Detection of human fecal pollution in environmental waters using human mitochondrial DNA and correlation with general and human-associated fecal genetic markers

Humanmitochondrial DNA (mtDNA) genetic markers are abundant in sewage and highly human-specific, suggesting a great potential for the environmental application as human fecal pollution indicators. Limited data are available on the occurrence and co-occurrence of human mtDNA with fecal bacterial markers in surface waters, and how the abundance of these markers is influenced by rain events. A 1-year sampling study was conducted in a suburban watershed impacted by human sewage contamination to evaluate the performance of a human mtDNA-based marker along with the bacterial genetic markers for human-associated Bacteroidales (BacHum and HF183) and Escherichia coli. Additionally, the human mtDNA-based assay was correlated with rain events and other markers. The mtDNA marker was detected in 92% of samples (n1⁄4 140) with a mean concentration of 2.96 log10 copies/100 ml throughout the study period. Human mtDNA was detected with greater abundance than human-associated Bacteroidales that could be attributed to differences in the decay of these markers in the environment. The abundance of all markers was positively correlated with rain events, and human mtDNA abundance was significantly correlated with various bacterial markers. In general, these results should support future risk assessment for impacted watersheds, particularly those affected by human fecal pollution, by evaluating the performance of these markers during rain events. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/). doi: 10.2166/wh.2019.197 om http://iwaponline.com/jwh/article-pdf/18/1/8/720524/jwh0180008.pdf 2022 A. B. M. Tanvir Pasha Jessica Hinojosa Duc Phan Adrianne Lopez Vikram Kapoor (corresponding author) Department of Civil and Environmental Engineering, University of Texas at San Antonio, San Antonio, TX 78249, USA E-mail: vikram.kapoor@utsa.edu This article has been made Open Access thanks to the generous support of a global network of libraries as part of the Knowledge Unlatched Select initiative.


INTRODUCTION
Identifying the sources of fecal pollution in environmental surface waters used for human recreation and/or fish breeding is the first step for reducing the potential for human contact with enteric pathogens (Simpson et  Hence, the accurate detection of human fecal pollution is imperative for mitigating human and environmental health risks, reducing economic losses, and maintaining water quality. For over a century, microbiological water quality has been assessed by measuring the densities of two fecal indicator bacteria (FIB), Escherichia coli and Enterococcus spp. As monitoring tools, these fecal bacterial indicators often provide a useful baseline for establishing public health risks.
However, many studies have recognized the limitations of conventional bacteria-based fecal source tracking, including the facts that most of these microorganisms are able to inhabit the intestines of several hosts, including humans, livestock, and wildlife, and are able to grow and survive naturally in the environment ( (Parker et al. ; Sidhu et al. ). Nevertheless, the contribution of fecal pollution due to rainfall is often not considered when assessing water quality, and limited information is available on the correlation between rainfall and source tracking markers, including human mtDNA.
Studies have examined the occurrence of human mtDNA in environmental waters; however, there is still little observational data on the occurrence and co-occurrence of human mtDNA with human-associated Bacteroidales and general E. coli markers in surface waters over extended temporal and spatial scales (Martellini et al. ; Kapoor et al. , ). In this study, qPCR assays targeting human mtDNA, human-associated Bacteroidales markers (HF183 and BacHum), and E. coli were applied to identify and quantify human fecal contamination in environmental surface waters. The Cibolo Creek in the Upper Cibolo Creek (UCC) watershed (Texas, USA) was used for sampling due to a history of significant human fecal contamination (Bass & Burger ). Surface waters of the Cibolo Creek receive discharge from two wastewater treatment systems and are subject to occasional stormwater runoff from nearby agricultural fields during rainfall events.
The objective of this study was to identify the sources of human fecal pollution in a suburban watershed by using general and human-associated fecal genetic markers, as well as human mtDNA as a direct and robust human fecal marker.
The study aimed to understand indicator prevalence in environmental surface waters and what that might suggest about the sources of these indicators. The influence of rainfall events on the abundance of these indicators was also evaluated. In addition, the correlation between human mtDNA, human-associated Bacteroidales markers, and E. coli was studied to better understand the efficacy of each marker for the detection of human fecal pollution. The results of this study provide valuable information on the sources of fecal pollution in the UCC watershed and important insights on the distribution of traditional (E. coli) and more novel indicators (human mtDNA and human-associated Bacteroidales).

