Occurrence and Fate of Ultramicrobacteria in a Full-Scale Drinking Water Treatment Plant

Jie Liu, Renxin Zhao, Jiayu Zhang, Guijuan Zhang, Ke Yu, Xiaoyan Li, Bing LiFrontiers in Microbiology 2018 · DOI

TL;DR

This study investigates the prevalence and behavior of ultramicrobacteria (UMB) in a full-scale drinking water treatment plant (DWTP). Researchers found that while overall bacterial numbers decrease, UMB become proportionally more abundant, suggesting they are more resilient to treatment processes. This resilience is linked to their unique genetic makeup and specialized metabolic functions.

What Problem Does This Paper Address?

Drinking water treatment plants (DWTPs) are designed to remove a wide range of microorganisms to ensure water safety. While the removal of larger bacteria (LB) is generally well-understood, the fate and characteristics of ultramicrobacteria (UMB) – bacteria with cell volumes less than 0.1 µm³ – within these systems remain poorly understood. UMB are known to be numerically dominant in freshwater environments, and their persistence or removal behavior in DWTPs is crucial for a complete picture of microbial dynamics and water quality. This paper aims to characterize the occurrence, community structure, and functional potential of UMB compared to LB throughout a DWTP.

Methods

The researchers employed a multi-pronged approach:

  • Flow Cytometry (FCM): Used to quantify bacterial cell density and differentiate between UMB and LB based on their size and nucleic acid content.
  • Size-Fractionation Filtration: A two-step filtration process using 0.22-µm and 0.10-µm pore-size filters was used to physically separate LB (captured on the 0.22-µm filter) from UMB (passing the 0.22-µm filter and captured on the 0.10-µm filter).
  • 16S rRNA Gene High-Throughput Sequencing: DNA extracted from the separated LB and UMB fractions was sequenced to determine their community composition and phylogenetic relationships.
  • Bioinformatic Analysis: This included:
  • LEfSe (Linear Discriminant Analysis Effect Size): To identify specific indicator taxa for LB and UMB.
  • NMDS (Non-metric Multidimensional Scaling) and Adonis analysis: To visualize and quantify community structure differences and the influence of various factors (e.g., filtration size, treatment stage).
  • Co-occurrence Network Analysis: To explore the complex interrelationships between different bacterial taxa.
  • Tax4fun package: To predict the functional and metabolic capabilities of the microbial communities based on 16S rRNA gene data.

Datasets

The study utilized water samples collected from a full-scale drinking water treatment plant in Guangdong province, China. Samples were taken at various stages of the treatment process: raw water (RW), flocculation-sedimentation effluent (FS), rapid sand filtration effluent (RSF), biological granular activated carbon filtration effluent (BAC), and chlorination water (CW). Multiple sampling events (DWTP1, DWTP2, DWTP3) were conducted to ensure robustness.

Key Insights and Ideas

  • UMB's Resilience: UMB are not simply removed along with larger bacteria; their proportion of the total bacterial community actually increases as water progresses through the DWTP, especially after biological activated carbon (BAC) treatment. This suggests UMB possess inherent resistance mechanisms.
  • Distinct Communities: UMB and LB communities are not just different in size but also in their phylogenetic makeup and predicted functions. UMB communities are less diverse but exhibit unique taxonomic groups.
  • Evolutionary Conservation of Size: The exclusive clustering of certain UMB phyla (e.g., Parcubacteria, Elusimicrobia, Saccharibacteria) at high taxonomic levels strongly suggests that their ultra-small size is an evolutionarily conserved trait, not just a temporary state.
  • Streamlined Genomes and Symbiosis: The "streamlined genome" characteristic of many UMB, particularly candidate phyla radiation (CPR) bacteria, is linked to a symbiotic or parasitic lifestyle. This may explain the observed non-random co-occurrence patterns within UMB communities, indicating strong interdependencies.
  • Functional Advantages: Predicted functional analysis points to UMB having enhanced capabilities in "environmental information processing" and "DNA replication and repair." These functions likely contribute to their higher resistance to stresses encountered during water treatment, such as disinfection.

