Article pubs.acs.org/est
Ruminants Contribute Fecal Contamination to the Urban Household Environment in Dhaka, Bangladesh Angela R. Harris,† Amy J. Pickering,†,‡ Michael Harris,§ Solaiman Doza,∥ M. Sirajul Islam,∥ Leanne Unicomb,∥ Stephen Luby,‡ Jennifer Davis,†,‡ and Alexandria B. Boehm*,† †
Environmental and Water Studies, Department of Civil and Environmental Engineering, Stanford University, Stanford, California 94305, United States ‡ Woods Institute for the Environment, Stanford University, Stanford, California 94305, United States § Emmett Interdisciplinary Program in Environment and Resources, Stanford University, Stanford, California 94305, United States ∥ icddr,b, Dhaka 1000, Bangladesh S Supporting Information *
ABSTRACT: In Dhaka, Bangladesh, the sensitivity and specificity of three human, three ruminant, and one avian source-associated QPCR microbial source tracking assays were evaluated using fecal samples collected on site. Ruminantassociated assays performed well, whereas the avian and human assays exhibited unacceptable cross-reactions with feces from other hosts. Subsequently, child hand rinses (n = 44) and floor sponge samples (n = 44) from low-income-households in Dhaka were assayed for fecal indicator bacteria (enterococci, Bacteroidales, and Escherichia coli) and a ruminant-associated bacterial target (BacR). Mean enterococci concentrations were of 100 most probable number (MPN)/2 hands and 1000 MPN/225 cm2 floor. Mean concentrations of Bacteroidales were 106 copies/ 2 hands and 105 copies/225 cm2 floor. E. coli were detected in a quarter of hand rinse and floor samples. BacR was detected in 18% of hand rinse and 27% of floor samples. Results suggest that effective household fecal management should account not only for human sources of contamination but also for animal sources. The poor performance of the human-associated assays in the study area calls into the question the feasibility of developing a human-associated marker in urban slum environments, where domestic animals are exposed to human feces that have been disposed in pits and open drains.
■
INTRODUCTION Dhaka, Bangladesh is one of the largest and fastest growing cities in the world.1 Only about 2% of fecal waste is effectively managed in Dhaka, with the remaining 98% directly entering the environment.2 Untreated fecal waste can contain high concentrations of human pathogens3 and exposure to this waste is a substantial health risk. For example, humans can shed up to 1012 enteric viruses per gram of feces.4 Individuals generate an estimated 135 to 270 g of feces per day.5 If proper management systems for human waste are not in place, pathogen loads to the environment can be substantial in densely populated settings.5 Animal feces can spread zoonotic bacterial and protozoan diseases to humans.6 A quantitative microbial risk assessment found that recreational exposure to water contaminated with cow feces may represent a similar risk to swimmer health as water contaminated with human feces.7 The model revealed the following pathogens in cow feces pose a substantial risk to human health: Campylobacter jejuni, Giardia spp., Cryptosporidium spp., and Escherichia coli O157:H7.7 Chicken feces can also contain human pathogens, such as Salmonella spp. and Campylobacter spp.8−10 A recent meta-analysis found a positive © XXXX American Chemical Society
relationship between exposure to domestic food producing animals, including chickens, goats, and cows, and diarrheal illness in humans.11 In particular, the study found that exposure to poultry increased the odds of Campylobacter spp. infection 2fold.11 Given that exposure to both human and animal feces can pose a risk to human health, understanding the variation and extent of fecal contamination in a particular environment can yield insight into potential risks and interventions to mediate those risks. Considering that fecal indicator bacteria can originate from various animal and human sources, microbial source tracking assays to detect host-associated enteric bacteria have been developed. Many of these assays are cultureindependent and detect gene-segments of enteric bacteria that are host-specific;12 however, there has been limited testing of the assays in developing countries. Prior to using these assays in Received: December 23, 2015 Revised: March 18, 2016 Accepted: April 5, 2016
A
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
were used to enumerate E. coli (EC) and enterococci (ENT). Serial dilutions were made of the fecal slurries, and three different dilutions of each composite were processed to ensure results within the range of quantification. Samples were processed following the manufacturer’s instructions using Quanti-tray 2000 (IDEXX), with the exception of EC enumeration trays being incubated at 44.5 °C for 18−22 h.23,24 The higher temperature was used because enumeration of fecal coliforms was conducted concurrently, although not discussed in the present study. The concentration of EC and ENT in a fecal composite was calculated from the lowest dilution of the composite sample yielding a quantifiable concentration of the indicator bacteria. Process blanks were created each sampling day, following the procedure described above with the exception of adding feces. Duplicate processing of one fecal composite per fecal source type was conducted. Household Visits. Locally trained enumerators visited study households and conducted interviews with the primary female caregiver and asked questions related to the youngest child in the household. Enumerators conducted interviews in Bengali and recorded answers on a hand-held computer using The Survey System software (Creative Research Systems, Petaluma, CA). Enumerators recorded descriptions of observed household water sources, reported water treatment behaviors, and descriptions of the household’s latrine, including the observed type and the reported number of households sharing it. Enumerators also asked respondents if animals were present within their living compound and then observed whether animals were present inside and within 20 m of the compound. Additionally, enumerators asked respondents about the material type of the child’s sleeping area floor as well as the methods used for cleaning the floor in the past 24 h, such as sweeping or wiping with rag. Enumerators also recorded the observed child activity prior to hand rinse; presence of dirt observed on child hands; respondent reported time since the child last washed his or her hands with soap; and the observed presence of a handwashing station with soap and water in the household. Household Sample Collection and Processing. Bangladeshi enumerators collected hand rinse samples from the youngest child within 44 households (Table S2). Each hand was inserted into a single Whirl-Pak Bag (Nasco, Fork Atkinson, WI) filled with 250 mL of autoclaved (121 °C for 15 min) 1/4-strength Ringer’s solution (Oxoid Ltd., Hampshire, UK), then massaged and agitated for 30 s per hand using established techniques.25,26 A Stanford researcher sampled the floor of the child sleeping area (44 households total, Table S2). An autoclaved (121 °C for 15 min), 225 cm2 aluminum foil stencil was placed on the floor of the child play area and a sterile Whirl-Pak Speci-Sponge (Nasco, Fort Atkinson, WI), prehydrated with 10 mL of neutralizing buffer, was rubbed in a systematic pattern over the surface, and then placed in a sterile Whirl-Pak Bag.27−30 The hand rinse and sponge samples were stored on ice in a cooler until taken to an icddr,b laboratory, where they were processed within 12 h of collection. To elute bacteria from the sponge, 250 mL of sterile 1/4strength Ringer’s solution was added in 3 incremental volumes (100, 100, and 50 mL).29 First, 100 mL of Ringer’s solution was added to the sponge bag. The sponge was then agitated in the bag for 15 s and then massaged for 15s; next, the Ringer’s solution was removed and added to a new Whirl-Pak bag. The next two volumes (100 and 50 mL) were processed in the same manner as the first volume, to create a total of 250 mL rinse of
