Inactivation efficiency of plasmid-encoded antibiotic resistance genes during water treatment with chlorine, UV, and UV/H₂O₂

Highlights

• Inactivation kinetics of ARB, e- and i-ARGs by chlorine, UV, UV/H₂O₂ were determined.

• Damages of ARGs occurred much more slowly than the inactivation of ARB (E. coli).

• Damages of i-ARGs were slower than e-ARGs due to protections by cellular components.

• 4-log ARG reductions require 40–70 (mg × min)/L for chlorine and 50–130 mJ/cm² for UV.

• ARG removals in a wastewater effluent were predictable from the ARG damage kinetics.

Abstract

This study assessed the inactivation efficiency of plasmid-encoded antibiotic resistance genes (ARGs) both in extracellular form (e-ARG) and present within Escherichia coli (intracellular form, i-ARG) during water treatment with chlorine, UV (254 nm), and UV/H₂O₂. A quantitative real-time PCR (qPCR) method was used to quantify the ARG damage to amp^R (850 bp) and kan^R (806 bp) amplicons, both of which are located in the pUC4K plasmid. The plate count and flow cytometry methods were also used to determine the bacterial inactivation parameters, such as culturability and membrane damage, respectively. In the first part of the study, the kinetics of E. coli inactivation and ARG damage were determined in phosphate buffered solutions. The ARG damage occurred much more slowly than E. coli inactivation in all cases. To achieve 4-log reduction of ARG concentration at pH 7, the required chlorine exposure and UV fluence were 33–72 (mg × min)/L for chlorine and 50–130 mJ/cm² for UV and UV/H₂O₂. After increasing pH from 7 to 8, the rates of ARG damage decreased for chlorine, while they did not vary for UV and UV/H₂O₂. The i-ARGs mostly showed lower rates of damage compared to the e-ARGs due to the protective roles of cellular components against oxidants and UV. The contribution of OH radicals to i-ARG damage was negligible in UV/H₂O₂ due to significant OH radical scavenging by cellular components. In all cases, the ARG damage rates were similar for amp^R versus kan^R, except for the chlorination of e-ARGs, in which the damage to amp^R occurred faster than that to kan^R. Chlorine and UV dose-dependent ARG inactivation levels determined in a wastewater effluent matrix could be reasonably explained by the kinetic data obtained from the phosphate buffered solutions and the expected oxidant (chlorine and OH radicals) demands by water matrix components. These results can be useful in optimizing chlorine and UV-based disinfection systems to achieve ARG inactivation.

Introduction

The rise of antibiotic resistance is a global issue for human and animal health, as this resistance lowers the efficacy of antibiotics and the treatability of infectious diseases (WHO, 2014). Antibiotics can select antibiotic-resistant bacteria (ARB) carrying genes (ARGs) responsible for antibiotic-resistance mechanisms (Keen and Montforts, 2012). The use of antibiotics can inevitably induce ARB, while the misuse or overuse of antibiotics is related to the prevalence of ARB in clinical and animal agriculture domains (Keen and Montforts, 2012). There has been growing concern regarding the presence of ARB and their resistance genes in aquatic environments, as antibiotic resistance can be disseminated by sharing ARGs among bacterial communities (Berendonk et al., 2015, Pruden, 2014). Mobile genetic elements, such as plasmids, integrons, and transposons, are involved in ARG sharing through horizontal gene transfer (HGT) processes, such as conjugation (cell-to-cell contact), transduction (virus mediated), and transformation (the uptake of exogenous genetic materials) (Keen and Montforts, 2012). Considering the HGT mechanisms, in both free form (extracellular) and within host cells (intracellular), ARGs are considered to be contaminants of concern in natural and engineered water systems (Berendonk et al., 2015, Dodd, 2012, Pruden, 2014).

Municipal wastewaters have been identified as one of the major sources of ARB and ARGs in aquatic environments (Rizzo et al., 2013). Conventional biological wastewater treatments do not fully eliminate the ARB and ARGs (Chen and Zhang, 2013, Rizzo et al., 2013), and may even lead to selective increases of multi-resistant bacterial species (Czekalski et al., 2012). Disinfection of wastewater effluent with chlorine or UV irradiation has been widely practiced to protect the microbiological quality of drinking water sources and sensitive receiving waters (Jacangelo and Trussell, 2002). Enhanced wastewater treatment with ozone or UV/H₂O₂ has also received increasing attention in many industrialized countries and has been implemented in some countries to eliminate various trace organic contaminants exerting adverse ecological effects, such as hormones and pharmaceuticals (Eggen et al., 2014). Under these circumstances, there has been growing interest in the efficiency of utilizing conventional and advanced wastewater disinfection processes as barriers against antibiotic-resistance dissemination by lowering the levels of ARB and ARGs.

