The impact of zinc pre-exposure on ciprofloxacin resistance development in E. coli

Abstract

Introduction: Antimicrobial resistance (AMR) is a global health crisis that is predicted to worsen in the coming years. While improper antibiotic usage is an established driver, less is known about the impact of other endogenous and exogeneous environmental factors, such as metals, on AMR. One metal of interest is zinc as it is often used as a supplement for diarrhea treatment prior to antibiotics.

Materials and methods: Here, we probed the impact of zinc on ciprofloxacin resistance in E. coli via altering zinc exposure time and order. We found that the order of exposure to zinc impacted resistance development. These impacted samples then underwent whole genome and RNA sequencing analysis.

Results: Zinc pre-exposure led to a subsequent acceleration of ciprofloxacin resistance. Specifically, we saw that 5 days of zinc pre-exposure led samples to have nearly a 4× and 3× higher MIC after 2 and 3 days of subinhibitory antibiotics, respectively, compared to samples not pre-exposed to zinc, but only if ciprofloxacin exposure happened in the absence of zinc. Additionally, for samples that underwent the same pre-exposure treatment, those exposed to a combination of zinc and ciprofloxacin saw delayed ciprofloxacin resistance compared to those exposed to only ciprofloxacin resulting in up to a 5× lower MIC within the first 2 days of antibiotic exposure. We did not observe any genetic changes or changes in antibiotic tolerance in cells after zinc pre-exposure, suggesting changes in gene expression may underlie these phenotypes.

Discussion: These results highlight the need to reexamine the role of zinc, and supplements more broadly, on antibiotic resistance evolution.

Keywords: Escherichia coli (E. coli); antimicrobial resistance (AMR); ciprofloxacin (CIP); conflict settings; environmental pollution; zinc.

FIGURE 1

Change in ciprofloxacin MIC overtime. (A) E. coli was pre-exposed to zinc for 3 or 5 days. On day zero, pre-exposed cells along with wild type, zero-day pre-exposed cells, were exposed to subinhibitory ciprofloxacin and tested daily for resistance development by assessing their MICs. (B) After 3 days of exposure to only subinhibitory concentrations of ciprofloxacin, E. coli pre-exposed to 0.5 mM Zn for 5 days showed higher MICs that were significantly different compared to samples not pre-exposed to Zn 2 and 3 days after exposure to ciprofloxacin. While a significant difference was observed for the 5-day Zn pre-exposure samples on days two as indicated by the double dagger (p = 0.0061) and three as indicated by the single dagger (p = 0.0176), no significant difference was ever observed for the 3-day pre-exposure samples. (C) Meanwhile, samples exposed to a combination of subinhibitory ciprofloxacin and 0.5 mM Zn after the initial 0.5 mM Zn pre-exposure period saw no significant differences between pre-exposed conditions and the non-pre-exposed condition. Values are plotted in GraphPad Prism as the average of N = 4 biological replicates with error bars indicating standard deviations for each data point. All comparisons are non-significant unless otherwise noted based on ordinary one-way ANOVA with multiple comparisons via Dunnett’s test using the no zinc pre-exposure as the control.

FIGURE 2

Impact of mutations from prior Zn exposure and Zn re-exposure on ciprofloxacin development. (A) All samples previously pre-exposed to Zn before being frozen, showed no significant difference in the development of resistance to ciprofloxacin after being exposed to subinhibitory ciprofloxacin after being thawed. An ordinary one-way ANOVA with multiple comparisons (Dunnett’s test) was performed between each of the three conditions and a control condition that was never exposed to Zn. Data represents the average MIC where N = 4 with error bars representing the SD. (B) To further examine if the earlier development of resistance among the Zn pre-exposed samples is a result of the specific mutations observed post Zn exposure, the same three samples previously pre-exposed and sequenced and a fresh wild type sample were thawed and pre-exposed to Zn for 5 days before being exposed to ciprofloxacin at a concentration of 40% of the MIC. An ordinary one-way ANOVA with multiple comparisons (Dunnett’s test) was performed between each of the four conditions and a new control condition that was never pre-exposed to Zn. Pre-exposing the samples to Zn post freezing was the only experimental difference between (A) and (B). Here, 5 days of Zn pre-exposure led to an earlier onset of elevated MIC levels. Significant differences were observed on day one and two as indicated by the dagger, and double dagger, respectively. After 1 day of exposure to ciprofloxacin all prior pre-exposed conditions saw significantly higher MICs compared to the non-pre-exposed condition (Sample #3 p = 0.0009, Sample #2 p = 0.0092, Sample #1 p = 0.0010). The newly pre-exposed Zn exposure condition saw an elevated MIC but was not significant compared to the no pre-exposure condition (p = 0.0536). By day two, this newly Zn pre-exposed sample was significantly higher (p = 0.0495) with differences still observed for Sample #3 (p = 0.0011) and Sample #1 (p = 0.0112). Although Sample #2 was still elevated, it was no longer significantly elevated (p = 0.1055). Statistical results were found by one-way ANOVA with multiple comparisons (Dunnett’s test). A full description of Samples #1, #2, and #3 can be found in Table 1.