a new environmental context, a validation study should be conducted to determine their effectiveness in the study area.13,14 In the present study, we are particularly interested in understanding fecal contamination in the household environment in Dhaka, Bangladesh. Previous work has established poor management of fecal waste in the study area2 and relatively high rates of diarrheal illness, with children under 5 years of age having an increased risk of illness relative to persons in all other age groups.15 We first assessed the sensitivity and specificity of selected microbial source tracking assays for their application in Dhaka. Next, we measured fecal contamination levels using both microbial source tracking molecular methods and fecal indicator bacterial assays. We investigated fecal contamination on the hands of children and on household floors, as these surfaces represent potential reservoirs for feces that can contribute to the fecal−oral route of exposure.16−18
■
METHODS Study Site and Sample Frame. The study was conducted in the Mirpur community of Dhaka, Bangladesh (23°42′ N, 90°24′ E). Participating households were randomly selected from the control arm of a cholera vaccine and behavior change randomized controlled trial conducted by icddr,b.19 For our study, we randomly selected (using a computer generated randomization sequence) 59 control households from the cholera vaccine study with children from birth to 24 months of age at the beginning of our sample collection period (July 2012). Informed consent was given by the male or female head of each participating household. Fecal Sample Collection and Processing. Fecal samples were collected from human subjects, as well as from the most commonly kept domesticated animals in the study site: chickens, ducks, cows, goats. Fifteen to twenty individual specimens were collected per fecal source type using a purposive sampling strategy (Table S1). Human samples were collected from study households from individuals of varying age and gender. For each human stool sample, a sample kitwhich consisted of a sterile stool collection tube with scoop, aluminum foil, and gloveswas provided to the household, along with verbal instructions on collecting the stool. A member of the research team returned to the household to collect the sample the following day. For each domesticated animal type, efforts were made to include both males and females, animals of varying age, and animals kept by households located throughout the study site. Sampled animals were not necessarily owned by households interviewed in the study. A sterile spoon and 50 mL centrifuge tube were used to collect >2 g of feces eliminated from each donor, and care was taken to avoid including soil in the sample. Staff attempted to collect fresh animal feces that appeared to be deposited within the past day; however, the exact age of each specimen is unknown. After collection, all fecal samples were stored on ice in a cooler until taken to an icddr,b laboratory, where they were processed within 12 h of collection. Aliquots of three or four individual fecal specimens of the same source type were combined at equal masses to form a 2.0 g composite (Table S1).20−22 Five composite samples were formed per fecal source type. The composite samples were made into fecal slurries in a DNA-sterile 50 mL centrifuge tube using molecular grade water (Thermo Fisher Scientific, Waltham, MA) to form a 20 mL solution. Defined-substrate assays (Colilert-18 and Enterolert, IDEXX, Westbrook, ME) B
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
as well as primer and probe concentrations, as described in the referring manuscript of each assay were used, with the exception of the BacR assay. For the BacR assay, the standard annealing temperature (60 °C) and time (60 s) for the ABI Universal Master Mix was used with 45 cycles. Bovine serum albumin (BSA) was added to all the reaction mixtures (Table S3). The source-associated microbial source tracking assays meeting sensitivity and specificity criteria (described in the Results section) were used for environmental sample analysis, with no more than one assay per source type being selected. All environmental samples were also processed with a molecular indicator for general fecal contamination (GenBac3). The DNA concentrations in the sample extracts were quantified using Nanodrop (Thermo-Scientific, Wilmington, DE) in order to determine whether the extract should be diluted for the qPCR reactions (only DNA extracts with less than 100 ng of nucleic acid per μL were added to qPCR reactions). Each qPCR plate processed included a standard curve run in triplicate with concentrations of standard (Table S3) ranging from 101 copies per μL to 105 copies per μL. Standards were quantified using Nanodrop. Triplicate no-template controls were included with each plate. All fecal samples were processed in triplicate. Environmental samples were processed in duplicate. Cow, goat, human, duck, and chicken duplicate fecal composites were processed with some of the microbial source tracking assays. For more details on molecular processing and tests for assay inhibition, see the SI. Data Analysis: Microbial Source Tracking Assay Performance. Microbial source tracking assay validation was conducted following an established approach.20 The concentration of the molecular target detected in a fecal composite sample was normalized to the amount of feces as measured by MPN of ENT (i.e., copies per MPN ENT). As discussed by Ervin et al.,39 there is no standard approach for quantifying the amount of feces in this type of analysis so we chose to normalize by MPN ENT to follow a previous large collaborative microbial source tracking method evaluation study20 (see the SI for the analysis conducted by normalizing to the amount of feces as measured by ng DNA). The sensitivity metric is that the median concentration of the microbial source tracking molecular target in target host feces should be higher than 10 copies per MPN ENT. A concentration of 10 copies per MPN ENT corresponds to the minimum concentration of the molecular target that would be detectable in a 100 mL environmental sample with 100 MPN ENT/100 mL. This was determined assuming bacteria from 100 mL of a sample were captured on a filter, 100 μL of eluent was obtained during DNA extraction from the filter, and 2 μL of undiluted extract was added to the qPCR reaction that has a lower limit of quantification of 10 copies/μL undiluted extract. In addition, an independent sample t test was used to compare the log10-mean concentrations of the molecular target in target host feces. For the specificity metric, the number of nontarget host fecal composites that had the molecular target detected at concentrations within the range detected in target host fecal samples should be zero. Only concentrations detected within the range of quantification were considered for the specificity analysis, following Boehm et al.20 The sensitivity and specificity of the assays was also determined using molecular results in a binary, presence/absence format. The methods associated with this analysis are found in the SI.
