Longitudinal Comparison of Bacterial Diversity and Antibiotic Resistance Genes in New York City Sewage

Abstract

Bacterial resistance to antibiotics is a pressing health issue around the world, not only in health care settings but also in the community and environment, particularly in crowded urban populations. The aim of our work was to characterize the microbial populations in sewage and the spread of antibiotic resistance within New York City (NYC). Here, we investigated the structure of the microbiome and the prevalence of antibiotic resistance genes in raw sewage samples collected from the fourteen NYC Department of Environmental Protection wastewater treatment plants, distributed across the five NYC boroughs. Sewage, a direct output of anthropogenic activity and a reservoir of microbes, provides an ecological niche to examine the spread of antibiotic resistance. Taxonomic diversity analysis revealed a largely similar and stable bacterial population structure across all the samples, which was found to be similar over three time points in an annual cycle, as well as in the five NYC boroughs. All samples were positive for the presence of the seven antibiotic resistance genes tested, based on real-time quantitative PCR assays, with higher levels observed for tetracycline resistance genes at all time points. For five of the seven genes, abundances were significantly higher in May than in February and August. This study provides characteristics of the NYC sewage resistome in the context of the overall bacterial populations.IMPORTANCE Urban sewage or wastewater is a diverse source of bacterial growth, as well as a hot spot for the development of environmental antibiotic resistance, which can in turn influence the health of the residents of the city. As part of a larger study to characterize the urban New York City microbial metagenome, we collected raw sewage samples representing three seasonal time points spanning the five boroughs of NYC and went on to characterize the microbiome and the presence of a range of antibiotic resistance genes. Through this study, we have established a baseline microbial population and antibiotic resistance abundance in NYC sewage which can prove to be very useful in studying the load of antibiotic usage, as well as for developing effective measures in antibiotic stewardship.

Keywords: New York City; antibiotic resistance; microbiome; sewage.

FIG 1

Geographical distributions of NYC WWTPs. This New York City map indicates the drainage areas for the 14 WWTPs across the five NYC boroughs from which samples were obtained for this study. The map was plotted with the Maptools package in R using GIS data obtained from Open Sewer Atlas NYC (http://openseweratlas.tumblr.com/data).

FIG 2

Alpha diversity in 102 sequenced sewage samples. Rarefaction was based on (A) phylogenetic diversity. (B) Observed number of OTUs. The x axes represent the numbers of sequences used in the rarefaction analysis. (C) Box-and-whisker plots depicting the species-level richness and diversity of OTUs based on the Shannon diversity index in the sewage samples. Within the boxes, the central line represents the median values, and the upper and lower boundaries indicate the 75th and 25th percentiles, respectively. The whiskers show the maximum and minimum values. Top panels show results based on location (NYC borough); bottom panels show results based on sampling season. Statistical significance was tested by one-way analysis of variance using the Kruskal-Wallis test, followed by multiple comparisons using Dunn’s test (****, P < 0.0001).

FIG 3

Beta diversity in sewage samples. PCoA results represent unweighted (A and C) and weighted (B and D) UniFrac distances by NYC major drainage area (lower panels) and sampling season (upper panels). Statistical significance was tested by the Adonis test using the Vegan package in R. Each of the data sets indicated distinctions between the tested groups at P < 0.001.

FIG 4

Relative taxonomic abundances of the assignments in 102 sewage samples. Samples are grouped by family according to NYC borough and sampling season. Taxonomy was assigned to the identified OTUs based on Greengenes database v13.8. Only families represented by >3% abundance are listed.

FIG 5

Differentially abundant OTUs identified in 102 sewage samples. LEfSe analyses were performed based on NYC borough (A) and sampling time point (B). Significantly abundant OTUs were determined based on an alpha value of <0.05 and a logarithmic LDA score (effect size) of >4.0.

FIG 6

16S rRNA gene copies in NYC sewage samples. (A) Abundances normalized against the volume of sample used. (B) Biomass concentrations of the sewage samples determined by the ratio of the 16S rRNA copies to the total DNA extracted (ng) for each sample. Each box represents a scatter of the mean values of the 17 replicated sewage samples collected at that particular time point. Error bars indicate the standard deviations at each time point. Statistical significance was determined by using a one-factor Friedman’s test, followed by Dunn’s multiple-comparison test (**, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 7

Antibiotic resistance gene representation in proportion to the 16S rRNA gene copies in the NYC sewage samples. Collections were made during February, May, and August 2015. (A) A Circos diagram represents the percent breakdown of the presence of the measured ARG concentrations in each month and vice versa indicated by the connecting ribbons. (B) Ratios were calculated by normalization of antibiotic resistance gene abundance against the 16S rRNA gene abundance for each sample at each time point. Each box represents a scatter of the mean values of the ratios for the 17 replicated sewage samples collected at that time point. The bars indicate the standard deviations at each time point. Statistical significance was determined using a one-factor Friedman’s test, followed by Dunn’s multiple-comparison test (**, P < 0.01; ***, P < 0.001; ****, P < 0.0001).