Article pubs.acs.org/est
Assessment of Genetic Markers for Tracking the Sources of Human Wastewater Associated Escherichia coli in Environmental Waters Ahmed Warish,*,†,‡ Cheryl Triplett,§ Ryota Gomi,∥ Pradip Gyawali,†,⊥ Leonie Hodgers,† and Simon Toze†,⊥ †
CSIRO Land and Water, Ecosciences Precinct, 41 Boggo Road, Brisbane, Queensland 4102, Australia Faculty of Science, Health and Education, University of the Sunshine Coast, Maroochydore, DC, Queensland 4558, Australia § Environmental Science, Spelman College, Atlanta, Georgia 30314, United States ∥ Department of Environmental Engineering, Graduate School of Engineering, Kyoto University, Katsura, Nishikyo-ku, 615-8540, Kyoto, Japan ⊥ School of Public Health, University of Queensland, Herston Road, Herston, Queensland 4006, Australia ‡
ABSTRACT: In this study, we have evaluated the performance characteristics (host-specificity and -sensitivity) of four human wastewater-associated Escherichia coli (E. coli) genetic markers (H8, H12, H14, and H24) in 10 target (human) and nontarget (cat, cattle, deer, dog, emu, goat, horse, kangaroo, and possum) host groups in Southeast Queensland, Australia. The overall hostsensitivity values of the tested markers in human wastewater samples were 1.0 (all human wastewater samples contained the E. coli genetic markers). The overall host-specificity values of these markers to differentiate between human and animal host groups were 0.94, 0.85, 0.72, and 0.57 for H8, H12, H24, and H14, respectively. Based on the higher host-specificity values, H8 and H12 markers were chosen for a validation environmental study. The prevalence of the H8 and H12 markers was determined among human wastewater E. coli isolates collected from a wastewater treatment plant (WWTP). Among the 97 isolates tested, 44 (45%) and 14 (14%) were positive for the H8 and H12 markers, respectively. A total of 307 E. coli isolates were tested from environmental water samples collected in Brisbane, of which 7% and 20% were also positive for the H8 and H12 markers, respectively. Based on our results, we recommend that these markers could be useful when it is important to identify the source(s) of E. coli (whether they originated from human wastewater or not) in environmental waters.
■
INTRODUCTION Fecal indicator bacteria (FIB) such as Escherichia coli and Enterococcus spp. have long been used as an indirect measure of public health risks associated with environmental waters.1,2 However, the value of FIB in translating the health risks has been questioned because of their poor correlation with pathogens, especially enteric viruses and protozoa.3−5 Some strains of FIB have the ability to adapt in the environment and persist in sediment and vegetation.6,7 The major limitation of FIB is that they cannot be assigned to their source of origin due to their cosmopolitan nature (found in all warm- and some cold-blooded animals).8,9 When environmental waters are polluted with elevated levels of FIB from diffuse sources, it becomes extremely difficult to implement a robust management plan without identifying their source(s). Over the last two decades, researchers have developed library-dependent and -independent microbial source tracking (MST) methods to identify the sources of fecal pollution in environmental waters. The earlier developed MST methods were library-dependent that required the collection and fingerprint matching of E. coli and Enterococcus spp. from variety of host groups and environmental waters.10,11 However, © XXXX American Chemical Society
the limitations of these methods based on FIB have been highlighted in the literature.8,12−14 For example, librarydependent methods can be costly and time-consuming to develop. In addition, the performance of a library can be affected by several factors such as geographical stability, temporal stability and complexity in statistical analysis.12,14 In contrast, library-independent methods primarily involve identifying a specific DNA sequence or a target gene of a bacterial species (predominantly 16S rRNA gene of Bacteroidales) found in host groups. Numerous polymerase chain reaction (PCR) based methods have been developed to identify, and in some cases quantify, the host-specific markers in environmental waters.9 The validation of these markers is assessed by several performance characteristics such as specificity, sensitivity, evenness, persistence, and relevance to health risks.9,11 However, none of the markers possess all of the desirable characteristics and Received: April 29, 2015 Revised: July 2, 2015 Accepted: July 7, 2015
A
DOI: 10.1021/acs.est.5b02163 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Table 1. Target and Nontarget Host Groups, Sources of Samples and the Concentration of DNA host groupsa Nontarget Host Groups cat cattle wastewater deer dog emu goat horse kangaroo possum Target Host Groups human wastewater a
sample type
sources of samples
weight or volume used for DNA extraction
range (ng) DNA per μL of extract
individual composite individual individual individual individual individual individual individual
veterinary hospital abattoir sanctuary veterinary hospital and parks emu park veterinary hospital horse racecourse sanctuary wild
180−220 mg 10 mL 180−220 mg 180−220 mg 180−220 mg 180−220 mg 180−220 mg 180−220 mg 180−220 mg
17.7−108 8.70−33.5 26.9−90.8 27.2−167 8.10−101 6.60−13.8 13.8−64.3 24.5−130 20.2−148
composite
wastewater treatment plants
10 mL
10.6−57.3
From each target and nontarget host group 10 samples were tested.