the sponge sample in the new sample bag. The sponge and hand rinse samples were processed for EC and ENT using Colilert-18 and Enterolert, respectively. For the hand rinse sample, 50 mL of the sample was added to 50 mL of distilled water to create the necessary 100 mL sample volume for the IDEXX assays. For the floor sponge eluent, 10 mL of the eluent was added to 90 mL of distilled water to create the 100 mL sample volume for the assays. The samples were processed following the same Colilert-18 and Enterolert methods as used for the fecal samples. Quanti-trays with zero positive wells were recorded as 0.5 most probable number (MPN) per volume tested. Trays with all positive wells were recorded as 2420 MPN per volume assayed. Concentrations of EC and ENT are reported per 2 hands rinsed and per 225 cm2 floor sponged. Each sampling day, laboratory blanks were generated following the same procedure, processing Sterile Ringer’s solution with IDEXX reagents in place of a field sample. Field blanks were also generated by processing unused hand rinse bags (n = 4) and unused Speci-Sponges (n = 2) returning from the household visits in the cooler. Processing of Fecal and Environmental Samples for Molecular Analysis. For each fecal composite, 2 mL of the fecal slurry (equivalent to 0.2 g wet weight of fecal material) was filtered using disposable Nalgene filter funnels (Thermo Fisher Scientific, Waltham, MA) and 47 mm diameter, 0.4 μm pore size polycarbonate membrane filters (Isopore, Millipore, Billerica, MA). For the floor and hand rinse samples, the polycarbonate membrane filters were used in conjunction with reusable filtration funnels sterilized using hydrochloric acid. The majority of the hand and floor samples were processed at 50 mL volumes (more details found in the Supporting Information (SI)). After the samples (i.e., the fecal slurries and environmental samples) were filtered, they were treated with 0.5 mL of RNAlater (Qiagen, Valencia, CA, USA) and allowed to sit for 5 min. The RNAlater was then vacuumed through the filter to remove any residual solution. The filters were then aseptically placed into a 2 mL tube with glass beads (Generite, North Brunswick, NJ, USA) and stored at −20 °C until transported to Stanford University, where they were stored at −80 °C until processed. As per U.S. Department of Agriculture regulations (permit # 122829), the avian fecal samples were heat treated at 74 °C for 30 min prior to their transport to the U.S. Because of time limitations in the laboratory, one human fecal composite slurry was stored at 4 °C for 12 h after it was processed using the IDEXX defined substrate assay and before it was processed using membrane filtration. DNA was extracted from each fecal sample filter and environmental sample filter using the commercial DNA EZ extraction kit (Generite, North Brunswick, NJ, USA). Ten to twenty samples were extracted at a time, and an extraction blank (i.e., no sample filter included) was created with each extraction set. The following established microbial source tracking Taqman qPCR assays were performed on the fecal DNA extracts: GenBac3,31 HF183 Taqman,32 BacHum,33 HumM2,34 BacCow,33 Rum2Bac,35 and BacR.36 These assays were selected because they performed well in previous studies in other regions of the world.14,20,32−37 An avian SYBR green assay termed “GFD” was also used because of its good performance in detecting chicken feces in several geographic areas.38 The master mixes used for the microbial source tracking assays20 are provided in Table S3. Cycling parameters, C
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Data Analysis: Microbial Source Tracking of Environmental Samples. Concentrations of the microbial source tracking molecular targets detected within the range of quantification are reported as copies per 2 child hands rinsed and copies per 225 cm2 floor sponged for the hand and floor samples, respectively. The lower limit of quantification for the environmental samples was 5000 copies per 225 cm2 floor sponged and 5000 copies per 2 child hands rinsed. For the Genbac3 assay, if a sample had the target detected but not quantifiable, a concentration of 2500 copies per 2 hands rinsed or per 225 cm2 floor sponged (i.e., half of the lower limit of quantification) was used to allow for statistical analysis of a continuous outcome. The detection of the host-associated microbial source tracking molecular targets in the environmental samples was analyzed in a binary format. If the sample contained the target, either detected within the range of quantification or detected but not quantifiable, the sample was recorded as positive. If the sample did not have the target detected (i.e., a nondetect), then the sample was recorded as negative. Statistical associations and comparisons were made using independent sample t tests and Fisher’s Exact tests. Results were considered statistically significant at p-values less than 0.05.
