Disinfection processes are key barriers to the spread of viable antibiotic resistant bacteria (ARB) within environmental systems and human populations. However, even if ARB cells are inactivated during disinfection, antibiotic resistance genes (ARGs) within cell debris may remain biologically active and able to transfer resistance traits to non-resistant bacterial populations via horizontal gene transfer processes (e.g., natural transformation). This project utilized real-time quantitative polymerase chain reaction (qPCR) and culture-based DNA activity assays to determine the effectiveness of disinfectants used in (waste)water and healthcare practice for not only inactivating ARB cells, but also degrading (i.e., chemically modifying) and deactivating (i.e., eliminating the biological activities of) ARGs associated with extra- and intracellular bacterial DNA. The ARGs and ARB strains investigated included the model ARG blt from Bacillus subtilis, and the clinically-relevant ARGs mecA, vanA, tetA, ampC, bla_NDM, and bla_KPCfrom Staphylococcus aureus, Enterococcus faecium, Escherichia coli, Pseudomonas aeruginosa, and Klebsiella pneumoniae, respectively. 

Kinetics measurements highlighted a wide range of ARG reactivities toward different disinfectants (Figure 1), with reactivities declining in the order hydroxyl radical (*OH) >> ozone (O₃) > free chlorine (HOCl) >> chlorine dioxide (ClO₂) > monochloramine (NH₂Cl) > hydrogen peroxide (H₂O₂), and generally high effectiveness observed for UV_254nm light. Ethanol (EtOH) and phenol were completely ineffective in degrading ARGs. For the investigated disinfectants, reactivity of the DNA sequence comprising a given ARG increased with either (i) # of AT+GC bps or (ii) # of intrastrand 5'-bipyrimidine-3' or 5'-GG-3' doublets per sequence length. The observed dependency of ARG reactivities on sequence length and/or nucleotide content enabled development of a modeling approach (validated using the above ARGs) that can be used to predict reactivities of any ARG toward these disinfectants provided the ARG's sequence is known.

When treating the model ARG blt with HOCl, NH₂Cl, ClO₂, O₃, or UV light, ARG deactivation kinetics correlated strongly with ARG degradation (as measured by qPCR) as long as the ARG segments analyzed by qPCR had lengths equivalent to minimum sequence lengths (~800-1000 bp) required for natural transformation of genes in B. subtilis. This indicates that in carefully selected cases, it may be feasible to use qPCR measurements (which are relatively cheap, rapid, and simple) as surrogates for direct measurements of ARG deactivation (which are much more expensive, slower, and labor intensive). However, for *OH and H₂O₂, ARG deactivation was much faster than anticipated based on qPCR measurements of ARG degradation, illustrating the need for caution in attempting to use qPCR measurements as surrogates for direct measurements of ARG deactivation in general.

Experiments with intracellular ARGs showed that for every disinfectant investigated, ARG degradation and deactivation lagged inactivation of ARB cells themselves (i.e., ARGs could indeed "survive" ARB inactivation). For HOCl, O₃, UV, and H₂O₂, ≥99% degradation and deactivation of ARGs was feasible at practical disinfectant exposure levels, whereas for NH₂Cl, ClO₂, EtOH, and phenol, little degradation or deactivation of ARGs was observed at even the most extreme treatment conditions likely to be applied in practice.

Bench- and full-scale chlorination and UV irradiation of the investigated ARGs and ARB strains in low- and high-ammonia wastewater effluents collected from two local wastewater treatment facilities - with and without nitrification/denitrification, respectively - showed that chlorination of high-ammonia wastewater effluents yielded minimal degradation of ARGs under practical treatment conditions - due to rapid conversion of HOCl to NH₂Cl. This shows that chlorination of non-nitrified effluent is unlikely to prevent ARG release into downstream waterbodies. In contrast, ≥90-99% ARG degradation could be achieved during chlorination of low-ammonia effluents (where added HOCl remains as HOCl), and ≥90-99% ARG degradation could be achieved during UV irradiation of high- or low-ammonia effluents, consistent with measured ARG reactivities toward each disinfectant.

Overall, this work has yielded extensive data on DNA reaction kinetics and treatment requirements for achieving effective ARG degradation with common (waste)water and healthcare disinfectants. The resulting knowledge can be used to improve selection and design of disinfection processes, and also illustrates that in some important treatment scenarios, certain disinfection processes may not prevent ARG dissemination within environmental or healthcare settings (e.g., chlorination of high-ammonia wastewater effluents).

Research findings from the project have been communicated to technical and non-technical audiences through peer-reviewed publications in leading journals, oral and poster presentations at regional, national, and international conferences, and outreach to the public through regular K-8 open house events hosted by the UW College of Engineering (Figure 2).

The project has also supported professional development of an early career investigator (the PI), five graduate students (three UW PhD students and two visiting PhD students), and four undergraduate students (three UW students and one visiting student). Additionally, the project has provided thirty-two high school teachers and sixty-one high school students (Figure 3) with hands-on training in water sciences and engineering through an annual workshop program focusing on inquiry-based curriculum development (see: http://faculty.washington.edu/doddm/Personal/WaterWorks.html).

Last Modified: 12/30/2019
Modified by: Michael Dodd