■
because of this, it has been recommended that a “toolbox” approach (analysis of multiple markers) should be used for the accurate identification of polluting sources.15−17 For MST studies, little focus has been given to the development of host specific E. coli and enterococci markers, even though for decades, these indicators have been routinely used for water quality monitoring. To the best of our knowledge, only cattle and pig specific E. coli toxin gene markers18,19 and the human specific enterococci surface protein (esp) marker found in Enterococcus faecium20 have been used for MST studies. One major limitation of the toxin gene markers is their low prevalence in target host group samples. Therefore, a cultural enrichment step is required prior to PCR to promote growth of injured and stressed cells. This technique is often used to detect pathogenic bacteria that generally occur at low concentrations in environmental waters.21 In recent years, terminal restriction fragment polymorphism (TRFLP), microarray, and next generation sequencing have allowed for a better characterization of bacterial communities from wastewater and environmental water samples.22−24 In addition, some of these methods allow us to identify specific marker(s) associated with specific microorganisms. A recent study has reported the development of a number of human wastewater, cattle, pig, and chicken feces associated E. coli genetic markers by sequencing the whole genome of E. coli isolates using Illumina sequencing.25 For the identification of host-specific genomic regions, the authors focused on E. coli accessory genes, which do not comprise the core genome. Performance characteristics data indicated that these markers may be a valuable addition to the existing library-independent MST “toolbox”. This is particularly important because E. coli has been used as an indicator of fecal pollution for environmental water bodies for decades. Therefore, knowing their source(s) is important in order to take necessary steps for remediation. The primary aim of this study was to evaluate the hostspecificity and -sensitivity of human wastewater-associated E. coli genetic markers (H8, H12, H14, and H24) by testing fecal samples from ten target and nontarget host groups in Southeast Queensland, Australia. Based on the performance characteristics, the best marker(s) were chosen for a validation environmental study. Real-time PCR detection of these markers in a collection of environmental E. coli isolates was then undertaken to support the presence of human wastewaterassociated E. coli in environmental waters.
