Genome-Wide Screens Identify Genes Responsible for Intrinsic Boric Acid Resistance in Escherichia coli

Abstract

Boric acid (BA) has antimicrobial properties and is used to combat bacterial infections, including Enterobacteria. However, the molecular mechanisms and cellular responses to BA are still unknown. This genomics study aims to provide new information on the genes and molecular mechanisms related to the antimicrobial effect of BA in Escherichia coli. The Keio collection of E. coli was used to screen 3985 single-gene knockout strains in order to identify mutant strains that were sensitive or hypersensitive to BA at certain concentrations. The mutant strains were exposed to different concentrations of BA ranging from 0 to 120 mM in LB media. Through genome-wide screens, 92 mutants were identified that were relatively sensitive to BA at least at one concentration tested. The related biological processes in the particular cellular system were listed. This study demonstrates that intrinsic BA resistance is the result of various mechanisms acting together. Additionally, we identified eighteen out of ninety-two mutant strains (Delta_aceF, aroK, cheZ, dinJ, galS, garP, glxK, nohA, talB, torR, trmU, trpR, yddE, yfeS, ygaV, ylaC, yoaC, yohN) that exhibited sensitivity using other methods. To increase sensitivity to BA, we constructed double and triple knockout mutants of the selected sensitive mutants. In certain instances, engineered double and triple mutants exhibited significantly amplified effects. Overall, our analysis of these findings offers further understanding of the mechanisms behind BA toxicity and intrinsic resistance in E. coli.

Keywords: Escherichia coli; Boric acid; Gene; Genome-wide screen.

Fig. 1

Effect of 50 mM BA on K-12 BW25113ΔtorR ΩKm^R mutant E. coli strain. LB plates without any BA (left panel) and LB plates with 50 mM BA (right panel) were used to culture 96 mutants from the plate in Keio collection. Among them, ΔtorR mutant failed to grow on the plate with 50 mM BA. Further investigation on pure colonies confirmed its hypersensitivity

Fig. 2

Spot tests to determine the BA sensitivity of the 18 mutant strains. The mutants were subjected to sequential spotting under different concentrations of BA, as described in the “Materials and Methods” section. Five different dilutions of the bacterial culture (1/1, 1/2, 1/4, 1/8, 1/16) and 13 different BA concentrations were used. The growth of the mutants and the wild-type strain is illustrated, with the particular strain name listed below each picture

Fig. 3

Strain numbers of subsystems. Each gene hit, which is associated with a BA-sensitive mutant strain, is mapped to specific cellular processes. It is important to note that gene hits may be associated with more than one process. The gene list was evaluated using Omics Dashboard (Pathway Tools) and the EcoCyc databases

Fig. 4

Enrichment of the corresponding genes in BA-sensitive mutants g. Clusters with three or more genes were considered (p value < 0.05). The DAVID gene functional classification (version 6.8) database was employed to assess the enrichment using E. coli K-12 genome. Only the particular processes with three or more genes were included

Fig. 5

String analysis of the BA-sensitive gene hits. The predicted functional associations between the genes are shown by lines, where the number of the lines represents the linkage based on curated databases, gene-fusions, co-expression, and experimental evidence. STRING (version 10.5) with medium confidence score of 0.4 was used