Study area and sampling sites
The Cibolo Preserve in the UCC watershed covers approxi- A total of nine sampling sites were selected within the study area based on proximity to the WWTPs, discharge to the Cibolo Creek and in an attempt to cover the extent of the preserve area ( Figure 1). Site 1 is located on the Currey Creek downstream of the STWT; since no flow upstream of the STWT was observed over the study period, no sampling site was selected in the upstream region of the Currey Creek.
Site 2 is located approximately 1 km downstream of site 1 and is the only recreational site in the study area. Sites 7 and 8 are located before and after the contribution from the WWTRC, respectively, and site 3 is located just after the confluence of the Cibolo Creek and the Menger Creek. The WWTRC discharges effluent water into the Menger Creek, which converges with the Cibolo Creek at the preserve boundary. Sites 4 and 5 are located before and after the converging point of the Browns Creek with the Cibolo Creek. Site 6 is located at the eastern edge of the preserve area and was selected to observe the quality of the water before leaving the preserve area. Site 9 is located on the Browns Creek, which converges with the Cibolo Creek inside of the preserve area. Extraction controls with autoclaved distilled water were used during filtration to monitor for potential extraneous DNA contamination. The manufacturer's protocol was followed to extract DNA from the sample membranes using the DNeasy PowerLyzer PowerSoil Kit (Qiagen, Germantown, MD). The purity and concentration of DNA were determined using the NanoDrop One C Spectrophotometer (Thermo Scientific, Wilmington, DE). DNA extracts were stored at À20 C for subsequent qPCR analyses.

qPCR analyses
TaqMan qPCR assays were used to measure both the presence and relative abundance of the four genetic markers in environmental water samples (Table 1) The methodology for the qPCR assays performed is described in detail in a previous study (Kapoor et al. ).  Creek. This map was created using the ESRI ArcGIS.
cross-contamination for each qPCR run. In addition, undiluted and 10-fold dilutions for each DNA extract were used as the DNA template in qPCRs to test for PCR inhibition.

Data analyses
All data for each genetic marker were calculated as the copy number per 100 ml of water for every sample analyzed using qPCR with a cycle threshold (C T ) value above background.
Before performing statistical analyses, all data were transformed to log 10 copies per 100 ml of the water sample. The

Performance of qPCR assays
The linear range and amplification efficiencies of the qPCR assays were determined by plotting the standard curves generated using serial dilutions of known copy numbers of each marker. The linear range of quantification for all the qPCR assays was between 10 and 10 6 copies per reaction. The qPCR amplification efficiencies for all the assays ranged from 86 to 105%, with r 2 values above 0.9. To determine PCR inhibition, reactions were performed with 10-fold dilutions of each DNA extract as described in a previous study (Kapoor et al. ). In these tests, PCR inhibition did not interfere with the amplification efficiency, since a C T value proportional to a 10-fold dilution relative to the undiluted DNA templates was observed for the reactions.
DNA extraction controls and no template controls indicated the absence of contamination in the qPCR assays.

Detection of human mtDNA and fecal bacterial markers
The concentrations of human mtDNA, human-associated Bacteroidales, and E. coli were measured for the water samples using TaqMan qPCR assays. The human mtDNA marker was detected in 92% of water samples (  samples, and both human mtDNA and HF183 were detected concurrently in 54% (76 out of 140) of water samples.

Spatial distribution of markers
The spatial distribution of the levels of markers across the study sites is represented in Figure 2. The two human-associ- Surprisingly, there were no notable differences in human mtDNA marker levels among the study sites, except for site  (Carey & Migliaccio ), and they suggested that excessive nutrients like nitrogen and phosphorus from the effluent water may be responsible for the high algae growth. These data suggest effluent from the WWTRC as a point source of contamination.

Correlation of markers with rainfall events
To discern the effect of rainfall events, the precipitation data reported 24 h before sample collection was obtained for each sampling event. The average concentrations for all markers at sites 2, 3, and 4 were used for the analysis due to their proximity to the USGS station. In total, there were five rainfall events with rainfall ranging from 0.51 to 41.91 mm of rain in a 24 h period. Results for marker copy number per 100 ml volume are shown in Figure 3 along with plots of the rainfall throughout the study. One-way analysis of variance (ANOVA) was performed on all fecal markers to assess if the concentrations of the markers were significantly affected by rainfall events. The abundance of all markers was positively correlated with rain events, indicating that the indicator marker concentration increased after a rainfall event. Furthermore, the abundance of all markers was found to be considerably different (p < 0.05) between dry and wet weather conditions.
The results of this study suggest that the loading of markers into surface waters may be influenced by precipitation events. As suggested in previous studies, it is possible that sediments and runoff could contribute to the input of bacterial markers into surface waters. Results presented here are consistent with previous studies in which higher numbers of bacterial markers were detected after rainfall events

Correlation of human mtDNA and fecal bacterial markers
Spearman's rank correlation coefficients were calculated between each pair of human mtDNA and fecal bacterial markers as measured by the qPCR (Table 4)  The detection of mitochondrial genes via PCR-based assays can be used to identify animal waste directly through its own discharged eukaryotic cells. In addition to our study, the application of mtDNA for fecal source tracking has also been demonstrated in a few recent studies. For example, qPCR assays targeting mitochondrial genes have been used to detect human-, bovine-, and swine-specific contamination