Influence of Wastewater Discharge on the Metabolic Potential of the Microbial Community in River Sediments Dong Li, Jonathan O. Sharp & Jörg E. Drewes
Microbial Ecology ISSN 0095-3628 Microb Ecol DOI 10.1007/s00248-015-0680-x
1 23
Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media New York. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy Microb Ecol DOI 10.1007/s00248-015-0680-x
ENVIRONMENTAL MICROBIOLOGY
Influence of Wastewater Discharge on the Metabolic Potential of the Microbial Community in River Sediments Dong Li 1,2 & Jonathan O. Sharp 1 & Jörg E. Drewes 1,2,3
Received: 19 February 2015 / Accepted: 15 September 2015 # Springer Science+Business Media New York 2015
Abstract To reveal the variation of microbial community functions during water filtration process in river sediments, which has been utilized widely in natural water treatment systems, this study investigates the influence of municipal wastewater discharge to streams on the phylotype and metabolic potential of the microbiome in upstream and particularly various depths of downstream river sediments. Cluster analyses based on both microbial phylogenetic and functional data collectively revealed that shallow upstream sediments grouped with those from deeper subsurface downstream regions. These sediment samples were distinct from those found in shallow downstream sediments. Functional genes associated with carbohydrate, xenobiotic, and certain amino acid metabolisms were overrepresented in upstream and deep downstream samples. In contrast, the more immediate contact with wastewater discharge in shallow downstream samples resulted in an increase in the relative abundance of genes associated with nitrogen, sulfur, purine and pyrimidine metabolisms, as well as restriction–modification systems. More diverse bacterial phyla were associated with upstream and deep
Electronic supplementary material The online version of this article (doi:10.1007/s00248-015-0680-x) contains supplementary material, which is available to authorized users. * Jörg E. Drewes
[email protected] 1
NSF Engineering Research Center ReNUWIt, Department of Civil and Environmental Engineering, Colorado School of Mines, Golden, CO, USA
2
Water Desalination and Reuse Center (WDRC), King Abdullah University of Science and Technology, Thuwal, Saudi Arabia
3
Chair of Urban Water Systems Engineering, Technical University of Munich, Am Coulombwall 8, 85748 Garching, Germany
downstream sediments, mainly including Actinobacteria, Planctomycetes, and Firmicutes. In contrast, in shallow downstream sediments, genera affiliated with Betaproteobacteria and Gammaproteobacteria were enriched with putative functions that included ammonia and sulfur oxidation, polyphosphate accumulation, and methylotrophic bacteria. Collectively, these results highlight the enhanced capabilities of microbial communities residing in deeper stream sediments for the transformation of water contaminants and thus provide a foundation for better design of natural water treatment systems to further improve the removal of contaminants. Keywords Metagenomics . Wastewater discharge . Sediment . Phylogenetics . Functional genes
Introduction Due to the limited availability of pristine freshwater water supplies in arid and semiarid regions, reuse of treated municipal wastewater has been proposed as a supplemental supply in many parts of the world. Natural water treatment systems including riverbank filtration, soil–aquifer treatment, and aquifer recharge and recovery, which all utilize the infiltration of wastewater effluents through biologically active porous media such as river sediments to further remove wastewaterderived constituents, have been used to augment groundwater supplies for decades and act as effective polishing zones for treated wastewater [9]. The discharge of treated municipal wastewater can result in an increase of nutrients, organic and inorganic constituents, and oxygen demand in receiving waters. This discharge can influence the phylogenetic composition of indigenous microbial communities in proximal freshwater sediments as demonstrated in prior research [11, 19, 23]. However, the impact that this disturbance might have on the
Author's personal copy D. Li et al.
