The opportunistic pathogen Legionella is the leading cause of reported waterborne disease outbreaks in the United States. Legionella can thrive under the warm, stagnant, low-disinfectant conditions characteristic of premise (i.e., building) plumbing systems, making it challenging to identify effective interventions for its control. Copper (Cu) is a promising antimicrobial that can be dosed directly to water via copper-silver ionization systems or released naturally via corrosion of Cu pipes to help control growth of Legionella and other pathogens. However, prior research has shown that Cu does not always reliably control Legionella and sometimes seems to even stimulate its growth. A deeper understanding of the mechanistic effects of Cu on Legionella, at both pure-culture and real-world scales, is critical in order to inform effective controls for Legionella. The overarching objective of the research embodied by this dissertation was aimed at elucidating the chemical and microbial interactions in premise plumbing that govern efficacy of Cu for Legionella control through a series of complementary bench-, pilot-, and field-scale studies.
A critical review and synthesis of the literature identified important knowledge gaps in relation to antimicrobial effects of Cu. In particular, changes in the pH, phosphate corrosion control, and rising levels of natural organic matter (NOM) in distributed water are predicted to be important controlling factors. The type of sacrificial anode rod material employed in water heaters was also identified as an underappreciated factor, which directly affects pH, evolution of hydrogen gas as a microbial nutrient, and release of metals (such as aluminum) that bind copper. Microbiological factors: including growth phase of Legionella (e.g., exponential or stationary), strain-specific Cu tolerance, background microbiome composition, and the possibility that viable but non-culturable (VBNC) Legionella might still cause human disease, were also identified as major confounding factors. These knowledge gaps are addressed from various dimensions across each chapter of the dissertation.
The effects of pH, orthophosphate corrosion inhibitor concentration, and NOM were examined in bench-scale pure culture experiments over a range of conditions relevant to drinking water. Cupric ions and antimicrobial effects were drastically reduced at pH >7.5, especially in the presence of phosphate, which precipitates copper, or NOM, which complexes the Cu in a form that is less bioavailable. Chick-Watson disinfection models indicated that soluble Cu was the most robust correlate with observed Cu antimicrobial effects across a range of tested waters. This new knowledge suggests that measuring soluble rather than total Cu would be much more informative to guide practitioners in dosing. The research also demonstrated that changes in pH or orthophosphate that have been made to control corrosion over the last few decades, have significantly altered Cu chemistry in buildings, undermining antimicrobial capacity and increasing likelihood of Legionella growth.
Pilot-scale experiments confirmed that soluble Cu is an effective indicator of Cu antimicrobial capacity, even in more complex environments represented by realistic hot water plumbing systems. In particular, dosing of orthophosphate, which is widely added by drinking water utilities to control corrosion, directly reduces soluble copper and overall antimicrobial capacity. In some cases, Cu added together with orthophosphate apparently promoted the growth of Legionella, providing an example of at least one circumstance where Cu addition can induce interactive effects that elevate Legionella compared to a control system with trace Cu.
It was also demonstrated for the first time that different water heater sacrificial anode types are subject to different corrosion processes, which indirectly influence Cu antimicrobial capacity. Specifically, aluminum ions released from aluminum anode corrosion at 1 mg/L can form an Al(OH)3 gel, which can remove >80% of the soluble Cu from water and reduce Cu antimicrobial effects by >2-log at pH=7. Corrosion from magnesium anodes was found to dramatically increase the pH from 6.8 to >8, which correspondingly reduces Cu antimicrobial capacity. Cu deposition further increased the anode corrosion rate and promoted evolution of hydrogen gas, which is a potent electron donor that stimulates autotrophic microbial growth especially with a magnesium anode. Electric powered anodes did not release metals or alter pH and thus did not diminish Cu antimicrobial capacity. Still, across the pilot-scale experiments, even very high levels of Cu (>1.2 mg/L) at low pH (<7) failed to fully eradicate culturable Legionella.
