Colicins of Escherichia coli Lead to Resistance against the Diarrhea-Causing Pathogen Enterotoxigenic E. coli in Pigs

Abstract

Gut microbes can affect host adaptation to various environment conditions. Escherichia coli is a common gut species, including pathogenic strains and nonpathogenic strains. This study was conducted to investigate the effects of different E. coli strains in the gut on the health of pigs. In this study, the complete genomes of two E. coli strains isolated from pigs were sequenced. The whole genomes of Y18J and the enterotoxigenic E. coli strain W25K were compared to determine their roles in pig adaptation to disease. Y18J was isolated from feces of healthy piglets and showed strong antimicrobial activity against W25K in vitro. Gene knockout experiments and complementation analysis followed by modeling the microbe-microbe interactions demonstrated that the antagonistic mechanism of Y18J against W25K relied on the bacteriocins colicin B and colicin M. Compared to W25K, Y18J is devoid of exotoxin-coding genes and has more secondary-metabolite-biosynthetic gene clusters. W25K carries more genes involved in genome replication, in accordance with a shorter cell cycle observed during a growth experiment. The analysis of gut metagenomes in different pig breeds showed that colicins B and M were enriched in Laiwu pigs, a Chinese local breed, but were scarce in boars and Duroc pigs. IMPORTANCE This study revealed the heterogeneity of E. coli strains from pigs, including two strains studied by both in silico and wet experiments in detail and 14 strains studied by bioinformatics analysis. E. coli Y18J may improve the adaptability of pigs toward disease resistance through the production of colicins B and M. Our findings could shed light on the pathogenic and harmless roles of E. coli in modern animal husbandry, leading to a better understanding of intestinal-microbe-pathogen interactions in the course of evolution.

Keywords: Escherichia coli; comparative genomics analysis; genome sequencing; microbe-pathogen interactions; secondary metabolite.

FIG 1

Bioinformatics-based screening of targeted bacteria with antibacterial activity. (A) Flowchart of targeted screening. (B) Numbers of genes with antibacterial activity in the pig metagenome. (C) Inhibitory effects of wild bacterial strain Y18J against the pathogen W25K.

FIG 2

Phylogeny and virulence factors present in Y18J and W25K. (A) Neighbor-joining phylogenetic tree obtained from gyrB gene sequences. The bootstrap consensus tree inferred from 500 replicates was taken to represent the evolutionary history of the taxa analyzed. Branches corresponding to partitions reproduced in less than 50% of bootstrap replicates were collapsed. (B) Virulence-associated genes were identified using VFDB in known pathogenic E. coli strains and probiotic E. coli strains. The bacteria in red are commercial probiotics. The bacteria in blue are pathogens. (C) Virulence-associated genes were identified using VFDB in W25K and Y18J. Black squares indicate the presence of factors; white squares indicate the absence of factors.

FIG 3

Homology genomic and growth curve comparisons between Y18J and W25K. (A) Circos plot showing homology regions between Y18J (including y1, y2, y3, and y4) and W25K (including w1, w2, w3, w4, w5, w6, w7, and w8). The homology regions are shaded in different colors. (B) COG function annotation of the different homology regions according to the length of homology region. The length of homology regions is represented by the letter “a.” (C) Growth curves for Y18J (wild type and mutant) and W25K. The Y18J mutant is a strain in which colicins B and M were knocked out. The optical density at 600 nm (OD₆₀₀) was measured from a starting OD₆₀₀ of 0.085. Data are means and standard deviations (SD) from three independent experiments.

FIG 4

Modeling of the microbe-microbe (W25K-Y18J) interactions in vitro. (A) Inhibitory effects of the Y18J wild-type and mutant strains at different growth times against the pathogen W25K. The antimicrobial products in the growth medium were inactivated with and without heat treatment. Con, control group treated with LB medium. (B) Inhibitory effects of W25K against Y18J at different growth times. (C) Y18J (wild-type or mutant) and W25K colonies at different coculture times. The colony without fluorescence is Y18J, and the fluorescent colony is W25K with a GFP expression plasmid. (D) Ratios of numbers of colonies of W25K and Y18J.

FIG 5

Levels of pathogenic and antagonistic factors from different origins. (A) Boars; (B) Duroc pigs; (C) Laiwu pigs. The pathogenic factors encoded by W_1 and W_2 in W25K are a heat-stable enterotoxin and a heat-labile enterotoxin, respectively. The antagonistic factors encoded by Y_1 and Y_2 in Y18J are colicin B and colicin M, respectively. Statistical analyses were performed using the Mann-Whitney U test (R version 3.5; a P value of <0.01 was considered statistically significant).