Metagenomic evaluation of bacteria in drinking water using full-length 16S rRNA amplicons.

Escherichia coli and total coliforms are important tools for identifying potential faecal contamination in drinking water. However, metagenomics offers a powerful approach for delving deeper into a bacterial community when E. coli or total coliforms are detected. Metagenomics can identify microbes native to water systems, track community changes and potential pathogens introduced by contamination events, and evaluate the effectiveness of treatment processes. Here, we demonstrate how the dual application of traditional monitoring practices and metagenomics can improve monitoring and surveillance for water resource management. The robustness of long-read metagenomics across replicates is demonstrated by the effect and interaction between manganese filters and bacterial communities, as well as the impact of chlorination after coliform detection. These examples reveal how metagenomics can identify the complex bacterial communities in the distribution system and the source waters used to supply drinking water treatment plants (DWTPs). The knowledge gained increases confidence in identified causes and mitigations of potential contamination events. By exploring bacterial communities, we can gain additional insights into the impact of faecal contamination events and treatment processes. This insight enables more precise remediation actions and enhances confidence in communicating health risks to drinking water operators and the public.

Development of a multiplex droplet digital PCR assay for simultaneous detection and quantification of Escherichia coli, E. marmotae, and E. ruysiae in water samples.

Escherichia coli are widely used by water quality managers as Fecal Indicator Bacteria, but current quantification methods do not differentiate them from benign, environmental Escherichia species such as E. marmotae (formerly named cryptic clade V) or E. ruysiae (cryptic clades III and IV). Reliable and specific techniques for their identification are required to avoid confounding microbial water quality assessments. To address this, a multiplex droplet digital PCR (ddPCR) assay targeting lipB (E. coli and E. ruysiae) and bglC (E. marmotae) was designed. The ddPCR performance was assessed using in silico analysis; genomic DNA from 40 local, international, and reference strains of target and non-target coliforms; and spiked water samples in a range relevant to water quality managers (1 to 1000 cells/100 mL). Results were compared to an analogous quantitative PCR (qPCR) and the Colilert method. Both PCR assays showed excellent sensitivity with a limit of detection of 0.05 pg/µL and 0.005 pg/µl for ddPCR and qPCR respectively, and of quantification of 0.5 pg/µL of genomic DNA. The ddPCR allowed differentiation and quantification of three Escherichia species per run by amplitude multiplexing and showed a high concordance with concentrations measured by Colilert once proportional bias was accounted for. In silico specificity testing underlined the possibility to further detect and distinguish Escherichia cryptic clade VI. Finally, the applicability of the ddPCR was successfully tested on environmental water samples where E. marmotae and E. ruysiae potentially confound E. coli counts based on the Most Probable Number method, highlighting the utility of this novel ddPCR as an efficient and rapid discriminatory test to improve water quality assessments.

Draft genome sequences of Escherichia spp. isolates from New Zealand environmental sources.

Escherichia coli is often used as a fecal indicator bacterium for water quality monitoring. We report the draft genome sequences of 500 Escherichia isolates including newly described Escherichia species, namely Escherichia marmotae, Escherichia ruysiae, and Escherichia whittamii, obtained from diverse environmental sources to assist with improved public health risk assessments.

High-resolution genomic analysis to investigate the impact of the invasive brushtail possum (Trichosurus vulpecula) and other wildlife on microbial water quality assessments.