Disinfection (oxidation) processes, such as chlorine, ozone, UV, and UV/H₂O₂, have been examined to inactivate ARB and ARGs in wastewater effluent matrixes in laboratory, pilot, and full-scale studies (Alexander et al., 2016, Chen and Zhang, 2013, Czekalski et al., 2016, Dodd, 2012, Ferro et al., 2017, Huang et al., 2013, Lüddeke et al., 2015, McKinney and Pruden, 2012, Pak et al., 2016, Sousaa et al., 2017, Yuan et al., 2015, Zhang et al., 2015, Zhuang et al., 2015 and references cited therein). As a general trend, wastewater disinfection processes under typical treatment conditions can significantly reduce overall ARB levels (e.g., by more than several logs), while at the same time resulting in ARB selection (i.e., an increase of the relative proportions of ARB amongst the surviving bacterial cells) (Alexander et al., 2016, Lüddeke et al., 2015). This complicates the assessment of disinfection's efficiency in lowering the potential of antibiotic-resistance dissemination. The reduction of ARG levels, which was determined by the quantitative polymerase chain reaction (qPCR) method, was usually much less significant compared to that of ARB, indicating the more resistant nature of DNA versus bacterial cells themselves. Overall, wastewater disinfection processes should be carefully evaluated and further optimized to achieve sufficient levels of ARG inactivation.

The inactivation efficiency of ARGs during wastewater disinfection can depend on various factors. First, the type of ARGs can be important. Whether ARGs are present within extracellular or intracellular DNA (hereafter denoted as e-ARG and i-ARG) may affect the efficacy of ARG damage by oxidants or UV (Dodd, 2012). The effect of ARG type on its inactivation efficiency via different disinfection processes is poorly understood. Second, different degrees of ARG inactivation can be measured depending on which analytical methods are utilized. In recent years, qPCR-based methods have been widely utilized to quantify ARG damage. Most previous studies used the qPCR method optimized for short amplicons (e.g., 100–200 bp). However, the short amplicon-based qPCR methods may underestimate ARG damage and associated loss of ARG biological function (Chang et al., 2017, McKinney and Pruden, 2012, Süβ et al., 2009). Third, the operational conditions of disinfection processes (e.g., oxidant or UV dose) and water qualities (e.g., dissolved organic matter-DOM) can significantly affect ARG inactivation levels. Most previous studies were conducted under the specific conditions created by disinfection processes and wastewater matrixes, which limits opportunities for inter-comparison and generalization of the results. In this respect, it is recommended to determine principle-based kinetic parameters for ARG inactivation in clean water matrixes (e.g., no oxidant demand). Only a limited number of studies have carefully determined generally applicable rate constants for ARG inactivation, such as for ozonation (Czekalski et al., 2016) and UV disinfection (Chang et al., 2017, McKinney and Pruden, 2012). In summary, despite significant recent research progress, it is still difficult to predict the inactivation levels of diverse types of ARGs during wastewater disinfection or oxidation at differing operational process or water matrix conditions.

In the present study, we have determined and compared the inactivation efficiencies of plasmid-encoded antibiotic-resistance genes in e-ARG and i-ARG forms (the latter as E. coli host) in bench-scale disinfection experiments with chlorine and UV_254nm (hereafter UV) utilized as conventional treatment methods and UV/H₂O₂ utilized as an advanced treatment method. It was of interest whether OH radicals (OH), as the main oxidant in the UV/H₂O₂ process, could enhance the inactivation efficiency of ARGs. An extended amplicon-length qPCR method was used to determine the damage to two differing ARG amplicons located in plasmid pUC4K (i.e., amp^R with 850 bp and kan^R with 806 bp). To obtain more generally applicable inactivation kinetic parameters, the first part of the experiments was carried out in synthetic phosphate buffered solutions, in which plasmid or E. coli was separately spiked and treated as a function of chlorine exposure (= C × T, chlorine concentration × contact time) or UV fluence (= I × T, UV intensity × irradiation time). The inactivation levels of the target ARGs were determined, together with the bacterial inactivation parameters, such as culturability and membrane damage in some cases. In the second part of the experiments, the plasmid or E. coli was separately spiked to filtered, secondary wastewater effluent, where the inactivation levels of target ARGs were determined at varying oxidant doses or UV fluences. The results were discussed in relation to the prediction of ARG inactivation efficiency during water and wastewater disinfection, based on the obtained kinetic parameters.