MATERIALS AND METHODS
Target and Nontarget Host Group Sampling. To determine the host-specificity and -sensitivity of the E. coli genetic markers, fecal and wastewater samples were collected from 10 target (human wastewater) and nontarget host groups (nonhuman) (Table 1). Individual animal fecal samples were collected from the veterinary hospital (University of Queensland, Gatton), animal sanctuaries, parks, farms and racecourses. Beef cattle composite wastewater (mixture of feces and urine) samples were collected from two different abattoirs located on the outskirts of Brisbane. Human wastewater samples were collected from the primary influent of two wastewater treatment plants (WWTPs) serving 100 000−500 000 people in Brisbane. A fresh fecal sample (approximately 300 mg to 1 g) was collected from the defecation of each individual animal. All samples were transported on ice to the laboratory, stored at 4 °C, and processed within 12−24 h. Concentration of Cattle and Human Wastewater Samples. The human and cattle wastewater samples were concentrated with Amicon Ultra-15 (30 K) Centrifugal Filter Devices (Merck Millipore Ltd.). In brief, 10 mL of wastewater sample was added to the Amicon device and centrifuged at 4750g for 10 min. 180−200 μL (entire volume) of concentrated samples were collected from the filter device sample reservoir using a pipet.26 The concentrated samples were stored at −20 °C for a maximum of 24 h prior to DNA extraction. DNA Extraction and Standardization. DNA was extracted from the concentrated wastewater samples using DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA). A QIAamp Stool DNA Kit (Qiagen) was used to extract DNA from 100 to 220 mg of fresh animal fecal sample. All DNA samples were quantified using a NanoDrop spectrophotometer (ND-8000; NanoDrop Technology, Wilmington, DE). Since DNA samples from different host groups had different DNA concentrations (ranging from 6.6 to 167 ng/μL of DNA), all DNA samples were standardized to a concentration of 5 ng/μL of DNA (Table 1). PCR Inhibition. To obtain information on the level of PCR inhibition, standardized DNA samples extracted from target and nontarget host groups were spiked with 10 pg of Oncorhynchus keta DNA (Sigma Chemical Co., St. Louis, Mo.) and tested with the Sketa22 real-time PCR assay.27,28 The CT values for 10 pg O. keta DNA were also determined in realtime PCR reactions with DNase and RNase free water. O. keta CT values in DNase and RNase free water were compared with B
DOI: 10.1021/acs.est.5b02163 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
from 65 to 98 °C at 0.5 °C increment. The correct real-time PCR products were verified by comparing the melting curves of the real-time PCR positive controls (obtained from Department of Environmental Engineering, Kyoto University, Japan) which showed peaks at melting temperature 93 ± 0.5 °C, 87 ± 0.5 °C, 85 ± 0.5 °C, 82 ± 0.5 °C for H8, H12, H14, and H24, respectively. All real-time PCR assays were performed using the Bio-Rad CFX96 thermal cycler (Bio-Rad Laboratories). Data Analysis. The host-specificity and -sensitivity of the markers were determined as follows: sensitivity = a/(a + b) and specificity = c/(c + d), where a is true positive (samples were positive for target host group), b is false negative (samples were negative for target host group), c is true negative (samples were negative for nontarget host groups), and d is false positive (samples were positive for nontarget host groups). Chi-square test was used to test the differences in the prevalence of the selected markers in E. coli isolates collected from human wastewater and environmental water samples using the statistical package SPSS 21.0 (IBM Inc., Chicago, IL).
those spiked O. keta in wastewater and fecal samples from target and nontarget host groups to obtain information on the PCR inhibition level. Isolation of E. coli from Human Wastewater Samples. Three human wastewater samples (100 mL) were collected from the primary influent of a WWTP on three separate occasions. A total number of 112 E. coli were isolated from the wastewater samples using a spread plate method. In brief, serial dilutions of wastewater samples were made, and streaked on Chromocult coliform agar (Merck, Germany) followed by incubation at 37 °C overnight. Individual, well-isolated typical E. coli colonies were picked from the agar plates. After purification, single colonies from agar plates were inoculated into 2 mL centrifuge tubes containing 1 mL nutrient broth (Oxoid, UK). Inoculated tubes were incubated overnight at 37 °C in the shaking platform incubator at 100 rpm. The overnight culture from each isolate was centrifuged at 10 000g for 5 min. The supernatant was decanted and the cell pellet was resuspended in 200 μL sterile water by vortexing. DNA was