were not present in any of the 5 human composites at concentrations higher than the 10 copies per MPN ENT sensitivity threshold. For the BacHum assay, only 1 of 5 human composites had concentrations of the target higher than the sensitivity threshold. Furthermore, the human assay targets were not detected in each human fecal composite. Additionally, none of the human assays satisfied the specificity metric. Composites of goat and chicken (nontarget hosts) had concentrations of the human-associated targets within the same range of concentrations as the human composites (target hosts). HumM2 and BacHum targets were also amplified in duck composites within the same concentration range as the human composites (Figure 1). The avian GFD target was detected in all chicken and duck composite fecal samples; however, only a single duck composite had the avian GFD target detected at a concentration above the sensitivity threshold. The avian GFD assay thus did not satisfy the sensitivity metric. The avian GFD target was also detected in one human fecal composite within the same concentration range as the chicken and duck composite samples (target hosts). Thus, the Avian GFD assay did not satisfy the specificity metric. All three ruminant-associated assaysBacR, Rum2bac, and BacCowsatisfied the sensitivity metric. The BacR target was detected at significantly higher concentrations per MPN ENT than the Rum2bac and BacCow targets in ruminant feces (paired sample t test comparing means of log10-transformed concentrations, p < 0.05). In addition, the BacR and Rum2bac assays had no cross-reactivity with nonruminant fecal samples; these two assays thus also satisfied the specificity metric. The BacCow target amplified in the avian fecal samples, although at lower concentrations than in the ruminant samples. The BacCow assay also satisfied the specificity metric. An analysis of assay sensitivity and specificity using data in a binary format are presented in the SI (Table S5). Sample Household Characteristics. Twenty-two percent of respondents owned their home. The median reported regular weekly expenditure was 1500 taka (18 USD). The majority of households (98%) used water provided by the Dhaka Water Supply and Sewerage Authority (WASA) utility, either through a hand pump, metal pipe connected to a tap, or flexible pipe connected to a tap. Most households used a pourflush toilet shared with their neighbors (88%); shared toilets are not considered improved sanitation based on international monitoring definitions.40 A quarter of the study households reported having animals within their compound. Seven percent (7%) had goats; 10% had chickens; 3% had ducks; and 12% had other types of birds (e.g., quails, pigeons, and parrots) in the compound. Also, enumerators observed animals inside and within 20 m of the household’s living compound. Enumerators observed goats (10% of households), chickens (24%), ducks (10%), cows (3%), and other types of birds (15%). Forty percent of the animals observed were inside the compound. Additional study household descriptive data is provided in Table S6. Household Contamination. In 90% of sample households, the floor of the child sleeping area was made of concrete. The remaining households had dirt (7%, n = 4) or tile/plastic (3%, n = 2) floors. Fifty-four percent (54%) of households reported sweeping the floor the day of sampling, prior to sample collection, and 15% of households reported using water and a mop or towel to clean the floor on the day and prior to sampling. The majority (58%) of children were held by a
■
RESULTS Quality Control and Quality Assurance. Field and technical blanks indicated that contamination was negligible. Biological replicates (duplicate samples) yielded similar results. Tests for inhibition of samples informed whether samples needed to be diluted prior to qPCR (Table S4). Further details on the results from the inhibition testing, as well as results from the blanks and duplicates, can be found in the SI. Concentrations of Fecal Indicator Bacteria in Fecal Samples. Concentrations of EC, ENT, and GenBac3 varied over several orders of magnitude within and between feces from different host species (Table 1, Figure S2). Analyzing all the Table 1. Log10-Mean Concentration (Standard Deviation) of General Fecal Indicators Per Gram Wet-Weight Feces by Fecal Host fecal host
mean (SD) GenBac3 log10 copies per gram feces
chicken duck cow goat human
7.9 (1.5) 7.9 (0.9) 9.8 (0.05) 9.7 (0.2) 10.7 (0.1)
mean (SD) log10 MPN EC per gram feces 8.5 7.6 6.8 7.8 7.7
(0.5) (0.9) (0.7) (0.1) (0.7)
mean (SD) log10 MPN ENT per gram feces 7.8 6.7 3.8 5.8 6.7
(0.9) (0.5) (0.4) (1.0) (1.2)
fecal composites collectively, ENT and EC per gram of feces are positively correlated (Pearson’s r correlation between log10transformed concentrations = 0.48, p = 0.01, n = 25). GenBac3 concentrations did not covary significantly with either EC or ENT concentrations. There were significant differences in EC, ENT, and GenBac3 concentrations in feces between hosts. Details are provided in the SI. Detection of Microbial Source Tracking Targets in Fecal Samples. Figure 1 shows the concentrations of the hostassociated molecular targets normalized by the amount of feces analyzed quantified as MPN ENT by fecal host. The three human assaysHF183, HumM2, and BacHumdid not satisfy the sensitivity metric. The HF183 and HumM2 targets D
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Concentrations of microbial source tracking assay target copies per MPN ENT in fecal material from humans, goats, cows, ducks, and chickens. The target source of the microbial source tracking assay is aligned to the left of the assay tick, and nontarget sources are aligned to the right of the assay tick. Samples in which the microbial source tracking target was detected but not quantified (DNQ), or in which the target was not detected (ND), are plotted at the bottom of y-axis. The red line marks the sensitivity threshold of 10 copies per MPN ENT.
The remaining 13 households with the BacR target detected in an environmental sample reported not having ruminants within their compound. There was also a significant positive association between the observation of goats around the household and BacR target detection in household samples (Fisher’s exact test, p = 0.01).
caregiver prior to having their hands rinsed, based on enumerator observation. Other children were observed playing on the floor or bed (31%), sleeping (9%), or eating (3%) prior to hand rinsing. According to adult caretakers, 15 percent (15%) of children had their hands washed with soap and water within 2 h prior to hand rinse sampling. ENT and GenBac3 were detected in nearly all child hand rinses and household floor samples (Figure 2). ENT were detected in 95% of hand rinse samples and 100% of floors samples, having log10-mean concentrations of 2.0 log10 MPN per 2 hands (standard deviation (SD) = 0.7, n = 44) and 3.0 log10 MPN per 225 cm2 of floor (SD = 1.1, n = 44). The GenBac3 target was detected in 100% of hand rinse and floor samples, having log10-mean concentrations of 6.0 log10 copies per 2 hands (SD = 0.7, n = 44) and 5.2 log10 copies per 225 cm2 of floor (SD = 1.1, n = 44). In contrast, EC was only detected in 25% of the floor samples, and 27% of the hand rinses. In floor samples, ENT and GenBac3 concentrations were positively correlated (Pearson’s r correlation of log10-transformed concentrations, r = 0.64, p < 0.01, n = 44). No other correlations between fecal indicators in samples were statistically significant. The BacR target was detected in 27% of floor samples (12 out of 44) and in 18% of child hand rinse samples (8 out of 44) (Figure 2). The BacR target was detected within the range of quantification of the assay in 11% of floor samples (5 out of 44) and 9% of child hand rinse samples (4 out of 44). The maximum concentration of the BacR target was 5.6 log10 copies per 225 cm2 of floor and 4.5 log10 copies per 2 child hands. In units of copies per MPN ENT, the maximum was 120 BacR copies per MPN ENT for floor samples and 590 BacR copies per MPN ENT for child hand rinse samples. Seventeen households (out of 59, 29%) had the BacR target detected in the child’s hand rinse sample and/or the floor sponge sample. Four households reported having goats in their compound, and all of these households had the BacR, ruminant-associated target detected in either their hand rinse or floor sample (significant positive association between goat ownership and BacR detection, Fisher’s exact test, p = 0.01).
