Development of Indicators for Human Fecal Pollution Using Deep-Sequencing of Microbial Communities

Abstract

The gut microbiome is shaped by both host physiology and environmental factors, which results in unique communities that contain certain members specific to a host. Microbial source tracking (MST) methods that rely on host-specific fecal microorganisms have been applied to detect human fecal pollution over the past two decades. The most common approach uses quantitative polymerase chain reaction (qPCR) to amplify certain sequences of these microorganisms, or human fecal markers. To date, most bacterial human fecal markers have focused on the HF183 cluster within the genus Bacteroides. However, cross-reactions with animals or variable Bacteroides abundance in humans have been found. In addition, the traditional clone library method to identify fecal markers targets order Bacteroidales, thereby excluding other taxonomic groups that might also contain host-specific members. Here we employed deep 16S ribosomal RNA (rRNA) gene next-generation sequencing (NGS) of sewage and animal fecal samples (n=469) to explore human-specific microorganisms. Multiple marker candidates were identified from the family Lachnospiraceae and non-HF183 cluster of Bacteroides. Assays were developed for two human-associated Lachnospiraceae (i.e., Lachno3 and Lachno12) and two sewer pipe-derived Bacteroides (i.e., BacV4V5-1 and BacV6-21). Validation studies of these qPCR assays in host and non-host samples demonstrated their specificity to human fecal source. Low-level animal cross-reactions have been reported for all bacterial human fecal markers, including our newly identified human- and sewage-associated markers; however, the mechanism is poorly understood. We examined cross-reactivity in 180 animal fecal samples using NGS and qPCR assays (i.e., Lachno3, multiplexed Escherichia coli and human Bacteroides, and multiplexed Enterococcus spp. and BacV6-21). All three human fecal markers showed over 90% specificity in both NGS and qPCR results. Human marker cross-reactions could correlate with certain composition of its corresponding genus and could putatively correlate with environmental factors. In particular, discrepancies between NGS and qPCR marker positives could primarily be explained by amplification of the marker’s closely-related organisms. Overall, this work provided a new generation of reliable human fecal markers, identified mechanisms for their cross-reactions both ecologically and technically and highlighted the utility of deep sequencing of microbial communities for MST method development.