Escherichia coli are routine indicators of fecal contamination in water quality assessments. Contrary to livestock and human activities, brushtail possums (Trichosurus vulpecula), common invasive marsupials in Aotearoa/New Zealand, have not been thoroughly studied as a source of fecal contamination in freshwater. To investigate their potential role, Escherichia spp. isolates (n = 420) were recovered from possum gut contents and feces and were compared to those from water, soil, sediment, and periphyton samples, and from birds and other introduced mammals collected within the Makirikiri Reserve, Dannevirke. Isolates were characterized using E. coli-specific real-time PCR targeting the uidA gene, Sanger sequencing of a partial gnd PCR product to generate a gnd sequence type (gST), and for 101 isolates, whole genome sequencing. Escherichia populations from 106 animal and environmental sample enrichments were analyzed using gnd metabarcoding. The alpha diversity of Escherichia gSTs was significantly lower in possums and animals compared with aquatic environmental samples, and some gSTs were shared between sample types, e.g., gST535 (in 85% of samples) and gST258 (71%). Forty percent of isolates gnd-typed and 75% of reads obtained by metabarcoding had gSTs shared between possums, other animals, and the environment. Core-genome single nucleotide polymorphism (SNP) analysis showed limited variation between several animal and environmental isolates (<10 SNPs). Our data show at an unprecedented scale that Escherichia clones are shared between possums, other wildlife, water, and the wider environment. These findings support the potential role of possums as contributors to fecal contamination in Aotearoa/New Zealand freshwater. Our study deepens the current knowledge of Escherichia populations in under-sampled wildlife. It presents a successful application of high-resolution genomic methods for fecal source tracking, thereby broadening the analytical toolbox available to water quality managers. Phylogenetic analysis of isolates and profiling of Escherichia populations provided useful information on the source(s) of fecal contamination and suggest that comprehensive invasive species management strategies may assist in restoring not only ecosystem health but also water health where microbial water quality is compromised.

Exploring the Bacterial Community in Aged Fecal Sources from Dairy Cows: Impacts on Fecal Source Tracking.

(1) Background: This paper discusses the impact of agricultural activities on stream health, particularly in relation to dairy cow fecal pollution. The study explores the fecal microbiome of cattle and the potential ecological implications of aging fecal pollution on waterways. (2) Methods: The study examines changes in the bacterial community available for mobilization from in-situ decomposing cowpats and the effects of simulated rainfall. The microbiome of individual cowpats was monitored over 5.5 months. We used 16S rRNA metagenomics and machine learning software, FEAST (Fast Expectation-mAximization for microbial Source Tracking), for bacterial and fecal source assignments. (3) Results: The phyla Bacillota and Bacteroidota are dominant in the fecal microbiota of fresh cow feces but shift to Pseudomonodota, Actinomycetota, and environmental Bacteroidota in aged cowpats. Potential impacts of these bacterial community shifts on inputs to local agricultural streams are discussed in relation to water quality monitoring and aging sources of fecal contamination. We identified taxon orders that are potential indicators of fresh cattle sources (Oscillospirales and Bacteroidales) and aged sources (Peptostreptococcales-Tissierellales) in water bodies. (4) The paper highlights that bacterial metagenomic profiling can inform our understanding of the ecology of microbial communities in aquatic environments and the potential impacts of agricultural activities on ecosystem health.

Mobilization of Escherichia coli and fecal source markers from decomposing cowpats.

In rural environments, the sources of fecal contamination in freshwater environments are often diffuse and a mix of fresh and aged fecal sources. It is important for water monitoring purposes, therefore, to understand the impacts of weathering on detection of the fecal source markers available for mobilization from livestock sources. This study targets the impacts of rainfall events on the mobilization of fecal source tracking (FST) markers from simulated cowpats decomposing in situ for five-and-a-half-months. The FST markers analysed were Escherichia coli, microbial source tracking (MST) markers, fecal steroids and a fecal ageing ratio based on the ratio between counts of river microflora and total coliforms. There was a substantial concentration of E. coli (104/100 mL) released from the ageing cowpats suggesting a long-term reservoir of E. coli in the cowpat. Mobilization of fecal markers from rainfall-impacted cowpats, however, was markedly reduced compared with fecal markers in the cowpat. Overall, the Bacteroidales bovine-associated MST markers were less persistent than E. coli in the cowpat and rainfall runoff. The ten fecal steroids, including the major herbivore steroid, 24-ethylcoprostanol, are shown to be stable markers of bovine pollution due to statistically similar degradation rates among all steroids. The mobilizable fraction for each FST marker in the rainfall runoff allowed generation of mobilization decline curves and the derived decline rate constants can be incorporated into source attribution models for agricultural contaminants. Findings from this study of aged bovine pollution sources will enable water managers to improve attribution of elevated E. coli to the appropriate fecal source in rural environments.

Whole-Genome Sequencing and Virulome Analysis of Escherichia coli Isolated from New Zealand Environments of Contrasting Observed Land Use.

