Metagenomic Analysis of Spring and Stream Waters in the Chesapeake and Ohio Canal National Historical Park

Abstract

In the current century, the most critical crises faced by human kind will likely be climate change, shortage of energy supplies, and pollution of the environment. A large variety of contaminants are susceptible to be released in the environment from households and from agricultural and industrial activities. During the last decades, physical, chemical, and biological technologies have been developed for pollution remediation and for assessing the extent of environmental contamination in water resources. Because of the large diversity of contaminants, the systematic and comprehensive analysis of elemental and compound pollutants cannot practically be conducted over an extensive network of water bodies. As a consequence, large-scale surface water monitoring programs frequently rely on biological assessment protocols based on macroinvertebrates, microalgae, or fishes, allowing to integrate the impact of many potential contaminants into single indices that are easy to interpret. However, standard bioassessment protocols are currently based on the morphological identification of representative sets of indicator organisms, which requires extensive stream sampling and laboratory observation in the laboratory and taxonomic identification. These operations are time- and personnel-consuming and require a great deal of experience. In this project, we have developed and validated an innovative water quality bioindicator based on the metagenomic analysis of the total prokaryotic microbial community in the water. Microorganisms are essential components of the aquatic ecosystem and their diversity, nature, and distribution typically reflect variations of the environmental conditions and water quality parameters. Although conventional, cultivation-based methods for microbial characterization are important in investigating the microbial communities, they are time and resources consuming. New polymerase chain reaction (PCR)-based molecular methods, such as metagenomic pyrosequencing, have the potential to quickly provide the detailed information on the microbial communities present in any environment. Advanced bioinformatics computing in connection with the resources of extensive genomic databases allow providing the detailed distribution of the microbial species present in the samples, which, in this project, was used as a fingerprint of water quality. The proposed research has been conducted using water samples collected from the Chesapeake and Ohio Canal National Historical Park (CHOH) in Maryland. Comprehensive characterization of the aquatic bacterial communities has been performed using metagenomic pyrosequencing. In parallel, a suite of relevant water quality parameters were monitored in the samples using standard methods. Using redundancy analyses (RDA), meaningful relationships were established between water characteristics and the metagenomic biomarker, showing its potential utilization as a general water quality indicator. This study provides the basis for the development of an innovative method for the fast and cost-effective assessment of water quality based on the aquatic prokaryotic microbiome. Phylogenetic analyses conducted on the metagenomic data revealed that the dominant prokaryotic phyla detected in the 19 samples are similar to the ones typically detected in freshwater environments. Microbial diversity indices showed that all 2012 samples were characterized by a low biodiversity, while 2013 samples were characterized by a higher diversity, which is likely the result of different meteorological conditions in 2012 and 2013. Clustering analysis and principal component analysis (PCA) were conducted to investigate the relationships between the relative abundance of the prokaryotic phyla and water quality parameters. The results showed that the samples collected from the same sites in different years cluster well together when compared based on the water quality parameters. On the contrary, the samples collected in 2012 made a separate group of cluster and same is true for 2013 samples when compared based on the prokaryotic phyla. These observations suggest a larger temporal variation of the microbial communities than the physico-chemical parameters of the water. PCA focusing on prokaryotic communities showed that Proteobacteria and Bacteroides phyla, including aerobic heterotrophic, fast growing bacteria – referred to as copiotrophic or 'r-type' organisms --, cluster together. On the other hand, the other phyla, including mostly anaerobic and/or autotrophic, slow growing bacteria – referred to as oligotropic or 'K-type' organisms --, form a rather distinct cluster. The dependence of the prokaryotic relative abundance on the water quality parameters for the 19 samples was then interrogated using RDA. As showed by PCA investigations, the r-type phyla cluster together and correlate with high alkalinity and conductivity. On the contrary, the K-type phyla cluster together and correlate collectively with sulfate and nitrate. As expected, the copiotrophic, fast-growing, r-type phyla also correlate with the stream samples, while the oligotrophic, slow-growing, K-type phyla correlate better with spring, cave, and mine samples. This study provides the basis for the development of an innovative method for the fast and cost-effective assessment of water quality based on the prokaryotic microbiome.