The much lower than expected antimicrobial efficacy of Cu in the pilot-scale hot water plumbing systems was found to be partially explained by the properties of the strain that colonized the systems. Based on fitting the data to a Chick-Watson disinfection model, the outbreak-associated strain that was inoculated into the systems was estimated to be 7 times more tolerant to Cu compared to the common lab strain applied in the bench-scale tests. Further, exponential growth phase L. pneumophila were found to be 2.5 times more susceptible to Cu relative to early stationary phase cultures. It is important to also recognize that, in the pilot-scale systems, drinking water biofilms and the amoeba hosts that colonize them can further shield Legionella from the antimicrobial effects of Cu.
Application of shotgun metagenomic sequencing offered the opportunity to more deeply examine the response of Legionella and other pathogens to Cu dosed to the pilot-scale hot water systems in the context of the broader microbiome. It was found that metagenomic analysis provided a sensitive indication of the bioavailability of Cu to the broader microbial community inhabiting the hot water systems, further confirming that the outbreak-associated strain of Legionella that colonized the rigs was relatively tolerant of Cu. Functional gene analysis provided further insight into the mechanistic action of Cu, suggesting multi-modal action of both membrane damage and interruption of nucleic acid replication. The metagenomic analysis further revealed that protozoan host numbers tended to increase in the pilot-scale systems with time, and this could also increase the potential for Legionella proliferation with time.
Additional pure culture studies aiming to further assess the mechanistic action of Cu provided strong evidence that Cu can induce a VBNC state for Legionella. This is a concern, given that other studies have indicated that VBNC Legionella are still capable of causing legionellosis. However, VBNC cells are not detected by conventional culturing. Multiple lines of evidence supported the conclusion that Cu induced a VBNC state for Legionella, including membrane integrity, enzyme activity, ATP generation, and Amoebae resuscitation assays applied to two different strains of L. pneumophila. After exposure to Cu, up to a 5-log (99.999%) reduction in culturable Legionella was observed, whereas corresponding reductions in the various viability measures were only by <1-log (90%). In other words, conventional culturing may miss up to 99.99% of the Legionella that is still capable of causing disease. To our knowledge, this is the first study that has assessed the potential for Cu-induced VBNC Legionella. Additional research is needed to further quantify the contribution of VBNC status to challenges in effective Cu-based control of Legionella in premise plumbing.
This research further examines, for the first time, the proteomic response of Legionella to Cu, comparing both presumably VBNC and culturable cells. Functional annotation of proteins that were differentially produced by the cells in response to Cu addition revealed that VBNC L. pneumophila modulated its proteome to favor cell membrane- and motility-related proteins, while reducing production of other proteins related to primary metabolism compared to culturable cells. These observations are consistent with the metagenomic-based observations and support the hypothesis that Cu inactivates cells by damaging the cell membrane. The findings also confirmed reduced general cell metabolism that is characteristic of a VBNC state.
This dissertation highlights the important and complex effects of Cu on Legionella growth in potable water systems as modified by water chemistry, water heater anode type, characteristics of the surrounding microbiome, and Legionella strain characteristics and growth status. The findings raise important questions about how to measure disinfectant efficacy and provide fundamental new knowledge that can help to better optimize the application of Cu as an antimicrobial to drinking water systems and better protect public health. / Doctor of Philosophy / The opportunistic pathogen Legionella is the leading cause of reportable waterborne disease outbreaks in the United States. Legionella is capable of growing in drinking water plumbing systems in homes, hospitals, hotels, and other buildings. Legionella is spread by inhaling tiny droplets of water that are suspended in the air when using the water, for example when showering, resulting in a severe and deadly form of pneumonia called Legionnaires' Disease. Copper is a promising antimicrobial that can be dosed directly into a building's water system by installing a copper-silver ionization system. There is also interest in understanding whether copper released naturally from copper pipes could help control Legionella. However, prior research indicates that copper sometimes inhibits, sometimes has no effect, and sometimes even seems to stimulate Legionella growth. The purpose of this dissertation was to better understand how the chemical properties of the drinking water, such as pH, presence of corrosion inhibitors that are commonly added to the water by utilities, and natural organic matter impact the ability of copper to kill Legionella. Impacts of the design of the drinking water system were also examined, for example, the material used in the anodes of water heaters to prevent corrosive damage to other system components was hypothesized to change the water chemistry in such a way that could also interfere with copper disinfection. Finally, the effect of the strain of Legionella, its growth phase (exponential or stationary), and culturability status (culturable versus viable but non-culturable) was also examined. Experiments were conducted over a wide range of conditions, from bench-scale pure culture experiments of a few days to full-scale plumbing systems over a period of 3.5 years. The complementary approaches maximize the strength of scientific conclusions about approaches that can more effectively control Legionella.
