<br/>Wastewater-based epidemiology has been proposed as an approach for measuring and monitoring population-level activity and public health. Despite its growing interest, however, the spatial and temporal variability of biomarkers found in wastewater remains relatively unknown, limiting the generalizability of single observations. Here, using untargeted metabolomics we analyzed longitudinal wastewater samples collected from three different buildings in a neighborhood over a period of three months. Sampling intervals commenced with near daily sampling for the first month, followed by two months of sampling on Wednesdays and Saturdays. Via liquid chromatography mass spectrometry, a total of 1,440 mass-to-charge features were obtained and, when possible, matched to putative compound names. Metabolites were stratified to separate highly variable from stable metabolites. Of the stable metabolites, 687 features were shared between all three buildings (78.9%, 74.9% and 72.9% of the stable feature) while 303 of the non-stable metabolites were shared (57.9%, 53.3%, and 60.8% of all unstable features). For stable and shared metabolites, the intensities of 25 were found to be statistically significant between the three buildings (false discovery rate corrected, Kruskal-Wallis test with p=0.00001), indicating that the majority of shared metabolites remain relatively constant between buildings, throughout the course of sampling. For non-stable but shared metabolites, three sub-classes were defined: those with peak intensities larger than the max in the stable metabolites, those with a median intensity below a low threshold and the remaining highly variable metabolites. These data indicate that numerous metabolites undergo significant day-to-day variation, commenting on the variety of human activities. To compare metabolite dynamics through time, feature time courses were z-normalized and distances calculated with a euclidean distance. This revealed minimal shared building-to-building metabolite dynamics (fewer than 10 per building combination) despite the large numbers of shared types of metabolites (stable or unstable). In contract, internal to the buildings, a significant proportion of metabolites displayed overlapping temporal dynamics with several general classes of dynamics being observed. These results reveal a rich, building-dependent metabolite profile and speak to the potential of near-source wastewater monitoring for understanding community health.

Spatiotemporal dynamics of metabolites in residential wastewater revealed through longitudinal sampling and untargeted metabolomics

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.