metabolic potential of the microbiome residing in these impacted regions is still unclear. Furthermore, indigenous microorganisms in these sediments can play a vital role in the transformation and mineralization of inorganic and organic pollutants present in wastewater discharge such as nitrogen, organic carbon, and trace organic contaminants of emerging concern such as pharmaceuticals, personal care products, or household chemicals [1, 12, 21]. Microbial community phylotype, metabolic potential, and enzymatic expressions for a given local environmental condition have been linked to biotransformation of organic compounds in river sediments [5, 20]. Thus, understanding the impact of environmental conditions on both the taxonomic composition and metabolic capabilities of microbiomes in sediments might offer opportunities for an improved biotransformation of wastewater-derived organic compounds in wastewater-impacted sediments, which will also provide important insights about microbial metabolic capabilities in natural water treatment systems. Soil and sediment ecosystems are well recognized to harbor highly diverse microbial communities [22]. This composition is further complicated by soil–niche heterogeneities at the microscale [27]. High throughput sequencing has helped to reveal the microbial ecology of analogous systems using phylogenetic markers such as 16S or 18S ribosomal RNA (rRNA) genes [13]. This is particularly true for metagenomics of marine and human-related microbiomes, but much less is known about soil or sediment metagenomics [3, 8]. Recent work has demonstrated that a comparably small number of sequences (∼100,000 per sample) is sufficient to differentiate metagenomes across different soil sites and identify key metabolic categories that might shift with a change of environmental parameters [10]. It is thus possible to conduct a metagenomic inquiry with this readily achievable sequencing depth to characterize and compare the metabolic potential of microbiomes in soil and sediments while accepting limitations associated with the resolution of rarer functional genes or the full assembly of genomes belonging to abundant microbial populations. With this in mind, the primary aim of the work described herein was to elucidate and compare potential shifts in the metabolic potential of microbiomes in river sediments upstream and downstream of wastewater discharge. We selected sediments from the Taif River in Saudi Arabia where a prior study revealed the microbial taxonomic composition in different locations and depths of river sediments [19]. The river receives secondary treated municipal wastewater from a local facility, and the water qualities of both upstream and downstream river locations were seasonally consistent. The results of 16S rRNA gene pyrosequencing in the prior study demonstrated similar microbial communities derived from analogous sampling zones in geographically distanced rivers and a remarkable stable microbial community composition at all sampling sites across seasons. Furthermore, dissolved organic carbon (DOC) exerted a strong influence on microbial community composition and diversity in
water infiltration systems. The metagenomic results of this study will deepen our understanding of how the metabolic potential of sediment ecosystems shift as a function of changing water quality and therefore substrate compositions and provide insights into regions where the biotransformation of more refractory trace organic chemicals might be more favorable. The results should be helpful for better design of natural water treatment systems to enhance the removal of trace organic contaminants of emerging concern during the water infiltration process.
Materials and Methods Sampling of River Sediments River sediment samples were collected in February, April, and June 2011 from four cross sections (CS) of the Taif River located in Taif, Saudi Arabia, as described previously [19]. Briefly, two cross sections (CS1 and CS2) are located upstream of a secondary treated municipal wastewater discharge point and two cross sections (CS3 and CS4) are downstream of the discharge (Figure SI-1, Supplemental Information). The design capacity of the wastewater treatment plant was 19,700 m 3 /day. Treatment processes at the plant included solid screening, biological nutrient removal (nitrifying/denitrifying), and secondary clarification followed by chlorination. For each cross section, river sediment samples were collected in 5 cm depth below the riverbed (named as CS1-5, CS2-5, CS3-5, and CS4-5) by using thin-wall tube auger. An additional unsaturated sample was collected at 5 cm depth outside the streambed of cross section CS1 (named CS1-C). Furthermore, to better understand transformation processes of wastewater constituents in deeper zones of impacted sediments, one sample was collected from 50 cm depth at CS3 (CS3-50) and two sediment samples were taken from 25 and 50 cm depths at CS4 (CS4-25 and CS4-50), representing the zone below the infiltration layer. Approximately 100 g sediment was obtained for each sample in addition to 4 L river water samples for each cross section. Samples were stored at 4 °C (water) and −20 °C (sediment) pending respective chemical and microbial analyses. Bulk water quality parameters including DOC, UVabsorbance (254 nm), SUVA (specific UVabsorbance, which is the ratio of UV absorbance to DOC), ammonia, nitrate, and other physicochemical parameters as well as sediment properties including porosity, hydraulic conductivity, and particle size distribution were analyzed according to standard methods [2]. Sediment pH was measured at a 1:2.5 ratio of solids to DI water. DNA Extraction, Sequencing, and Annotation The DNA extraction procedure was similar to those described previously [19]. Briefly, for each sediment sample, replicate DNA extractions were performed using the PowerSoil kit (MO BIO laboratories, Carlsbad, CA). The extracted DNA
Author's personal copy Influence of Wastewater Discharge on Sediment Microbes
was quantified using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE) and the DNA samples taken from the same site during three sampling campaigns were further pooled together. Totally, 5–6 μg DNA was obtained for metagenomic analysis of each sampling site. DNA was nebulized and tagged using the GS-FLX-Titanium Rapid Library MID Adapters Kit (454 Life Sciences, Branford, CT, USA) before submitting to the Genomics Core Laboratory at KAUST for pyrosequencing on a 454 FLX Titanium sequencer (Roche). The run yielded 1.18 Gb of data, 3.12 million reads in total, with an average read length of 378 bp. Metagenomic sequences were demultiplexed and quality filtered using MG-RAST Version 3.2 with default settings [24] (submission ID: 4491385-4491392), and the obtained sequences were further submitted to the US Department of Energy Joint Genome Institute IMG/M system (http://img. jgi.doe.gov/cgi-bin/m/main.cgi) (submission ID: 37113, and 37115-37121) for gene calling and functional annotation mainly with Enzyme Commission (EC) numbers and Clusters of Orthologous Groups (COG) database. Resultant eight metagenomic libraries contained 0.25 to 0.56 M reads each with GC content ranging from 60 to 67 % (more details in Table S2). Metagenomic Differences Between Samples The relative abundance of EC and COG gene in each library were transformed (Log(x+1)) for normalization, and subsequently, a similarity matrix for all samples was calculated using Bray–Curtis distance. Principal coordinate analysis (PCA) and hierarchical clustering based on similarity matrix of all eight samples were performed using PRIMER 6 [6]. SIMPROF test was conducted to validate the clustering of eight samples with 999 simulation permutations. Furthermore, the clustering of metagenomic libraries was also validated using the R package clValid [4]. UPGMA, K-means, and PAM clustering algorithms were tested for clustering the samples. Optimal number of clusters was estimated using the Calinski–Harabasz (CH) index and further validated using silhouette width measurement and stability measurements including average proportion of non-overlap (APN) and the average distance between means (ADM). The identification of individual genes varying significantly between two clustering groups of samples was conducted using the R package ShotgunFunctionalizeR [15] based on the Poisson model and the algorithm LEfSe (the linear discriminant analysis effect size) using non-parametric factorial Kruskal–Wallis test followed by linear discriminant analysis [26]. P values in the package ShotgunFunctionalizeR were corrected for multiple tests using the Benjamini–Hochberg correction factor. Here, default parameter settings of the algorithm LEfSe (alpha parameter of 0.05 for pairwise tests set and the threshold on the logarithmic LDA score as 2.0) were used.