Generic Escherichia coli is commonly used as an indicator of fecal contamination to assess water quality and human health risk. Where measured E. coli exceedances occur, the presence of other pathogenic microorganisms, such as Shiga toxin-producing E. coli (STEC), is assumed, but confirmatory data are lacking. Putative E. coli isolates (n = 709) were isolated from water, sediment, soil, periphyton, and feces samples (n = 189) from five sites representing native forest and agricultural environments. Ten E. coli isolates (1.41%) were stx2 positive, 19 (2.7%) were eae positive, and stx1-positive isolates were absent. At the sample level, stx2-positive E. coli (5 of 189, 2.6%) and eae-positive isolates (16 of 189, 8.5%) were rare. Using real-time PCR, these STEC-associated virulence factors were determined to be more prevalent in sample enrichments (stx1, 23.9%; stx2, 31.4%; eae, 53.7%) and positively correlated with generic E. coli isolate numbers (P < 0.05) determined using culture-based methods. Whole-genome sequencing (WGS) was undertaken on a subset of 238 isolates with assemblies representing seven E. coli phylogroups (A, B1, B2, C, D, E, and F), 22 Escherichia marmotae isolates, and 1 Escherichia ruysiae isolate. Virulence factors, including those from extraintestinal pathogenic E. coli, were extremely diverse in isolates from the different locations and were more common in phylogroup B2. Analysis of the virulome from WGS data permitted the identification of gene repertoires that may be involved in environmental fitness and broadly align with phylogroup. Although recovery of STEC isolates was low, our molecular data indicate that they are likely to be widely present in environmental samples containing diverse E. coli phylogroups. IMPORTANCE This study takes a systematic sampling approach to assess the public health risk of Escherichia coli recovered from freshwater sites within forest and farmland. The New Zealand landscape is dominated by livestock farming, and previous work has demonstrated that "recreational exposure to water" is a risk factor for human infection by Shiga toxin-producing Escherichia coli (STEC). Though STEC isolates were rarely isolated from water samples, STEC-associated virulence factors were identified more commonly from water sample culture enrichments and were associated with increased generic E. coli concentrations. Whole-genome sequencing data from both E. coli and newly described Escherichia spp. demonstrated the presence of virulence factors from E. coli pathotypes, including extraintestinal pathogenic E. coli. This has significance for understanding and interpreting the potential health risk from E. coli where water quality is poor and suggests a role of virulence factors in survival and persistence of E. coli and Escherichia spp.

Application of crAssphage, F-RNA phage and pepper mild mottle virus as indicators of human faecal and norovirus contamination in shellfish.

Shellfish growing waters contaminated with inadequately treated human wastewater is a major source of norovirus in shellfish and poses a significant human health risk to consumers. Microbial source tracking (MST) markers have been widely used to identify the source (s) of faecal contamination in water but data are limited on their use for shellfish safety. This study evaluated the source specificity, sensitivity, occurrence and concentration of three viral MST markers i.e. cross-assembly phage (crAssphage), F-specific RNA bacteriophage genogroup II (F-RNA phage GII) and pepper mild mottle virus (PMMoV) using animal faeces (n = 119; 16 animal groups), influent wastewater (n = 12), effluent wastewater (n = 16) and shellfish (n = 33). CrAssphage, F-RNA phage GII and PMMoV had source specific values of 0.97, 0.99 and 0.91, respectively. The sensitivity of MST markers was confirmed by their 100% detection frequency in influent wastewaters. The frequency of detection in effluent wastewater ranged from 81.3% (F-RNA phage GII) to 100% (PMMoV). Concentration of F-RNA phage GII was one log10 (influent wastewater) and 2-3 log10 (effluent wastewater) lower than crAssphage and PMMoV, respectively. Despite lower prevalence of F-RNA phage GII in oysters and mussels compared to crAssphage and PMMoV, concentrations of the three MST markers were similar in mussels. As an indicator of norovirus contamination in shellfish, crAssphage and PMMoV had greater predictive sensitivity (100%; [95% CI; 81.5%-100%)]) and F-RNA phage GII had greater predictive specificity (93.3%; [95% CI; 68.1%-99.8%]). In contrast, crAssphage and F-RNA phage GII have similar accuracy for predicting norovirus in shellfish, however, PMMoV significantly overestimated its presence. Therefore, a combination of crAssphage and F-RNA phage GII analysis of shellfish could provide a robust estimation of the presence of human faecal and norovirus contamination.

