|M.Sc Student||Berenbaum Dina|
|Subject||CRISPR Evolution within Natural Microbial Communities|
|Department||Department of Energy||Supervisor||Professor Roy Kishony|
|Full Thesis text|
The CRISPR-Cas system, an adaptive immune system providing bacteria defense against phages, could be of great value for protecting microbial communities, in natural systems and in synthetic bio-energy applications.
While CRISPR-based adaptation in the form of spacer acquisitions was observed in the lab, the nature of evolution within natural communities still remains unknown. Specifically, the extent to which additional evolutionary pathways, such as spacer deletion or spacer duplication, influence selection in bacterial populations remains unclear. Moreover, the rate at which CRISPRs evolve in the natural environment is still unknown. To understand the unique features of natural CRISPR evolution, we employed two strategies: computational
analysis of a publicly available database of CRISPRs and an experimental methodology that aims to produce reliable data of the entire repertoire of CRISPRs directly from a human gut microbial community. The experimental method was itself divided into two approaches --a high-throughput, low-quality method and a high-quality, low-throughput method - and collected CRISPR sequencing data directly from human stool samples of multiple donors and multiple time points.
Overall, both strategies show that spacer deletion and spacer duplication are common evolutionary events that occur within microbial populations. Looking at the computational study, the ubiquitous nature of spacer deletion and spacer duplication was highly apparent, present across all the phylogeny. The functional groups of spacers (phage, plasmid and self) are equally represented in the deleted and duplicated pool and in the general pool of spacers, suggesting perhaps a neutral evolution. Interestingly, in the experimental approach, the results for Bifidobacteria, derived from both methods, show evolution occurring in real time within human hosts leading to the diversity of CRISPRs in a single sample. This diversity is caused primarily via deletions, which were observed in both the high-throughput and the low-throughput methods. However, the number of deletions was significantly higher in the high-throughput method, possibly suggesting artificial deletion due to amplification with PCR. We end with proposing the BifidoFish method. This revised method aims to produce data for characterizing CRISPR adaptation in natural microbial communities, both in high-throughput and with high-quality.
In total these results highlight a rapid diversification of CRISPRs across a range of time scales via spacer deletion and duplication. The evolution occurs in real time within a single host. The suggested lack of selection for the process could be implying on an even higher frequency of the ubiquitous events.