Improving Single Cell ‘Omics Methods for Investigating Microbial Dark Matter

  • chair:

    Sobol, M. S. / Kaster, A.- K. /Schwartz, T. (2023)

  • place:


  • Date: August 2023
  • Abstract (englisch):

    The vast majority of microbial life still remains undiscovered and understudied. We refer to these microorganisms as microbial dark matter (MDM) because they have not yet been successfully cultured. Within MDM hide potentially novel and important solutions for sustainable energy, bioremediation of contaminated environments, and the war against rising antibiotic resistance. The use of culture-independent methods to study microorganisms at the community-level, such as metagenomics and metatranscriptomics, have significantly advanced our understanding of MDM. However, these methods still struggle to reliably assemble individual genomes and transcriptomes, especially from low abundant organisms in highly diverse communities. Strain variation, the misattribution of sequences, highly repetitive sequence regions, and mobile genetic elements are a few of the problems that metagenomics faces. Likewise, in metatranscriptomics, the natural phenotypic heterogeneity of cells and diversity of microbial communities, results in complex transcriptional profiles that cannot be fully captured. Therefore, single-cell genomics (SCG) and transcriptomics (SCT), which together are referred to as single-cell ‘omics (SC ‘omics), were developed to overcome the disadvantages of metagenomics and metatranscriptomics by enabling the analysis of an individual cell.
    The application of SCG has become an important tool for expanding our knowledge of MDM, for example, by enabling the recent discovery of several novel candidate phyla, which are currently only represented by single-amplified genomes (SAGs). However, complete SAGs from many organisms, especially minority members, are statistically hard to capture due to the high costs and many technical challenges throughout the workflows. On the other hand, microbial SCT is faced with the many challenges of working with RNA, such as the short half-life of mRNA and low levels of gene expressions, which is why SCT has not yet been widely applied in microbiology. Thus, the anticipated effects of SC ‘omics have not yet been fulfilled.
    In a typical SCG workflow, after samples are collected, cells can be labeled with fluorescent dyes prior to single-cell isolation. After isolation, the cells are lysed and the genome has to be amplified for subsequent library preparation, which is followed by sequencing and data analysis. In this thesis, difficulties in the cell labelling, isolation, lysis, and whole genome amplification (WGA) steps were improved upon to overcome remaining challenges in SCG. First, a targeted-cell labeling approach was established, which enabled the enrichment of low abundant microorganisms from environmental samples. This approach aided in the discovery of novel phylogenies and metabolisms from rare members of the microbial community, which would have otherwise been overlooked by conventional metagenomics. Additionally, by targeting organisms of interest, this approach helped to reduce the costs of SCG by preventing the need to sequence tens of thousands of single-cells in order to access low abundant minority members. Next, improvements were made to the cell isolation and cell lysis steps to minimize both physical cell damage and DNA degradation, respectively, which helped to increase the success of the downstream genome amplification step. As for the WGA, a volume reduction approach was applied to significantly improve genome coverage uniformity and completeness. These findings highlighted the unnecessary need for further volume reduction down to nL or pL and costs could be reduced by 97.5%. It is anticipated that these advancements will increase the throughput of SCG and encourage the use of this approach in more research groups.
    Since SCG alone only provides information on phylogeny, genetic structure and metabolic potentials, but not on the actual activity of a cell, a microbial SCT pipeline was developed in this thesis, to help understand the cell’s individual functions in a community. Currently, very few methods for microbial SCT exist and the ones that do, remain widely unused outside of their respective groups due to their difficult handling and low accessibility. Therefore, in this study, modifications and improvements to a eukaryotic single-cell RNA sequencing (RNA-seq) method were made to enable its use in prokaryotes. Importantly, the addition of DTT in the lysis buffer was found to likely inhibit DNase I, leading to DNA contamination. The single-cell RNA-seq results herein revealed reliable transcriptional profiles when compared to bulk RNA-seq samples. This was also confirmed through a proof of principle experiment comparing heat-shock and non-treated Escherichia coli cells. Furthermore, evidence for unique responses involved in secondary metabolite synthesis and CRISPR-Cas editing were found upregulated in the single cell versus the bulk data, highlighting the importance for studying heterogeneity of functional subpopulations at the single-cell level. Overall, the improved SCG and SCT methods established in this work are anticipated to allow for more widespread use for further understanding of MDM diversity and function in the environment.