GLOBAL SURVEY OF MICROBIAL POPULATIONS IN UNTREATED DRINKING WATER SOURCES

Over 800 million people worldwide do not have access to basic drinking water services, obtaining their water from unimproved sources or surface waters. Over 800,000 deaths per year are attributed to waterborne illnesses. Advances in DNA sequencing technologies have revolutionized the characterization of microbial communities in numerous environments, including drinking water sources, but meta-analysis of data sets for patterns in microbial communities that may be related to the quality of the water is confounded by the acquisition of data by many different labs and through many different protocols. In an effort to understand how microbial communities found in drinking water sources across the globe are related, we have designed a unified sampling approach that characterizes three components of water samples: 1) bacterial communities through 16S rRNA sequencing, 2) suspended load particulate matter through SEM-EDS, and 3) dissolved metals through ICP-OES. In the field, trained NGO personnel captured the microbial communities and suspended particulate matter by filtration through 0.1 micron hollow-fiber membrane filters; filtered water was exposed to metal chelating polyurethane foam blocks to capture dissolved metals. In the laboratory, filters were back flushed using high pressure to remove microbes and trapped sediment. The resulting back flushed material was processed for DNA isolation and amplicon sequencing of the V4 region of 16S rRNA on an Illumina MiSeq platform to characterize the bacterial communities. To date, 316 filters have been sequenced, representing 52 water sources in 18 countries. Bacterial communities were analyzed in context with location, type of water source, climate, geographic factors, particulate load and dissolved metals. Factors that most significantly affect calculated distances between communities are country of origin and water source, though many significant factors are observed. Continued collection of water samples using consistent methods will begin to build a resource for understanding the relationships among microbial communities, geographical environment and potential chemical contaminants.