Impact of suppression of the SOS response on protein expression in clinical isolates of Escherichia coli under antimicrobial pressure of ciprofloxacin

Abstract

Introduction/objective: Suppression of the SOS response in combination with drugs damaging DNA has been proposed as a potential target to tackle antimicrobial resistance. The SOS response is the pathway used to repair bacterial DNA damage induced by antimicrobials such as quinolones. The extent of lexA-regulated protein expression and other associated systems under pressure of agents that damage bacterial DNA in clinical isolates remains unclear. The aim of this study was to assess the impact of this strategy consisting on suppression of the SOS response in combination with quinolones on the proteome profile of Escherichia coli clinical strains.

Materials and methods: Five clinical isolates of E. coli carrying different chromosomally- and/or plasmid-mediated quinolone resistance mechanisms with different phenotypes were selected, with E. coli ATCC 25922 as control strain. In addition, from each clinical isolate and control, a second strain was created, in which the SOS response was suppressed by deletion of the recA gene. Bacterial inocula from all 12 strains were then exposed to 1xMIC ciprofloxacin treatment (relative to the wild-type phenotype for each isogenic pair) for 1 h. Cell pellets were collected, and proteins were digested into peptides using trypsin. Protein identification and label-free quantification were done by liquid chromatography-mass spectrometry (LC-MS) in order to identify proteins that were differentially expressed upon deletion of recA in each strain. Data analysis and statistical analysis were performed using the MaxQuant and Perseus software.

Results: The proteins with the lowest expression levels were: RecA (as control), AphA, CysP, DinG, DinI, GarL, PriS, PsuG, PsuK, RpsQ, UgpB and YebG; those with the highest expression levels were: Hpf, IbpB, TufB and RpmH. Most of these expression alterations were strain-dependent and involved DNA repair processes and nucleotide, protein and carbohydrate metabolism, and transport. In isolates with suppressed SOS response, the number of underexpressed proteins was higher than overexpressed proteins.

Conclusion: High genomic and proteomic variability was observed among clinical isolates and was not associated with a specific resistant phenotype. This study provides an interesting approach to identify new potential targets to combat antimicrobial resistance.

Keywords: Enterobacteriaceae; Escherichia coli; SOS response; antimicrobial resistance; proteome profile.

Figure 1

Number of significantly over- and underexpressed proteins upon deletion of recA for ATCC 25922 and the five clinical isolates, with at least a 2-fold change in their relative abundance (log2 fold change >1 for increased and <−1 for decreased). Proteins were considered significantly changed with an adjusted p-value (=q-value) < 0.05. All experiments were done using 1xMIC concentration of ciprofloxacin.

Figure 2

Overlap between differentially expressed proteins (log2 fold change >1 and <−1, p-value < 0.05) following exposure to ciprofloxacin (1xMIC relative to each wild-type) between isolates with suppressed SOS response relative to wild-type and the control strain in the same conditions. Venn diagrams show the overlap. Numbers on the diagram refer to the number of proteins with significantly altered expression levels. Susceptible phenotype: FI 10; LLQR phenotypes: FI 4 and FI 20; Resistant phenotype: FI 19 and FI 24.

Figure 3

Proteome profile by strain. Orange: Proteins with log2 fold change (FC) < −1; Blue: Proteins with a log2 FC > 1; Green: Proteins with log2 FC = 0 to −1; Yellow: Proteins with log2 FC = 0 to 1. Black: no significant proteins (p > 0.05). Labeled proteins with log2 FC between >2.5 and < −2.5. wt, wild-type.

Figure 4

Heatmap of relative protein expression based on label-free quantification by liquid chromatography/mass spectrometry (LC-MS). The main function associated with each protein is shown.