Reconstruction of high-quality genomes of bacteria in wastewater treatment reactors by hybrid sequencing strategy

Abstract

The thesis's primary research objective was to pierce into the microbial dark matter in wastewater treatment reactors by reconstructing high-quality genomes. I developed novel bioinformatics approaches to produce high-resolution profiles of the microbial composition and genome structure from studied wastewater treatment reactors by leveraging state-of-the-art sequencing technologies, i.e., the high-accuracy short-read and error-prone long-read sequencing. Genome-centric metagenomics has been a critical approach to expanding our understanding of the microbial diversity, ecology, evolution, and interactions within various ecosystems, particularly with cost-effective short-read sequencing and advanced assembly/binning tools recently. However, the metagenome-assembled genome (MAG) from the high-accuracy short reads is always highly fragmented, with hundreds of contigs/scaffolds, limiting our ability to obtain a full picture of the microbial structure. Starting with an enrichment ecosystem, partial-nitritation/anammox (PNA) bioreactor, I developed an iterative hybrid assembly (IHA) approach to reconstruct 49 metagenome-assembled genomes (N50 of 2.2 Mbp; 92.3% of the community), including 27 high-quality and high-contiguity genomes with contig number ≤ 5. Among them, eight were circular genomes. Notably, I also retrieved the first complete genome of the genus Candidatus Brocadia, providing full insight into the genome structure and function of this genus. Given the suitability of the IHA method in resolving complex metagenomic ecosystems, a novel Hierarchical Clustering Based Hybrid Assembly (HCBHA) approach to reconstruct (near-) complete genomes accurately from high-complexity environmental samples was developed. The key idea of the workflow is to phase reads to haplotypes and decompose the metagenomic assembly into assembly processes of many small microbial subsets. I applied the iterative HCBHA approach to assemble the complex activated sludge (AS) microbiome and reconstructed 557 MAGs (N50 of 574 Kbp; 61.1% of the community), including 125 high-quality MAGs and 14 closed genomes. The high-contiguity MAGs demonstrated the versatility of the AS microbiome. I applied both metagenomics and metatranscriptomics to investigate the metabolic interactions in the microbial community from an anammox upflow anaerobic sludge blanket (UASB) bioreactor. Based on the reconstructed 74 MAGs (N50 of 891 Kbp; 89.1% of the community), I profiled the expression patterns of the microbial members and associated pathways. Synergistic interaction were found in the ecosystem that the anammox bacterium supported some primary metabolites for the community and benefited from the additional nitrite via nitrate respiration of the side population. With the rapid development of the Nanopore sequencing, e.g., improving base-pair accuracy, I further discussed the possibility of using the long-read-only method integrated with short-read polishing to reconstruct genomes from the metagenomic datasets. Although for the complex metagenome, e.g., AS, this approach currently encountered difficulties, the continuous improvements of Nanopore sequencing, such as raw reads accuracy and sequencer throughput, promise an alternative practice soon. Overall, the advanced workflows developed in my thesis leveraged the short- and long-read sequencing while overcoming their limitations. The proposed methodologies significantly improved the capacity of (near-) complete genome reconstruction. Furthermore, this will highlight the value of the genome-centric approaches for resolving microbiomes of the uncultured microbial majority from diverse environmental niches, such as the microbiome from the ocean, human gut, and groundwater aquifers.