Bacterial Community Structure in the Drinking Water Microbiome Is Governed by Filtration Processes

Department of Civil and Environmental Engineering, University of Michigan, USA.
Environmental Science & Technology (Impact Factor: 5.33). 07/2012; 46(16):8851-9. DOI: 10.1021/es302042t
Source: PubMed


The bacterial community structure of a drinking water microbiome was characterized over three seasons using 16S rRNA gene based pyrosequencing of samples obtained from source water (a mix of a groundwater and a surface water), different points in a drinking water plant operated to treat this source water, and in the associated drinking water distribution system. Even though the source water was shown to seed the drinking water microbiome, treatment process operations limit the source water's influence on the distribution system bacterial community. Rather, in this plant, filtration by dual media rapid sand filters played a primary role in shaping the distribution system bacterial community over seasonal time scales as the filters harbored a stable bacterial community that seeded the water treatment processes past filtration. Bacterial taxa that colonized the filter and sloughed off in the filter effluent were able to persist in the distribution system despite disinfection of finished water by chloramination and filter backwashing with chloraminated backwash water. Thus, filter colonization presents a possible ecological survival strategy for bacterial communities in drinking water systems, which presents an opportunity to control the drinking water microbiome by manipulating the filter microbial community. Grouping bacterial taxa based on their association with the filter helped to elucidate relationships between the abundance of bacterial groups and water quality parameters and showed that pH was the strongest regulator of the bacterial community in the sampled drinking water system.

Download full-text


Available from: Lutgarde Raskin, Jun 24, 2014
1 Follower
165 Reads
  • Source
    • "Previous studies have suggested that chlorinated drinking water systems are less 111 species-rich than non-chlorinated systems (Hoefel et al., 2005; Pinto et al., 2012; 112 Lautenschlager et al., 2013). "
    [Show abstract] [Hide abstract]
    ABSTRACT: In this study we collected water from different locations in 32 drinking water distribution networks in the Netherlands and analysed spatial and temporal variation in microbial community composition by high-throughput sequencing of 16S rRNA gene amplicons. We observed that microbial community compositions of raw source and processed water were very different for each distribution network sampled. In each network major differences in community compositions were observed between raw and processed water, although community structures of processed water did not differ substantially from end-point tap water. End-point water samples within the same distribution network revealed very similar community structures. Network specific communities were shown to be surprisingly stable in time. Biofilm communities sampled from domestic water meters varied distinctly between households and showed no resemblance to planktonic communities within the same distribution networks. Our findings demonstrate that high-throughput sequencing provides a powerful and sensitive tool to probe microbial community composition in drinking water distribution systems. Furthermore, this approach can be used to quantitatively compare the microbial communities to match end-point water samples to specific distribution networks. Insight in the ecology of drinking water distribution systems will facilitate the development of effective control strategies that will ensure safe and high-quality drinking water. This article is protected by copyright. All rights reserved.
    Environmental Microbiology 01/2015; 17(7). DOI:10.1111/1462-2920.12739 · 6.20 Impact Factor
  • Source
    • "It does not require labelled primers/nucleotides or gel electrophoresis and allows a large number of samples to be pooled (Ronaghi, 2001; Fakruddin and Chowdhury, 2012). This technique has recently been applied for the identification of species present in water during treatment (Wakelin et al., 2011; Pinto et al., 2012) and distribution (Henne et al., 2008; Hong et al., 2010; Hwang et al., 2012; Lin et al., 2013; Liu et al., 2013a; Lautenschlager et al., 2013). The studies using pyrosequencing have proved the value of identifying bacterial groups, for the evaluation of e.g. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The combination of flow cytometry (FCM) and 16S rRNA gene pyrosequencing data was investigated for the purpose of monitoring and characterizing microbial changes in drinking water distribution systems. High frequency sampling (5 min intervals for 1 h) was performed at the outlet of a treatment plant and at one location in the full-scale distri-bution network. In total, 52 bulk water samples were analysed with FCM, pyrosequencing and conventional methods (adenosine-triphosphate, ATP; heterotrophic plate count, HPC). FCM and pyrosequencing results individually showed that changes in the microbial com-munity occurred in the water distribution system, which was not detected with conven-tional monitoring. FCM data showed an increase in the total bacterial cell concentrations (from 345 ± 15 Â 10 3 to 425 ± 35 Â 10 3 cells mL À1) and in the percentage of intact bacterial cells (from 39 ± 3.5% to 53 ± 4.4%) during water distribution. This shift was also observed in the FCM fluorescence fingerprints, which are characteristic of each water sample. A similar shift was detected in the microbial community composition as characterized with pyro-sequencing, showing that FCM and genetic fingerprints are congruent. FCM and pyrose-quencing data were subsequently combined for the calculation of cell concentration changes for each bacterial phylum. The results revealed an increase in cell concentrations of specific bacterial phyla (e.g., Proteobacteria), along with a decrease in other phyla (e.g., Actinobacteria), which could not be concluded from the two methods individually. The combination of FCM and pyrosequencing methods is a promising approach for future drinking water quality monitoring and for advanced studies on drinking water distribution pipeline ecology.
    Water Research 10/2014; 63:179-189. DOI:10.1016/j.watres.2014.06.020 · 5.53 Impact Factor
  • Source
    • "Within Proteobacteria, α-Proteobacteria was the most dominant class in each water sample, followed by β-, γ-and δsubdivisions (Fig. 1), which is consistent with previous study (Kwon et al., 2011) revealing that α-subdivision was the more abundant than other classes of Proteobacteria in drinking water based on pyrosequencing. However, Pinto et al. (2012) showed that β-Proteobacteria (40%) was more abundant than "
    [Show abstract] [Hide abstract]
    ABSTRACT: In order to comprehensively investigate bacterial virulence in drinking water, 454 pyrosequencing and Illumina high-throughput sequencing were used to detect potential pathogenic bacteria and virulence factors (VFs) in a full-scale drinking water treatment and distribution system. 16S rRNA gene pyrosequencing revealed high bacterial diversity in the drinking water (441-586 operational taxonomic units). Bacterial diversity decreased after chlorine disinfection, but increased after pipeline distribution. α-Proteobacteria was the most dominant taxonomic class. Alignment against the established pathogen database showed that several types of putative pathogens were present in the drinking water and Pseudomonas aeruginosa had the highest abundance (over 11‰ of total sequencing reads). Many pathogens disappeared after chlorine disinfection, but P. aeruginosa and Leptospira interrogans were still detected in the tap water. High-throughput sequencing revealed prevalence of various pathogenicity islands and virulence proteins in the drinking water, and translocases, transposons, Clp proteases and flagellar motor switch proteins were the predominant VFs. Both diversity and abundance of the detectable VFs increased after the chlorination, and decreased after the pipeline distribution. This study indicates that joint use of 454 pyrosequencing and Illumina sequencing can comprehensively characterize environmental pathogenesis, and several types of putative pathogens and various VFs are prevalent in drinking water.
    Ecotoxicology and Environmental Safety 08/2014; 109C:15-21. DOI:10.1016/j.ecoenv.2014.07.029 · 2.76 Impact Factor
Show more