CrAssphage May Be Viable Markers of Contamination in Pristine and Contaminated River Water

Abstract

Viruses are the most biologically abundant entities and may be ideal indicators of fecal pollutants in water. Anthropogenic activities have triggered drastic ecosystem changes in rivers, leading to substantial shifts in chemical and biological attributes. Here, we evaluate the viability of using the presence of crAssphage as indicators of fecal contamination in South African rivers. Shotgun analysis revealed diverse crAssphage viruses in these rivers, which are impacted by chemical and biological pollution. Overall, the diversity and relative abundances of these viruses was higher in contaminated sites compared to pristine locations. In contrast to fecal coliform counts, crAssphage sequences were detected in pristine rivers, supporting the assertion that the afore mentioned marker may be a more accurate indicator of fecal contamination. Our data demonstrate the presence of diverse putative hosts which includes members of the phyla Bacteroidota, Pseudomonadota, Verrucomicrobiota, and Bacillota. Phylogenetic analysis revealed novel subfamilies, suggesting that rivers potentially harbor distinct and uncharacterized clades of crAssphage. These data provide the first insights regarding the diversity, distribution, and functional roles of crAssphage in rivers. Taken together, the results support the potential application of crAssphage as viable markers for water quality monitoring. IMPORTANCE Rivers support substantial populations and provide important ecosystem services. Despite the application of fecal coliform tests and other markers, we lack rapid and reproducible approaches for determining fecal contamination in rivers. Waterborne viral outbreaks have been reported even after fecal indicator bacteria (FIB) were suggested to be absent or below regulated levels of coliforms. This indicates a need to develop and apply improved indicators of pollutants in aquatic ecosystems. Here, we evaluate the viability of crAssphage as indicators of fecal contamination in two South African rivers. We assess the abundance, distribution, and diversity of these viruses in sites that had been predicted pristine or contaminated by FIB analysis. We show that crAssphage are ideal and sensitive markers for fecal contamination and describe novel clades of crAss-like phages. Known crAss-like subfamilies were unrepresented in our data, suggesting that the diversity of these viruses may reflect geographic locality and dependence.

Keywords: bacteria; bacteriophages; crAssphage; faecal pollution; metagenome assembled genomes; phylogeny; viruses.

FIG 1

Map of sampling sites and a phylogenetic tree of Bacteroidota. (A) The six sampling sites in the Limpopo, Province of South Africa. The letters designate specific sampling locations as with L and B for Letsitele and Thabina (known locally as the Bathabina) River, respectively. Specific sites include the following: L1, upstream of the settlements and irrigation; L2, midstream; L3, downstream of the settlements; B1, upstream of the settlements; B2, midstream; B3, downstream of the settlements. (B) Bacteroidota maximum-likelihood phylogenetic tree. The metagenome assembled genomes (MAGs) obtained from this study are indicated in green. Those retrieved from the NCBI database are shown in red. The tree illustrates separate clustering between the MAGs obtained in this study and those from the NCBI. The genomes from this study form a separate and distinct cluster. Bootstrap values were calculated to support the robustness of the different clades and are indicated by the insert. The Limpopo map was sourced from dmaps (https://d-maps.com/pays.php?num_pay=1641&lang=en), and the inset was sourced from Google Maps.

FIG 2

The relative abundances of crAssphage viruses. The relative abundance was calculated by read mapping sequences from the different sampling locations. The six sites (L1, L2, L3, B1, B2, B3) are shown in the x axis. The labels B and L correspond to Thabina and Letsitele river sampling sites, respectively. The cluster analysis was based on Euclidean distances. The putative crAssphage sequences, obtained from this study, are shown on the right y axis. The dark blue color indicates high crAssphage abundances, in a specific site, while white shows the absence of crAssphage per location. The heatmap shows generally higher diversity and distribution of crAssphage observed in contaminated sites compared with pristine sampling locations.

FIG 3

Phylogenetic tree of crAssphage sequences. The tree was constructed using TerL protein sequences from crAssphage. CrAssphage, obtained from this study, are colored black. CrAssphage sequences obtained from a recent study (28) representing Zeta, Epsilon, Gamma, Alpha, Beta, and Delta clusters are shown in blue, green, purple, brown, red, and green, respectively. CrAssphage obtained in this study clustered with the Zeta, Epsilon, and Delta subfamilies. In general, crAssphage retrieved from this study clustered separately from those obtained from the Yutin et al. (28) study. This suggests that they may potentially represent novel, and as yet, uncharacterized crAssphage sequences.

FIG 4

Genomic structure comparison of near complete Epsilon crAssphage. The crAssphage sequences from this study were compared with the prototypical crAssphage (p-crAssphage), and two other gut associated crAssphage. Regions of amino acid sequence homology are shown. Blue indicates 100% homology and red indicates the lowest similarities. The RNAP subunit and the capsid gene module were found in all the genomes analyzed.