The heatmap of representative EC genes showing the largest differences between groups of samples as determined by ShotgunFunctionalizeR and LEfSe analyses was visualized and clustered using the function heatplot of the R package Made4 [7]. Phylogenetic Assignment and Analyses Metagenomic sequences were also phylogenetically assigned using the best hit classification in MG-RAST with GenBank as the annotation source and the following parameter settings: maximum e-value cutoff of 1E−05, minimum identity cutoff of 60 %, and minimum alignment length cutoff of 150 bp. PCA and hierarchical clustering based on similarity matrix of microbial species abundance in all eight samples were also performed using PRIMER 6. Microbial taxonomic groups that are significantly associated to each group of samples were determined using the algorithm LEfSe. The undirected, weighted microbial species networks were constructed by calculating the pairwise Pearson correlations between all species data across all samples using the R package WGCNA [17]. The soft thresholding power β based on the criterion of approximate scale-free topology was chosen to which co-occurrence similarity is raised to calculate adjacency. The adjacency was transformed into topological overlap matrix and the corresponding dissimilarity was further calculated. Species are organized into modules using the topological overlap measurement in a hierarchical cluster analysis. The data network was further exported to Cytoscape for fundamental statistics calculation [25]. The global descriptors of the modules including network density, network centralities, clustering coefficient, etc. were calculated. Additional statistical analyses including ANOSIM, etc. were performed using the SPSS package (version 16.0) or PAST [14]. P values of less than 0.05 were considered significant.
Results Influence of Wastewater Discharge As expected, wastewater discharge to the Taif River resulted in increased DOC, ammonia, nitrite, chloride, and phosphate concentrations observed at downstream locations (CS3 and CS4) compared to samples collected upstream (CS1 and CS2) of the receiving stream (Table SI-1). In addition to higher concentrations associated with wastewater discharge, SUVA values revealed that organic matter present downstream of the discharge had a lower fraction of aromatic organic moieties and presumably more readily biodegradable carbon than at the upstream location. Furthermore, no significant differences were observed regarding sediment properties including
Author's personal copy D. Li et al.
porosity, hydraulic conductivity, and particle size distribution among all sediment samples collected in this study (Table SI1, Supplemental Information). Microbiome Metabolic Potential in River Sediments A sediment microbiome, representative of functional gene profiles and presumably one that could correlate to metabolic potential for microbial degradative processes, was analyzed to determine if correlations could be derived to alterations in water properties or substrate conditions. Totally, eight resultant sediment samples from these transects were analyzed by PCA and hierarchical clustering of metagenome functional genes using Bray–Curtis similarity matrices. This analysis revealed two distinct groups where bacterial genes derived from shallow sediment samples downstream of wastewater discharge (CS3-5, CS4-5, and CS4-25 as group 2) were distinct from upstream sediments (CS1-5 and CS2-5), deeper downstream sediments (CS3-50 and CS4-50), and the unsaturated sediment sample CS1-C (as group 1) (Fig. 1a). This grouping followed a pattern of increasing concentrations of DOC, ammonia, and nitrite after discharge and was further validated using a SIMPROF test in PRIMER 6 in addition to the CH index, silhouette width, APN, and ADM measurements using the R package clValid (data not shown). ANOSIM analysis indicated that the difference between these two groups of samples was significant (P=0.018). We identified 247 KEGG Enzyme Commission (EC) numbers whose proportional representation differed significantly between these two groups using the R package ShotgunFunctionalizeR; 132 ECs were significantly more abundant in group 1 (P