Fecal indicator bacteria from environmental sources; strategies for identification to improve water quality monitoring.

In tropical to temperate environments, fecal indicator bacteria (FIB), such as enterococci and Escherichia coli, can persist and potentially multiply, far removed from their natural reservoir of the animal gut. FIB isolated from environmental reservoirs such as stream sediments, beach sand and vegetation have been termed "naturalized" FIB. In addition, recent research suggests that the intestines of poikilothermic animals such as fish may be colonized by enterococci and E. coli, and therefore, these animals may contribute to FIB concentrations in the aquatic environment. Naturalized FIB that are derived from fecal inputs into the environment, and subsequently adapted to maintain their population within the non-host environment are termed "naturalized enteric FIB". In contrast, an additional theory suggests that some "naturalized" FIB diverged from enteric FIB many millions of years ago and are now normal inhabitants of the environment where they are referred to as "naturalized non-enteric FIB". In the case of the Escherichia genus, the naturalized non-enteric members are identified as E. coli during routine water quality monitoring. An over-estimation of the health risk could result when these naturalized, non-enteric FIB, (that is, not derived from avian or mammalian fecal contamination), contribute to water quality monitoring results. It has been postulated that these environmental FIB belonging to the genera Escherichia and Enterococcus can be differentiated from enteric FIB by genetic methods because they lack some of the genes required for colonization of the host intestine, and have acquired genes that aid survival in the environment. Advances in molecular tools such as next generation sequencing will aid the identification of genes peculiar or "enriched" in particular habitats to discriminate between enteric and environmental FIB. In this appraisal, we have reviewed the research studying "naturalized" FIB, and discussed the techniques for their differentiation from enteric FIB. This differentiation includes the important distinction between enteric FIB derived from fresh and non-recent fecal inputs, and those truly non-enteric environmental microbes, which are currently identified as FIB during routine water quality monitoring. The inclusion of tools for the identification of naturalized FIB (enteric or environmental) would be a valuable resource for future studies assessing water quality.

Evaluation of pepper mild mottle virus as an indicator of human faecal pollution in shellfish and growing waters.

Bivalve molluscan shellfish grown in areas impacted by human faecal pollution are at risk of being contaminated with multiple enteric viruses. To minimise the public health risks associated with shellfish consumption, determining the presence of faecal contamination in shellfish and their growing waters is crucial. In this study, we evaluated the use of pepper mild mottle virus (PMMoV) as an indicator of human faecal contamination in oysters, mussels, cockles and shellfish growing waters in New Zealand. Using reverse transcription quantitative polymerase chain reaction (RT-qPCR) the presence, and where applicable, the concentration of PMMoV was determined in faeces from 11 different animal species, influent (untreated) wastewater, shellfish and shellfish growing waters. Non-human faecal samples (from seagull, Canada goose, black swan and dog) were RT-qPCR positive for PMMoV. The faecal source specificity of PMMoV was 0.83 (maximum value of 1) when 'detected but not quantifiable' (DNQ) values were used. However, when 'lower limit of quantification' (LLOQ) values were used, the specificity increased to 0.92. The PMMoV concentration in influent wastewater (n = 10) ranged from 6.3 to 7.7 log10 genome copies (GC)/L with a mean (±standard deviation) of 7.1 ±â€¯0.5 log10 GC/L. The overall occurrence of PMMoV in shellfish and shellfish growing waters from four different areas was 46/51 (90%) and 29/52 (56%), respectively. Of the cockles collected from an area known to be impacted by effluent wastewater, 14/14 (100%) contained PMMoV concentrations above the LLOQ. In contrast, only 13/37 (35%) shellfish and 6/52 (11.5%) growing water samples collected from three areas with low anthropogenic impact contained PMMoV concentrations above the LLOQ. The high concentration of PMMoV in influent wastewater indicates that PMMoV may be a promising indicator of human faecal contamination. The presence of PMMoV in shellfish and growing waters with a low anthropogenic impact may be of avian origin, and this needs to be considered if using PMMoV for monitoring shellfish and shellfish growing water quality in New Zealand.