■
DISCUSSION We found evidence of fecal contamination from ruminants on child hands and household floors in the study area. Households in dense urban areas of low-income countries find themselves living in close proximity with their animals, which presents a potential for the spread of zoonotic diseases. The results of our study suggest that households owning ruminants are particularly at risk for ruminant fecal contamination in their household environment. At the same time, we also found that even households not living with ruminants within their compound contained ruminant fecal contamination, suggesting that ruminant contamination is pervasive in this dense, urban, low-income setting. Close contact with animals has been identified as a risk factor associated with children exhibiting signs of environmental enteropathy, a disorder related to impaired growth in children.41 The results of our study further corroborate the recommendation that proper management of nonhuman animal feces should be considered to protect human health in densely populated low-income communities. The sensitivity and specificity of the human-associated assays in our study area are low compared to other studies.14,20,21,42 Notably, Ahmed et al.42 assessed a SYBR version of the HF183 human assay (not tested herein) in Dhaka and found it to perform well, with a sensitivity of 87% and specificity of 93% (percentages reported from a binary analysis). In our study, the human assays’ sensitivities ranged from 40% to 80% and specificities from 40% to 55%, using a binary analysis. However, goat and duck samples were not collected in the Ahmed et al.42 study, and chicken fecal samples were collected from poultry market chickens, not from household chickens as our study did. To determine human-associated fecal assay sensitivity, some researchers use raw sewage instead of individual human fecal E
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Dhaka potentially have on the intestinal bacteria of animals in the area. The human assays target the 16S rRNA gene or a functional gene of intestinal bacteria of the order Bacteroidales. These bacteria are found in high abundance in the intestinal tracks of animals. Endemic distributions (e.g., host-associated genetic clusters) of fecal Bacteroidales have been found in certain geographic settings, resulting in the identification of host-associated gene targets for the use in fecal source identification.43 However, endemic distributions are expected to occur when physical contact between different host-species is absent, thus limiting horizontal transmission of fecal bacteria between hosts.43 The environmental conditions in Dhaka may not promote endemic distributions of fecal bacteria uniquely associated with the human gastrointestinal tract. Poor management of human excreta in Dhaka creates substantial opportunity for domestic animals to consume food and water contaminated with human feces. Such exposure could explain the cross-reactivity of the human-associated assays in nonhuman feces. Analyzing the genetic distributions of fecal bacteria from different animal hosts in Dhaka could inform the feasibility of identifying a human-associated genetic target in an environment where environmental conditions may promote horizontal transmission of fecal bacteria between hosts. Also, future work that explores the horizontal transmission of gut-associated microorganisms, as well as the ecological conditions that promote host-specific distributions of intestinal bacteria, would be useful for determining the potential of identifying a robust human-associated genetic target. Other source tracking methods that do not rely on Bacteroidales species could be explored as well, targeting organisms that exhibit higher hostspecificity. For instance, recent studies found human virus assays and microbial community-based assays (e.g., PhyloChip and TRFLP methods) to exhibit high specificity for human fecal material.20,37 However, viral assays lack sensitivity.44 Our study further highlights the need for an improved human microbial source tracking assay,14 and if one cannot be developed, then direct detection of human fecal pathogens would be required to understand exposure risks to fecal contamination in these environments. The Avian GFD assay has been previously tested only in the US, Canada, and New Zealand.38 In these locations, the assay was 100% specific, as compared to 33% found in our study (specificities determined using binary analysis).38 At our study site, where cattle, humans, and birds lived in close proximity and with widespread fecal contamination, cross-reactivity was observed between the bird-associated marker and human and cow feces. Future work is needed to evaluate this assay in other geographic regions to determine its robustness as an indicator of avian fecal contamination. The BacR and Rum2Bac assays are sensitive and specific for detecting cow and goat fecal contamination in Dhaka, Bangladesh. Ruminant-associated assays have performed well across multiple countries and tend to be a robust indicator of ruminant fecal contamination.14 The unique digestive system of ruminants potentially creates an environmental niche for certain species of bacteria, perhaps explaining the groups of Bacteroidales genetic sequences uniquely associated with ruminant feces.43 The results of our study agree with similar studies conducted in the United States20 and Europe.35,36 A study in India found similar performance of the BacCow assay as was found in our study, with cross-reactivity occurring in chicken and dog fecal samples.21
Figure 2. Concentrations of ENT (MPN) and Genbac3 (target copies) per 2 hands rinsed and per 225 cm2 floor sponged. Box plots show the range of 25th to 75th percentile of samples as the bottom and top edges of the box, with the center line of the box displaying the median concentration. The top and bottom lines show the 10th and 90th percentile range. Percentage of samples tested that were positive for EC and the BacR marker in hand and floor samples is represented by the proportion of the circle that is shaded black.
specimens, as sewage integrates the feces of many individuals. In some cases, the sensitivities of the human-associated molecular targets are higher when evaluated for detection in raw sewage as compared to individual fecal samples.21 In Dhaka, only a minority of households are connected to sewerage systems,2 so raw sewage was not sampled in this study. The poor performance of the human fecal assays in our study calls into question the feasibility of developing a humanassociated marker for an urban slum environment where domestic animals are exposed to human feces that are disposed of in pits and open drains. The poor performance of the human fecal assays may be explained by the influence the environmental conditions in F
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Stanford Graduate Fellowship. We also acknowledge Dan Wang, Ehteshamul Islam, Zahid Mahmud, Shabiha Begum, Shamima Islam, Nusrat Islam, Razia Shultana, Rashedul Islam, and Shamim Khan.
