Metagenomic analysis of the effects of toll-like receptors on bacterial infection in the peritoneal cavity following cecum ligation and puncture in mice

Abstract

Objective: To establish the composition of bacteria in mice following cecum ligation and puncture (CLP) through metagenomic analysis and investigate the role of TLRs on the composition of bacteria.

Methods: Total DNA extraction was done from the ascites, blood, and fecal samples from C57BL/6 mice sacrificed at 0, 4, 8, and 16 h, as well as from Tlr2-/-, Tlr4-/-, Tlr5-/-, and NF-κB-/-mice sacrificed at 16 h following CLP. Amplification of the V3-V4 regions of the bacterial 16S rRNA genes by PCR and the Illumina MiSeq sequencer was used for deep sequencing. Hierarchical clustering of the isolates was performed with Ward's method using Euclidean distances. The relative abundance according to operational taxonomic unit (OTU) number or taxa was used to compare the richness among subgroups in the experiments.

Results: There were 18 taxa that had significantly different abundances among the different samples of the C57BL/6 mice at 16 h following CLP. Various dynamic changes in the infectious bacteria inside the peritoneal cavity after CLP were found. While knockout of Tlr5 and NF-κB impaired the ability of bacterial clearance inside the peritoneal cavity for some kinds of bacteria found in the C57BL/6 mice, the knockout of Tlr4 enhanced clearance for other kinds of bacteria, and they presented excessive abundance in the peritoneal cavity despite their scarce abundance in the stool.

Conclusion: NF-κB and TLRs are involved in bacterial clearance and in the expression pattern of the bacterial abundance inside the peritoneal cavity during polymicrobial infection.

Fig 1. The microbiota composition of phyla in the samples of ascites, blood, and stool of the sham-operated C57BL/6 mice (0 h) and of the C57BL/6 mice receiving CLP for 4 h, 8 h, and 16 h.

Fig 2. Hierarchical clustering of 18 bacterial taxa that had significantly different abundances among different samples of ascites, blood, and stool from the metagenomic analysis in the C57BL/6 mice.

Fig 3. These bacteria can be divided into four groups according to their pattern of abundance across different samples at 16 h after cecum ligation and puncture (CLP) in C57BL/6 mice.

Group A, the abundance of the bacterium between ascites and stool was similar but was much lower in the blood; Group B, the abundance of the bacterium in the ascites was much higher, regardless of a much lower abundance in the blood and stool; Group C, the abundance of the bacterium was much higher in the blood, but not in the ascites or stool; Group D, the abundance of the bacterium in the stool was high, but was lower in the ascites and blood. * indicated a significant different OUT number than those in the ascites.

Fig 4. Time-dependent expression bacterial 16s rRNA from the metagenomic analysis of the isolated ascites, blood, and stool samples of C57BL/6 mice 0, 4, 8, and 16 h after CLP experiment.

* indicated a significant different OUT number than those in the same sample (ascites, blood, or stool) from the sham-control C57BL/6 mice, which were indicated as 0 h.

Fig 5

(A) The microbiota composition of phyla in the stool samples of the Tlr2^–/–, Tlr4^–/–, Tlr5^–/–, and NF-κB^–/–mice; (B), Hierarchical clustering of Groups B and C bacterial taxa from the ascites and stool samples from the metagenomic analysis in the Tlr2^–/–, Tlr4^–/–, Tlr5^–/–, and NF-κB^–/–mice. Quantity of 16S rRNA in the blood was not enough for the metagenomic analysis in the Tlr and NF-κB knockout mice.