Novel method of probe design for characterising unclassified microbial taxa in wastewater
Tan, Shi Ming
Date of Issue2017
Interdisciplinary Graduate School (IGS)
The Singapore Centre for Environmental Life Sciences Engineering (SCELSE)
Floccular microbial communities inhabiting the activated sludge biosphere of wastewater treatment plants are widely exploited for their metabolic activities in removing anthropogenic carbon and inorganic nutrients. Although numerous studies have identified key microorganisms responsible for catalysing important bioprocesses, the activated sludge system is still regarded as a ‘black box’ because only a small proportion of the community has been taxonomically classified. ‘Black box’ of the sludge biosphere was demonstrated through deep sequencing to saturation of the floccular sludge community obtained from a municipal wastewater treatment plant in Singapore. The tag sequence of 37 out of the 50 most abundant OTUs could not be taxonomically classified, and none of the draft genomes obtained through genomic analysis were close to completion. Characterising the unclassified bacteria taxa has several implications. One apparent advantage is the broadening of our knowledge on the prokaryotic tree of life, which can subsequently lead to the design of more accurate PCR primers and FISH probes. This thesis describes the improvisation of a methodological approach that involves the targeted enrichment of unclassified bacterial taxa from floccular sludge community through FISH-FACS to obtain its “mini-metagenome” prior to downstream multiple displacement amplification (MDA), genomic sequencing and analyses. Even without access to a dedicated sterile environment and FACS machine for MDA-experiments, the FISH-FACS methodology outlined in this thesis provides a comprehensive guideline that manages DNA contamination and sterilisation of a communal FACS machine. Target cells were labelled with a new variant of fluorescence in situ hybridisation (FISH) probes which are termed as RiboProbes, and subsequently isolated via a highly-sensitive fluorescence-activated cell sorting (FACS) sorter. In contrast to canonical FISH probe design, RiboProbes were designed from short sequencing reads present in omics dataset (metagenomics and metatranscriptomics) that corresponded to highly variable regions (V4-V7) of the 16S rRNA. RiboProbes were affiliated with a high taxonomic resolution that allowed the differentiation of unclassified bacterial taxa down to the species level. However, truncation of the original length of the RiboProbe (33 bp) is necessary for combinatorial use with canonical FISH probes (average length: 17-25 bp) at a standardised hybridisation temperature. FISH-FACS was initially validated on a specific Thauera species in an axenic culture as a reference organism, followed by activated sludge samples where high level of enrichment of Thauera (>97%) was reproduced across many replicates. FISH-FACS with a canonical FISH probe produced lower level of enrichment (63%) for Thauera due to lower specificity of the probe, which subsequently led to the inclusion of non-target bacterial taxa in the sorted samples. FISH-FACS methodology was further extrapolated to two unclassified bacterial taxa in activated sludge, where target population was enriched through a collection of sorted samples that contained multiple events (5-1000). FISH-FACS resulted in one of the unclassified taxa (Candidatus Shimingles) being enriched to 99% and the other taxon (Haliangium clustero) being enriched to 44.14%. Level of enrichment for Haliangium clustero was lower than Ca. Shimingles due to co-sorting of closely-related Haliangium species into the sample. Due to the low mean read depth of sequencing (2.56X ± 2.97), no RiboTags could be assigned to Ca. Shimingles and Haliangium clustero in the pre-sorted samples of activated sludge. Thus far, FISH-FACS with RiboProbe is the only method that has been used for the enrichment and genome recovery of these two unclassified taxa. No other enrichment methods have been described for these novel taxa, except for culture-dependent methods used for the isolation of Haliangium species from costal and terrestrial regions. Taxonomic identification was made possible through phylogenetic assignment of full-length 16S rRNA gene sequences obtained from clone libraries. Through the 16S rRNA phylogenetic analysis, Ca. Shimingles was a novel bacterial genus categorised to an unclassified family of the order Sphingobacteriales, and Haliangium clustero was a novel bacterial species categorised to the genus Haliangium. Two distinct clades were observed in the 16S rRNA phylogenetic analysis of the Ca. Shimingles; distinct clades were in line with two draft genomes recovered through genomic binning. The two genomic bins were recovered with a size of 4.92 Mbp and 4.49 Mbp respectively, and a completeness of more than 90% and a contamination of less than 4%. The other draft genome recovered from Haliangium clustero had a size of 2.37 Mbp, with a completeness of 21.70% completeness and 0.02% contamination. A more complete genome of Haliangium clustero could not be obtained due to the complications of genome binning from strain heterogeneity. Even though the sorted samples were sequenced on the same sequencing flow cell as the pre-sorted samples with the same sequencing depth, genome coverage of both the novel taxa had vastly improved with the FISH-FACS enrichment procedure. The draft genomes exhibited low average amino acid identity and average nucleotide identity to the closest reference genomes in the database, therefore demonstrating novelty of these draft genomes. The novel Haliangium species formed aggregated structures that were distinct from other documented Haliangium species in activated sludge. Fruiting bodies of the novel Haliangium species were species-specific even in a mixed microbial environment. In addition, the formation of spherical myxospores was shown to be a characteristic of certain Haliangium species, and it was not strongly associated with the novel Haliangium species. The use of FISH probe to label myxospores and FACS to sort myxospores from the vegetative cells is unprecedented, and this method was first established in this thesis. Here, a reproducible and robust approach of overcoming the current limitations of conventional FISH probes design and metagenomics for the visualisation, identification and recovery of draft genomes from previously unclassified microbial taxa in wastewater is presented.