Several discoveries were made as a result of this research that can help to improve the use of copper for controlling Legionella in drinking water systems. In particular, it was found that it is best to keep the pH less than 7.5, because above pH 7.5 copper reacts with orthophosphate corrosion inhibitor or natural organic matter in the water in a manner that makes it less potent to microbes. Through disinfection modeling it was found that soluble copper was the best predictor of the ability to kill Legionella. Therefore, it is recommended from this research that practitioners should monitor soluble copper instead of total copper for the purpose of assessing Legionella control.
From the pilot-scale experiments, it was further found that the type of anode installed in the water heater can affect the ability of copper to kill Legionella. Magnesium anodes performed the worst, likely because they raised the pH above the recommended level of 7.5. Aluminum anodes were also a problem because aluminum ions released form an aluminum hydroxide gel that can remove more than 80% of the soluble copper from water. Electric powered anodes did not reduce copper antimicrobial effects by raising pH or forming a gel, but they are much less commonly used.
A surprising finding throughout this study was that very high levels of copper (>1.2 mg/L) were required to measurably reduce Legionella in the pilot-scale systems. In the pure culture experiments, it was found that the outbreak-associated strain from Quincy, IL, that was inoculated into the system was highly copper tolerant. This demonstrated that the strain of Legionella that colonizes a particular drinking water system could be the reason why copper is sometimes less effective. Pure culture experiments also found that stationary phase Legionella are more difficult to kill than exponential phase Legionella, which could explain some discrepancies among lab studies reported in the literature. A particularly noteworthy discovery of this research was that copper can make it appear as if Legionella have been killed, because the traditional culture media indicate that there is no growth on the Petri dish; however, they are in fact still alive and capable of causing human disease. This is referred to as a "viable but non-culturable (VBNC)" state. The VBNC state of Legionella was confirmed using an array of techniques (membrane integrity, enzyme activity, ATP generation, and amoebae resuscitation) for two strains of L. pneumophila. We also examined how VBNC Legionella cellular functions were impacted by copper using whole cell proteome, i.e., analysis of all of the proteins extracted from Legionella. Copper induced VBNC Legionella modulated its proteome to favor cell membrane and motility related proteins, and reduced others related to primary metabolism compared with culturable cells. These results were consistent with those obtained via shotgun metagenomic analysis of the microbial community DNA in the pilot-scale water systems. Given the potential for VBNC organisms to prevail in systems disinfected with copper, it is recommended to supplement culture-based monitoring with molecular-based monitoring, e.g., with quantitative polymerase chain reaction.
This dissertation highlights the important and complex effects of copper on Legionella growth in potable water systems. The findings help to inform guidance on how to improve the antimicrobial effect of copper, through adjusting the water chemistry, selecting appropriate water heater anodes, and optimizing the overall hot water system design. The dissertation also helps to inform improved strategies for monitoring the efficacy of copper for killing Legionella in real-world systems. Overall, the findings can help to improve policy and practice aimed at reducing the incidence of Legionnaires' Disease and protecting public health.
Identifer | oai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/107994 |
Date | 28 January 2022 |
Creators | Song, Yang |
Contributors | Civil and Environmental Engineering, Edwards, Marc A., Pruden, Amy, Rhoads, William J., Sarver, Emily A. |
Publisher | Virginia Tech |
Source Sets | Virginia Tech Theses and Dissertation |
Language | English |
Detected Language | English |
Type | Dissertation |
Format | ETD, application/pdf, application/pdf |
Rights | In Copyright, http://rightsstatements.org/vocab/InC/1.0/ |
Page generated in 0.0027 seconds