ion mode (A) and negative ion mode (B). The variable importance in the projection generated in the OPLS-DA analysis between the antibiotic group and control group in negative ion mode (C). (D) Log₁₀ fold-change of the metabolites involved in cholesterol metabolism, plant hormone signal transduction, arginine biosynthesis, bile secretion, and choline metabolism in cancer pathways. (E) Enriched pathways between the antibiotic group and control group in negative ion mode. “*” meansP values smaller than 0.05; “**” means P values smaller than 0.01; “***” means P values smaller than 0.001.

Figure 6

Heatmap of correlation coefficients between the top ten genera and metabolites involved in primary bile acid biosynthesis, secondary bile acid biosynthesis, and bile secretion pathways. Significant correlations are represented by ‘*’ with FDR corrected P < 0.05, ‘**’ with FDR corrected P < 0.01, and ‘***’ with FDR corrected P < 0.001.