There are several limitations associated with this study. The assays selected for the microbial source tracking validation were assays that exhibited high sensitivity and specificity in other study areas. However, there are some existing human and ruminant assays that we did not test which could potentially be more effective in detecting and distinguishing feces in our study area. We also tested the assays on a small number of fecal specimens, although on par with other validation studies.20,33,34,45,46 We created composites to increase the number of individual specimens tested; however, this limits our ability to understand variation across individuals. The relationship between the detection of microbial source tracking targets and infectious pathogens in various environmental media also remains unclear. This information would be useful to inform the health risks associated with the detection of these targets in environmental samples. In addition, improved understanding of the decay rates of microbial source tracking targets in media other than water is needed. Further work assessing microbial source tracking target decay rates compared to general fecal indicators and fecal pathogens is important to understanding how to interpret the detection of these targets and if they serve as effective indicators of fecal contamination and health risks. Microbial source tracking could be a valuable tool for evaluating the impact of water, sanitation, and hygiene interventions in low-income countries.42,47 Many studies take measurements of fecal contamination using fecal indicator bacteria to determine if the intervention reduces environmental contamination, such as in water or on hands or surfaces. If an intervention only targets fecal contamination from a human source (e.g., improving household toilets), using reduction of fecal indicator bacteria (which can come from a variety of different sources) as an outcome measure could lead to inaccurate conclusions about intervention effectiveness. The results of our study suggest that nonhuman fecal contamination exists in the environment of low-income, urban households. Therefore, general fecal indicators may not serve as an accurate measure of fecal contamination from human sources in these settings.
■
■
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b06282. Details on blanks, duplicates, inhibition, sample collection, molecular methods, indicator concentrations in feces, and measured hand and floor contamination (Tables S1−S5) (PDF).
■
REFERENCES
(1) Alirol, E.; Getaz, L.; Stoll, B.; Chappuis, F.; Loutan, L. Urbanisation and infectious diseases in a globalised world. Lancet Infect. Dis. 2011, 11 (2), 131−141. (2) Peal, A.; Evans, B.; Blackett, I.; Hawkins, P.; Heymans, C. Fecal sludge management (FSM): analytical tools for assessing FSM in cities. J. Water, Sanit. Hyg. Dev. 2014, 4 (3), 371−383. (3) Feachem, R. G.; Bradley, D. J.; Garelick, H.; Mara, D. D. Sanitation and Disease: Health aspects of excreta and wastewater management; Washington, DC, 1983. (4) Atmar, R. L.; Opekun, A. R.; Gilger, M. a; Estes, M. K.; Crawford, S. E.; Neill, F. H.; Graham, D. Y. Norwalk virus shedding after experimental human infection. Emerging Infect. Dis. 2008, 14 (10), 1553−1557. (5) Mara, D. Domestic Wastewater Treatment in Developing Countries; Earthscan: London, 2004; p 290. (6) WHO. Waterborne Zoonoses: Identification, Causes, and Control; WHO: London, 2004. (7) Soller, J. a; Schoen, M. E.; Bartrand, T.; Ravenscroft, J. E.; Ashbolt, N. J. Estimated human health risks from exposure to recreational waters impacted by human and non-human sources of faecal contamination. Water Res. 2010, 44 (16), 4674−4691. (8) Bull, S. A.; Allen, V. M.; Domingue, G.; Jørgensen, F.; Ure, R.; Whyte, R.; Tinker, D.; Corry, J. E. L.; Humphrey, T. J. Sources of Campylobacter spp. Colonizing Housed Broiler Flocks during Rearing. Appl. Environ. Microbiol. 2006, 72 (1), 645−652. (9) Cox, N. A.; Stern, N. J.; Musgrove, M. T.; Bailey, J. S.; Craven, S. E.; Cray, P. F.; Buhr, R. J.; Hiett, K. L. Prevalence and Level of Campylobacter in Commercial Broiler Breeders (Parents) and Broilers. J. Appl. Poult. Res. 2002, 11, 187−190. (10) Hutchison, M. L.; Walters, L. D.; Avery, S. M.; Munro, F.; Moore, A. Analyses of Livestock Production, Waste Storage, and Pathogen Levels and Prevalences in Farm Manures. Appl. Environ. Microbiol. 2005, 71 (3), 1231−1236. (11) Zambrano, L. D.; Levy, K.; Menezes, N. P.; Freeman, M. C. Human diarrhea infections associated with domestic animal husbandry: a systematic review and meta-analysis. Trans. R. Soc. Trop. Med. Hyg. 2014, 108, 313−325. (12) Bernhard, A. E.; Field, K. G. Identification of Nonpoint Sources of Fecal Pollution in Coastal Waters by Using Host-Specific 16S Ribosomal DNA Genetic Markers from Fecal Anaerobes. Appl. Environ. Microbiol. 2000, 66 (4), 1587−1594. (13) Vierheilig, J.; Farnleitner, A. H.; Kollanur, D.; Blöschl, G.; Reischer, G. H. High abundance of genetic Bacteroidetes markers for total fecal pollution in pristine alpine soils suggests lack in specificity for feces. J. Microbiol. Methods 2012, 88 (3), 433−435. (14) Reischer, G. H.; Ebdon, J. E.; Bauer, J. M.; Schuster, N.; Ahmed, W.; Åstrom, J.; Blanch, A. R.; Byamukama, D.; Coakley, T.; Ferguson, C.; et al. Performance Characteristics of qPCR Assays Targeting Human- and Ruminant-Associated Bacteroidetes for Microbial Source Tracking across Sixteen Countries on Six Continents. Environ. Sci. Technol. 2013, 47, 8548−8556. (15) Chowdhury, F.; Khan, I. A.; Patel, S.; Siddiq, A. U.; Saha, N. C.; Khan, A. I.; Saha, A.; Cravioto, A.; Clemens, J.; Qadri, F.; et al. Diarrheal Illness and Healthcare Seeking Behavior among a Population at High Risk for Diarrhea in Dhaka, Bangladesh. PLoS One 2015, 10 (6), e0130105. (16) Curtis, V.; Cairncross, S. Reviews Effect of washing hands with soap on diarrhoea risk in the community: a systematic review. Lancet Infect. Dis. 2003, 3, 275−281. (17) Cattaneo, M. D.; Galiani, S.; Gertler, P. J.; Martinez, S.; Titiunik, R. Housing, Health, and Happiness. American Economic Journal: Economic Policy. 2009, 1 (1), 75−105.
