Perks and considerations when targeting functional non-coding regions with CRISPR/Cas9

Abstract

Since the CRISPR system was discovered as an adaptive immune response in prokaryotic cells, the past decade has witnessed the engineering and deployment of CRISPR/Cas9 as one of the most efficient and powerful molecular tools. By leveraging the nuclease activity of CRISPR/Cas9, researchers are able to probe the biological functions of genetic elements and dissect molecular interactions by disrupting, activating or inactivating genes. In addition to biological research, the CRISPR/Cas9 toolkit has profoundly revolutionized gene therapy and agricultural products. However, there are many challenges regarding its efficiency, specificity and safety. Continuous efforts are being made to advance techniques and characterize the consequences of genome editing. In this thesis, we describe considerations when targeting genomic regions with CRISPR/Cas9 and provide methods to address some concerns related to efficiency and safety.

In Paper I, we introduced a non-hazardous method of transfecting human cells with large-size CRISPR/Cas9 vectors. By co-transfecting small-size vectors (3 kb) to cells, the delivery efficiency of CRISPR/Cas9 vectors (15 kb) and cell viability was significantly increased. The performance of the method has been verified in a number of hard-to-transfect human cell lines with both electroporation- and liposome-based transfection.

In Paper II, we revealed the complexity of CRISPR/Cas9-induced on-target genomic alterations by combining an advanced droplet-based target enrichment method followed by long-read sequencing and de novo assembly-based analysis. This approach enabled us to dissect the on-target sequence content in the order of kilobases, which was very challenging with many other available methods. With this tool, we uncovered the co-occurrence of multiple on-target rearrangements including duplication, inversion, as well as integrations of exogenous DNA and clustered interchromosomal rearrangements in CRISPR/Cas9-modified human cells. Furthermore, our study demonstrated that unintended genomic alterations could lead to the expression of DNA derived from both the target region and exogenous sources, as well as affect cell proliferation.

In Paper III, we reported a large unexpected genomic deletion in the HAP1 cell line, which is the one of most popular models used in CRISPR/Cas9-mediated experiments. This 287 kb deletion located on Chromosome 10 contains four widely-expressed protein-coding genes including the PTEN gene locus. We detected changes in histone acetylation and transcriptomes in HAP1 cells carrying the deletion. The loss of this genomic locus was not induced by Cas9 off-target nuclease activity. However, the generation of CRISPR/Cas9-modified cells significantly enhanced the frequency of the deletion among cell clones. Furthermore, our analysis indicated that this deletion initially found in HAP1 cells resembled a frequent deletion pattern driven by the PTEN gene in cancer patients.

In conclusion, we have presented two methods: one to improve delivery efficiency and another to detect on-target sequence content with higher resolution. Furthermore, we have revealed unintended genomic aberrations at targeted and non-targeted sites. These observations should be taken into consideration when modifying the genome with CRISPR/Cas9, and a comprehensive genomic validation is necessary.