Relationships between chemical and microbial faecal source tracking markers in urban river water and sediments during and post-discharge of human sewage.

This study explores the relationships between faecal source tracking (FST) markers (quantitative Polymerase Chain Reaction (qPCR) markers and steroids), microbial indicators, the faecal ageing ratio of atypical colonies/total coliforms (AC/TC) and potential human pathogens (Giardia, Cryptosporidium and Campylobacter). Faecal source PCR markers tested were GenBac3, HumM3, HumBac (HF183-Bac708R); Bifidobacterium adolescentis, wildfowl and canine-associated markers. Sediment and water samples from the Avon River were collected during and post-discharge of untreated human sewage inputs, following a series of earthquakes, which severely damaged the Christchurch sewerage system. Significant, positive Spearman Ranks (rs) correlations were observed between human-associated qPCR markers and steroid FST markers and Escherichia coli and F-specific RNA bacteriophage (rs 0.57 to 0.84, p < 0.001) in water samples. These human source indicative FST markers demonstrated that they were also effective predictors of potentially pathogenic protozoa in water (rs 0.43-0.74, p ≤ 0.002), but correlated less well with Campylobacter. Human-associated qPCR and steroid markers showed significant, substantial agreement between the two FST methods (Cohen's kappa, 0.78, p = 0.023), suggesting that water managers could be confident in the results using either method under these contamination conditions. Low levels of fluorescent whitening agents (FWA) (mean 0.06 µg/L, range 0.01-0.40 µg/L) were observed in water throughout the study, but steroids and FWA appeared to be retained in river sediments, months after continuous sewage discharges had ceased. No relationship was observed between chemical FST markers in sediments and the overlying water, and few correlations in sediment between chemical FST markers and target microorganisms. The low values observed for the faecal ageing ratio, AC/TC in water, were significantly, negatively correlated with increasing pathogen detection. This study provides support for the use of the AC/TC ratio, and qPCR and steroid FST markers as indicators of health risks associated with the discharge of raw human sewage into a freshwater system.

Fecal source tracking methods to elucidate critical sources of pathogens and contaminant microbial transport through New Zealand agricultural watersheds - A review.

In New Zealand, there is substantial potential for microbial contaminants from agricultural fecal sources to be transported into waterways. The flow and transport pathways for fecal contaminants vary at a range of scales and is dependent on chemical, physical and biological attributes of pathways, soils, microorganisms and landscape characteristics. Understanding contaminant transport pathways from catchment to stream can aid water management strategies. It is not practical, however to conduct direct field measurement for all catchments on the fate and transport of fecal pathogens due to constraints on time, personnel, and material resources. To overcome this problem, fecal source tracking can be utilised to link catchment characteristics to fecal signatures identifying critical sources. In this article, we have reviewed approaches to identifying critical sources and pathways for fecal microorganisms from agricultural sources, and make recommendations for the appropriate use of these fecal source tracking (FST) tools.

Identifying avian sources of faecal contamination using sterol analysis.

Discrimination of the source of faecal pollution in water bodies is an important step in the assessment and mitigation of public health risk. One tool for faecal source tracking is the analysis of faecal sterols which are present in faeces of animals in a range of distinctive ratios. Published ratios are able to discriminate between human and herbivore mammal faecal inputs but are of less value for identifying pollution from wildfowl, which can be a common cause of elevated bacterial indicators in rivers and streams. In this study, the sterol profiles of 50 avian-derived faecal specimens (seagulls, ducks and chickens) were examined alongside those of 57 ruminant faeces and previously published sterol profiles of human wastewater, chicken effluent and animal meatwork effluent. Two novel sterol ratios were identified as specific to avian faecal scats, which, when incorporated into a decision tree with human and herbivore mammal indicative ratios, were able to identify sterols from avian-polluted waterways. For samples where the sterol profile was not consistent with herbivore mammal or human pollution, avian pollution is indicated when the ratio of 24-ethylcholestanol/(24-ethylcholestanol + 24-ethylcoprostanol + 24-ethylepicoprostanol) is ≥0.4 (avian ratio 1) and the ratio of cholestanol/(cholestanol + coprostanol + epicoprostanol) is ≥0.5 (avian ratio 2). When avian pollution is indicated, further confirmation by targeted PCR specific markers can be employed if greater confidence in the pollution source is required. A 66% concordance between sterol ratios and current avian PCR markers was achieved when 56 water samples from polluted waterways were analysed.