AUTHOR INFORMATION
Corresponding Author
*A. B. Boehm. Phone: 650-724-9128. Fax: 650-723-7058. Address: Jerry Yang & Akiko Yamazaki Environment & Energy Building, 473 Via Ortega, Room 189 MC: 4020, Stanford, CA 94305. Email:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by funding from UPS, USAID, and the Center for International Security and Cooperation at Stanford University. While conducting this work, Angela Harris was funded by a NSF Graduate Research Fellowship and a G
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology (18) Pickering, A. J.; Julian, T. R.; Marks, S. J.; Mattioli, M. C.; Boehm, A. B.; Schwab, K. J.; Davis, J. Fecal contamination and diarrheal pathogens on surfaces and in soils among Tanzanian households with and without improved sanitation. Environ. Sci. Technol. 2012, 46 (11), 5736−5743. (19) Qadri, F.; Ali, M.; Chowdhury, F.; Khan, A. I.; Saha, A.; Khan, I. A.; Begum, Y. A.; Bhuiyan, T. R.; Chowdhury, M. I.; Uddin, M. J.; et al. Feasibility and effectiveness of oral cholera vaccine in an urban endemic setting in Bangladesh: a cluster randomised open-label trial. Lancet 2015, 386, 1362−1371. (20) Boehm, A. B.; Van De Werfhorst, L. C.; Griffith, J. F.; Holden, P. a; Jay, J. a; Shanks, O. C.; Wang, D.; Weisberg, S. B. Performance of forty-one microbial source tracking methods: a twenty-seven lab evaluation study. Water Res. 2013, 47 (18), 6812−6828. (21) Odagiri, M.; Schriewer, A.; Hanley, K.; Wuertz, S.; Misra, P. R.; Panigrahi, P.; Jenkins, M. W. Validation of Bacteroidales quantitative PCR assays targeting human and animal fecal contamination in the public and domestic domains in India. Sci. Total Environ. 2015, 502, 462−470. (22) Green, H. C.; Haugland, R. a; Varma, M.; Millen, H. T.; Borchardt, M. a; Field, K. G.; Walters, W. a; Knight, R.; Sivaganesan, M.; Kelty, C. a; et al. Improved HF183 quantitative real-time PCR assay for characterization of human fecal pollution in ambient surface water samples. Appl. Environ. Microbiol. 2014, 80 (10), 3086−3094. (23) Yakub, G. P.; Castric, D. A.; Stadterman-knauer, K. L.; Tobin, M. J.; Blazina, M.; Heineman, T. N.; Yee, G. Y.; Frazier, L. Applications Evaluation Defined of Colilert Substrate and Enterolert for Methodology Applications Wastewater. Water Environ. Res. 2002, 74 (2), 131−135. (24) Characklis, G. W.; Dilts, M. J.; Simmons, O. D.; Likirdopulos, C. a; Krometis, L.-A. H.; Sobsey, M. D. Microbial partitioning to settleable particles in stormwater. Water Res. 2005, 39 (9), 1773− 1782. (25) Ram, P. K.; Jahid, I.; Halder, A. K.; Nygren, B.; Islam, M. S.; Granger, S. P.; Molyneaux, J. W.; Luby, S. P. Variability in hand contamination based on serial measurements: implications for assessment of hand-cleansing behavior and disease risk. Am. J. Trop. Med. Hyg. 2011, 84 (4), 510−516. (26) Pickering, A. J.; Davis, J.; Walters, S. P.; Horak, H. M.; Keymer, D. P.; Mushi, D.; Strickfaden, R.; Chynoweth, J. S.; Liu, J.; Blum, A.; et al. Hands, water, and health: fecal contamination in Tanzanian communities with improved, non-networked water supplies. Environ. Sci. Technol. 2010, 44 (9), 3267−3272. (27) Dorsa, W. J.; Cutter, C. N.; Siragusa, G. R. Evaluation of six sampling methods for recovery of bacteria from beef carcass surfaces. Lett. Appl. Microbiol. 1996, 22, 39−41. (28) Dorsa, W. J.; Siragusa, G. R.; Cutter, C. N.; Berry, E. D. Efficacy of using a sponge sampling method to recover low levels of Escherichia coli O157: H7, Salmonella typhimurium, and aerobic bacteria from beef carcass surface tissue. Food Microbiol. 1997, 14, 63−69. (29) Buttner, M. P.; Cruz, P.; Stetzenbach, L. D.; Cronin, T. Evaluation of two surface sampling methods for detection of Erwinia herbicola on a variety of materials by culture and quantitative PCR. Appl. Environ. Microbiol. 2007, 73 (11), 3505−3510. (30) Sites, J. E.; Walker, P. N.; Burke, A.; Annous, B. A. Development and Validation of a Pilot Scale Enhanced Biosafety Level Two Containment for Performance Evaluation of Produce Disinfection Technologies. Applied Biosafety 2008, 13 (1), 30−44. (31) Siefring, S.; Varma, M.; Atikovic, E.; Wymer, L.; Haugland, R. A. Improved real-time PCR assays for the detection of fecal indicator bacteria in surface waters with different instrument and reagent systems. J. Water Health 2008, 6 (2), 225−237. (32) Haugland, R. A.; Varma, M.; Sivaganesan, M.; Kelty, C.; Peed, L.; Shanks, O. C. Evaluation of genetic markers from the 16S rRNA gene V2 region for use in quantitative detection of selected Bacteroidales species and human fecal waste by qPCR. Syst. Appl. Microbiol. 2010, 33 (6), 348−357. (33) Kildare, B. J.; Leutenegger, C. M.; McSwain, B. S.; Bambic, D. G.; Rajal, V. B.; Wuertz, S. 16S rRNA-based assays for quantitative