extracted from the pellet using the InstaGene matrix according to the manufacturer’s instructions (Bio-Rad Laboratories, Richmond, CA). Sources of Environmental E. coli Isolates. A total of 307 environmental E. coli isolates were obtained from 21 water samples collected from six urban catchments in Brisbane, Australia in a previous study.29 The six sites (Brisbane River, Cabbage Tree Creek, Enoggera Creek, Fitzgibbon Drain, Oxley Creek and Pine River) represent diverse sources of human and animal fecal pollution through an urban environment. The procedures for E. coli isolation, DNA extraction and E. coli confirmatory test have been described elsewhere.29 Real-Time PCR Assays. O. keta DNA was amplified using previously published primers, probe and cycling parameters.28 Amplification for O. keta was performed in 25 μL reaction mixtures using iQ Supermixes (Bio-Rad Laboratories). The real-time PCR assay mixtures contained 12.5 μL of Supermixes, 300 nM of each primer, 400 nM of probe, 10 ng (2 μL) of template DNA sample and 10 pg of O. keta DNA. To confirm the presence of E. coli in each standardized fecal DNA sample from the target and nontarget host groups, 23S rRNA E. coli gene was amplified using previously published primers, probe and cycling parameters.30 DNA samples extracted from presumptive E. coli isolates from human wastewater were also confirmed by E. coli 23S rRNA gene real-time PCR assay. E. coli 23S rRNA real-time amplification was performed in 25 μL reaction mixtures using iQ Supermixes (Bio-Rad Laboratories). The real-time PCR mixtures contained 12.5 μL of Supermixes, 800 nM of each primer, 80 nM of probe and 10 ng (2 μL) of template DNA. For the real-time PCR detection of the E. coli genetic markers, previously published primers and cycling parameters were used with slight modification of primer concentration.25 Real-time PCR amplifications were performed in 20 μL reaction mixtures using SsoFast EvaGreen Supermix (Bio-Rad Laboratories, Richmond, CA). The real-time PCR mixtures contained 10 μL of Supermix, 900 nM of each primer and 10 ng (2 μL) of template DNA from fecal samples and 2 μL of template DNA for E. coli isolates. For each real-time PCR assay, positive (DNA from control strains) and negative (sterile water) controls were included. To separate the specific product from nonspecific products, including primer dimers, melting curve analysis was performed for each real-time PCR run. During melting curve analysis, the temperature was increased
■
RESULTS Host-Specificity and -Sensitivity of the E. coli Genetic Markers. In all, 100 fecal and wastewater samples were collected from 10 target and nontarget host groups. None of the DNA samples from the target and nontarget host groups tested in this study showed the sign of PCR inhibition at a test concentration of 10 ng of DNA. All fecal and wastewater samples were also PCR positive for E. coli 23S rRNA gene at a test concentration of 10 ng of DNA. All four E. coli genetic markers were detected in all human wastewater samples (Table 2). Of the 90 nontarget host groups tested in this study, 85 (94%) were PCR negative for the H8 marker, however, five cattle wastewater samples were PCR positive. Similarly, 77 (85%) were PCR negative for the H12 marker, however, five cattle wastewater, one emu, five kangaroo, and six possum fecal samples were PCR positive. H14 and H24 markers were more frequently detected in nontarget host groups. The overall Table 2. PCR Positive Results for Human WastewaterAssociated E. coli Genetic Markers in Total Fecal DNA Samples from Target and Nontarget Host Groups in Southeast, Queensland, Australia no. of E. coli marker PCR positive samplesb host groupsa Nontarget host groups cat cattle wastewater deer dog emu goat horse kangaroo possum Target Host-Group human wastewater host-specificity host-sensitivity
H8
H12
H14
H24
0 5 0 0 0 0 0 0 0
0 5 0 0 1 0 0 5 6
1 7 7 4 5 0 4 5 6
1 8 0 2 8 1 0 4 1
10 0.94 1.00
10 0.85 1.00
10 0.57 1.00
10 0.72 1.00
a
From each host group 10 samples were included for host-specificity and host-sensitivity analysis. bSamples were tested at a concentration of 10 ng of DNA. C
DOI: 10.1021/acs.est.5b02163 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
host-specific in Japan, little is known regarding their hostspecificity and -sensitivity worldwide. Host-specificity and -sensitivity are two important performance characteristics of genetic markers because markers with low specificity and sensitivity may result in false positive and false negative identification of fecal pollution in environmental waters.11,31 It is desirable that a marker should be highly host-specific (value of 1). However, the specificity of a particular marker may vary from study to study.32−35 Because of this, it has been recommended that a marker with a specificity