The impact of major earthquakes and subsequent sewage discharges on the microbial quality of water and sediments in an urban river.

A series of large earthquakes struck the city of Christchurch, New Zealand in 2010-2011. Major damage sustained by the sewerage infrastructure required direct discharge of up to 38,000 m(3)/day of raw sewage into the Avon River of Christchurch for approximately six months. This allowed evaluation of the relationship between concentrations of indicator microorganisms (Escherichia coli, Clostridium perfringens and F-RNA phage) and pathogens (Campylobacter, Giardia and Cryptosporidium) in recreational water and sediment both during and post-cessation of sewage discharges. Giardia was the pathogen found most frequently in river water and sediment, although Campylobacter was found at higher levels in water samples. E. coli levels in water above 550 CFU/100 mL were associated with increased likelihood of detection of Campylobacter, Giardia and Cryptosporidium, supporting the use of E. coli as a reliable indicator for public health risk. The strength of the correlation of microbial indicators with pathogen detection in water decreased in the following order: E. coli>F-RNA phage>C. perfringens. All the microorganisms assayed in this study could be recovered from sediments. C. perfringens was observed to accumulate in sediments, which may have confounded its usefulness as an indicator of fresh sewage discharge. F-RNA phage, however, did not appear to accumulate in sediment and in conjunction with E. coli, may have potential as an indicator of recent human sewage discharge in freshwater. There is evidence to support the low-level persistence of Cryptosporidium and Giardia, but not Campylobacter, in river sediments after cessation of sewage discharges. In the event of disturbances of the sediment, it is highly probable that there could be re-mobilisation of microorganisms beyond the sediment-water exchange processes occurring under base flow conditions. Re-suspension events do, therefore, increase the potential risk to human health for those who participate in recreational and work-related activities in the river environment.

Sunlight inactivation of human polymerase chain reaction markers and cultured fecal indicators in river and saline waters.

Decay rates for sunlight inactivation of polymerase chain reaction (PCR) markers for total Bacteroidales, human-specific Bacteroidales, Escherichia coli, and Bifidobacterium adolescentis relative to cultured E. coli were investigated. The experiment used 100-L chambers of fresh water and seawater (paired with dark controls) seeded with human sewage and exposed to natural sunlight over three summer days. Culturable E. coli levels in sunlight-exposed chambers decreased by at least 3 logs on day 1, and by day 3 a total reduction of 4.5 to 5.5 logs was achieved in fresh water and seawater, respectively. In contrast, PCR detection of the four gene targets in sunlight-exposed chambers reduced by no more than 2 logs over the duration of the study (k(t) < 0.071 log(e) units h(-1)). Under these experimental conditions, PCR markers are considerably more conservative than culturable E. coli and can persist for extended periods of time following inactivation of E. coli.

A PCR marker for detection in surface waters of faecal pollution derived from ducks.

Detection of the faecal pollution contribution from wildfowl is an important adjunct in determining the sources of faecal pollution in waterways. This is particularly true, where human waste and other animal faecal sources have been eliminated as the pollution source. A polymerase chain reaction (PCR) marker was developed as a duck-specific marker of faecal pollution. The semi-nested primer system targeted an unknown bacterium (E2) isolated from mallard ducks. E2 had the closest 16S rRNA sequence similarity to members of the Desulfovibrio genus, which was further confirmed by phenotypic characterisation of the bacterium. Testing of the prevalence of E2 identified the marker in 76% of duck faecal samples (n=42), 20% of swan faecal samples (n=10) and 15% of Canada geese faecal samples (n=20). It was also identified in the faeces of two out of 15 domestic goats (13%). The marker was not detected in any samples derived from human faeces or effluent, dairy cows or sheep. It is proposed that this PCR marker would be useful in conjunction with faecal taxation indicators in the determination of pollution derived from duck faecal inputs in waterways.