detection of universal, human-, cow-, and dog-specific fecal Bacteroidales: A Bayesian approach. Water Res. 2007, 41 (16), 3701−3715. (34) Shanks, O. C.; Kelty, C. A.; Sivaganesan, M.; Varma, M.; Haugland, R. A. Quantitative PCR for genetic markers of human fecal pollution. Appl. Environ. Microbiol. 2009, 75 (17), 5507−5513. (35) Mieszkin, S.; Yala, F.; Joubrel, R.; Gourmelon, M. Phylogenetic analysis of Bacteroidales 16S rRNA gene sequences from human and animal effluents and assessment of ruminant faecal pollution by realtime PCR. J. Appl. Microbiol. 2010, 108 (3), 974−984. (36) Reischer, G. H.; Kasper, D. C.; Steinborn, R.; Mach, R. L.; Farnleitner, A. H. Quantitative PCR method for sensitive detection of ruminant fecal pollution in freshwater and evaluation of this method in alpine karstic regions. Appl. Environ. Microbiol. 2006, 72 (8), 5610− 5614. (37) Mayer, R. E.; Bofill-Mas, S.; Egle, L.; Reischer, G. H.; Schade, M.; Fernandez-Cassi, X.; Fuchs, W.; Mach, R. L.; Lindner, G.; Kirschner, a.; et al. Occurrence of human-associated Bacteroidetes genetic source tracking markers in raw and treated wastewater of municipal and domestic origin and comparison to standard and alternative indicators of faecal pollution. Water Res. 2016, 90, 265− 276. (38) Green, H. C.; Dick, L. K.; Gilpin, B.; Samadpour, M.; Field, K. G. Genetic markers for rapid PCR-based identification of gull, Canada goose, duck, and chicken fecal contamination in water. Appl. Environ. Microbiol. 2012, 78 (2), 503−510. (39) Ervin, J. S.; Russell, T. L.; Layton, B. a; Yamahara, K. M.; Wang, D.; Sassoubre, L. M.; Cao, Y.; Kelty, C. a; Sivaganesan, M.; Boehm, A. B.; et al. Characterization of fecal concentrations in human and other animal sources by physical, culture-based, and quantitative real-time PCR methods. Water Res. 2013, 47 (18), 6873−6882. (40) Joint Monitoring Program. Meeting the MDG Drinking Water and Sanitation Target: The Urban and Rural Challenge of the Decade; Geneva, 2006. (41) George, C. M.; Oldja, L.; Biswas, S. K.; Perin, J.; Lee, G. O.; Ahmed, S.; Haque, R.; Sack, R. B.; Parvin, T.; Azmi, I. J.; et al. Fecal Markers of Environmental Enteropathy are Associated with Animal Exposure and Caregiver Hygiene in Bangladesh. Am. J. Trop. Med. Hyg. 2015, 93 (2), 269−275. (42) Ahmed, W.; Yusuf, R.; Hasan, I.; Goonetilleke, A.; Gardner, T. Quantitative PCR assay of sewage-associated Bacteroides markers to assess sewage pollution in an urban lake in Dhaka, Bangladesh. Can. J. Microbiol. 2010, 56, 838−845. (43) Dick, L. K.; Bernhard, A. E.; Brodeur, T. J.; Domingo, J. W. S.; Simpson, J. M.; Walters, S. P.; Field, K. G. Host Distributions of Uncultivated Fecal Bacteroidales Bacteria Reveal Genetic Markers for Fecal Source Identification. Appl. Environ. Microbiol. 2005, 71 (6), 3184−3191. (44) Harwood, V. J.; Boehm, A. B.; Sassoubre, L. M.; Vijayavel, K.; Stewart, J. R.; Fong, T.-T.; Caprais, M.-P.; Converse, R. R.; Diston, D.; Ebdon, J.; et al. Performance of viruses and bacteriophages for fecal source determination in a multi-laboratory, comparative study. Water Res. 2013, 47 (18), 6929−6943. (45) Jenkins, M. W.; Tiwari, S.; Lorente, M.; Gichaba, C. M.; Wuertz, S. Identifying human and livestock sources of fecal contamination in Kenya with host-specific Bacteroidales assays. Water Res. 2009, 43 (19), 4956−4966. (46) Ahmed, W.; Masters, N.; Toze, S. Consistency in the host specificity and host sensitivity of the Bacteroides HF183 marker for sewage pollution tracking. Lett. Appl. Microbiol. 2012, 55 (4), 283− 289. (47) Knappett, P. S. K.; Escamilla, V.; Layton, A.; McKay, L. D.; Emch, M.; Williams, D. E.; Huq, R.; Alam, J.; Farhana, L.; Mailloux, B. J.; et al. Impact of population and latrines on fecal contamination of ponds in rural Bangladesh. Sci. Total Environ. 2011, 409 (17), 3174− 3182.
H
DOI: 10.1021/acs.est.5b06282 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
