Rachel Case
Research Mentor: Brian Athey
Mentor Department: Computational Medicine & Bioinformatics, Medicine
Author(s): Not Available
Session: Session 4 (1:00 PM – 1:50 PM)
Presentation Type: Oral Presentation
Abstract
In the field of pharmacogenomics, a person’s genes are used to assist in determining their specific response to medications and to inform the correct dosage. Currently, the most effective solution for visualizing exact sequences of difficult genes – such as ones with many variations and a complex genome locus – is to use long read sequencing technology, such as the nanopore-based sequencing devices provided by Oxford Nanopore Technologies (ONT). This tool sequences DNA by passing molecules through nanopores and measuring the change in electrical current, which is used to infer DNA sequences (reads). The DNA molecules passed through the nanopore need to be modified with a motor protein in order to slow the sequencing down and produce visible and interpretable electrical signals. Normally, whole genome sequencing is used to get reads over the entire organism’s genome, but oftentimes only a specific part of a gene (such as one that metabolizes drugs) is needed. In targeted sequencing, many reads are needed from a single genomic locus in order to accurately predict sequences, so it is necessary to enrich a specific region of the genome. During targeted sequencing, an enzyme called Cas9 is used to selectively recognize and cut double-stranded DNA at specific genomic sites, allowing the targeted regions to be modified with sequencing adaptors and sequenced. This technique of using Cas9 to cut DNA originated from the CRISPR bacterial defense system used to protect against viruses. The main components of Cas9 cutting is the guide RNA (gRNA), which directs the Cas9 to the correct DNA sequence, and the Cas9 enzyme, which cuts the DNA into two pieces. Originally, Cas9 was reported to function as a blunt-end nuclease, cutting double stranded DNA without leaving base pair overhangs. The gRNA and Cas9 complex (also called the ribonucleoprotein or RNP complex) has been demonstrated to cut three base pairs upstream of the PAM site (the unique NGG sequence on the non-target strand of DNA) resulting in two DNA products – the PAM proximal DNA fragment, which includes the PAM site, and the PAM distal DNA fragment. During the lab’s initial experiments in the fall of 2024, we noticed that reads were found to be unequally distributed from the initial cut site when performing targeted sequencing using Cas9, with more reads starting at the PAM proximal end. This result was unexpected, because sequencing requires the ligation of adaptors; if both DNA ends were blunt, then they should receive adapters at the same rate and produce reads at the same rate. Upon looking into the literature, the same issue was found. Oxford Nanopore hypothesized the uneven pattern of reads was due to the Cas9 binding post-cleavage to the PAM distal end.1 However, after conducting follow-up experiments and using Proteinase K to remove the remaining Cas9, the same read bias was observed, suggesting post-cleavage Cas9 binding does not explain the asymmetry of reads. This observation was concerning because the bias reduces accuracy in targeted sequencing. With further research, we came across a paper claiming that Cas9 does not originally but blunt and may also have post-cleavage trimming.2 Therefore, we hypothesize that Cas9 is producing non-blunt DNA ends in addition to blunt ends primarily on the PAM distal fragment, explaining the observed bias because non-blunt DNA ends are not compatible with the current adapter? ligation strategy. To test this hypothesis, we designed an experiment to detect potential overhang structures generated by Cas9 cleavage and evaluate how these structures affect adaptor ligation efficiency during targeted sequencing. For this experiment, duplexed adaptors with various known overhangs will be ligated to Cas9 cut DNA and Eco53KI cut DNA. Eco53KI is confirmed to be a blunt cutting enzyme and will act as the control to compare to the Cas9 cut DNA. A combination of Oxford Nanopore Sequencing, TapeStation gel electrophoresis, and capillary electrophoresis will confirm or refute the presence of overhang structures. If the result is supportive, this experiment should show Cas9’s cutting mode pattern and the presence of specific overhang structures. Based on literature evidence, we anticipate observing both blunt and non-blunt DNA ends produced by Cas9, with variation due to the gRNA and target used, explaining the bias in the reads previously mentioned. Once the presence and nature of Cas9-induced overhangs are characterized at individual loci, this approach will be expanded into a high-throughput assay to systematically evaluate cutting behavior/efficiency across many guide RNAs. To accomplish this, 100 – 200 different guide RNAs will be used to explore Cas9 cutting efficiencies and overhang formation patterns. By incorporating the bioinformatics side of the Athey lab, a predictive computational model for the Cas9-induced overhang structures will be developed. Overall, the first part of this experiment will help develop an accurate assay for targeted sequencing and will provide a better understanding of Cas9. In turn, there will be more precision in clinical lab results, leading to improved drug metabolism predictions. The second part of this experiment will advance gene editing techniques using Cas9 and become a useful tool for genetics researchers more broadly. Together, these experiments will provide a foundation for effectively integrating Cas9-targeted sequencing into pharmacogenomics and significantly improve personalized medicine. Works Cited Targeted, amplification-free DNA sequencing using CRISPR/Cas. Oxford Nanopore Technologies. https://nanoporetech.com/document/targeted-amplification-free-dna-sequencing-using-crispr-cas Stephenson, A. A., Raper, A. T., & Suo, Z. (2018). Bidirectional Degradation of DNA Cleavage Products Catalyzed by CRISPR/Cas9. Journal of the American Chemical Society, 140(10), 3743–3750. https://doi.org/10.1021/jacs.7b13050