Key Results

  • Cell Density: Both LB and UMB cell densities generally decreased along the treatment process, with the exception of the BAC unit where diversity and density increased.
  • UMB Proportion: The proportion of UMB in the total bacterial community significantly increased from 0.05-9.05% in raw water to 1.52-37.4% in treated water.
  • Diversity: UMB communities were less diverse than LB communities across most treatment stages.
  • Indicator Taxa: Specific indicator taxa were identified for both LB and UMB. UMB indicator taxa were predominantly affiliated with α/β/γ-Proteobacteria, Deinococcus-Thermus, Firmicutes, Acidobacteria, and Dependentiae. Notably, phyla like Parcubacteria, Elusimicrobia, and Saccharibacteria were exclusively found as UMB indicators.
  • Functional Gene Abundance: Predicted enzyme-encoding genes related to energy metabolism, environmental information processing, environmental stress resistance, and DNA replication/repair were found at higher relative abundances in UMB compared to LB.

Limitations and Future Work

  • Filtration Imperfections: The authors acknowledge that size-based filtration is not perfect, and some LB might pass through the 0.22-µm filter, while some UMB might be captured on it due to matrix effects. This could lead to some overlap in the separated fractions.
  • BAC Maintenance: The BAC unit was undergoing maintenance during one sampling event, limiting the data for that specific stage.
  • Future Work: The study provides a strong foundation for understanding UMB in DWTPs. Future research could focus on direct experimental validation of the predicted functional roles of UMB, particularly their stress resistance mechanisms. Further investigation into the symbiotic relationships and the implications of their streamlined genomes for water treatment efficiency would also be valuable. The findings offer a basis for potentially manipulating drinking water microbiomes for improved water quality.

Summary updated

Abstract

Ultramicrobacteria (UMB) are omnipresent and numerically dominate in freshwater, as microbes can present in drinking water systems, however, the UMB communities that occur and their removal behaviors remain poorly characterized in drinking water treatment plants (DWTPs). To gain insights into these issues, we profiled bacterial cell density, community structure and functions of UMB and their counterpart large bacteria (LB) using flow cytometry and filtration paired with 16S rRNA gene high-throughput sequencing in a full-scale DWTP. Contrary to the reduction of bacterial density and diversity, the proportion of UMB in the total bacteria community increased as the drinking water treatment process progressed, and biological activated carbon facilitated bacterial growth. Moreover, UMB were less diverse than LB, and their community structure and predicted functions were significantly different. In the DWTP, UMB indicator taxa were mainly affiliated with α/β/γ-<i>Proteobacteria, Deinococcus-Thermus, Firmicutes, Acidobacteria</i>, and <i>Dependentiae</i>. In particular, the exclusive clustering of UMB at the phylum level, e.g., <i>Parcubacteria, Elusimicrobia</i>, and <i>Saccharibacteria</i>, confirmed the fact that the ultra-small size of UMB is a naturally and evolutionarily conserved trait. Additionally, the streamlined genome could be connected to UMB, such as candidate phyla radiation (CPR) bacteria, following a symbiotic or parasitic lifestyle, which then leads to the observed high connectedness, i.e., non-random intra-taxa co-occurrence patterns within UMB. Functional prediction analysis revealed that environmental information processing and DNA replication and repair likely contribute to the higher resistance of UMB to drinking water treatment processes in comparison to LB. Overall, the study provides valuable insights into the occurrence and fate of UMB regarding community structure, phylogenetic characteristics and potential functions in a full-scale DWTP, and it is a useful reference for beneficial manipulation of the drinking water microbiome.
BibTeX
@inproceedings{liu2018occurrence,
  author = {Jie Liu and Renxin Zhao and Jiayu Zhang and Guijuan Zhang and Ke Yu and Xiaoyan Li and Bing Li},
  title = {{Occurrence and Fate of Ultramicrobacteria in a Full-Scale Drinking Water Treatment Plant}},
  booktitle = {Frontiers in Microbiology},
  year = {2018},
  doi = {10.3389/fmicb.2018.02922},
}
Figure