Improved Base Editing Technologies With Novel Editors and Assays

Abstract

Genome engineering is a rapidly evolving area of study. One driver of the breakneck speed with which the field is moving forward is the application of CRISPR/Cas9. Since its introduction in 2013, CRISPR/Cas9 has completely changed the ease and utility of genome engineering and has revolutionized the field. The use of CRISPR/Cas9 to directly edit genes, increase or decrease gene expression, or even image genomic loci is widely accepted and extensively used in models ranging from bacteria to mammals. The recent development of second-generation CRISPR editing tools has opened even more doors into how the human genome can be manipulated. Past methods to introduce a single-base substitution into genomes involved creating a double-strand break and taking advantage of the cellular repair pathway homologous recombination to incorporate a donor plasmid into the genomic sequence. These methods are inefficient and can result in introduction of the donor template into numerous unrelated loci throughout the genome, unwanted insertions or deletions, or chromosomal translocations. Base editing unlocks a method to introduce single-base substitutions without the need for a donor template or the creation of a double-strand break. By fusing rat APOBEC1, a natural cytosine deaminase, to Cas9 nickase and uracil DNA glycosylase inhibitor, the Liu lab was the first to create an RNA-guided base editor that can change target cytosines to thymines. In just a year and a half, significant improvements have been made on this front, including making the original base editor more efficient and specific and even introducing editors with the power to mutate adenosines to guanosines. Despite the advancements constantly being made to this technology, there is still room for improvement. The current base editors such as BE3 can edit unintended cytosines in the target sequence at high rates. In addition, the combination of the nickase activity of Cas9 and the abasic site created after deamination of the target cytosine can create a pseudo double-strand break resulting in creation of unwanted insertions or deletions. Finally, targeting at endogenous loci continues to hover between 30% and 80%, depending on the method used and model genome being targeted. Targeting rates could benefit from growth, especially as this technology is being considered for therapeutic applications. A bottleneck in the process of developing new and improved base editing technology is the time and effort that is required to quantify editing efficiencies. Most studies use next-generation sequencing to quantify editing rates. The preparation of samples and quality control required can take up to six weeks. If using a core at a larger university or research institution there is additional time spent waiting in a queue to use a sequencing instrument. The field is in need of a rapid method to quantify base editing in real time that is transferable to multiple cellular systems. Here I report two bicistronic, fluorescence-based systems for the quantification of base editing activity. By changing a 5’-TT-3’ dinucleotide motif to a 5’-TC-3’ dinucleotide motif in eGFP or mCherry, I simultaneously ablate fluorescence and create an APOBEC-preferred mutational hotspot. When the cytosine is reverted back to a thymine, fluorescence is restored. This tight off-to-on system allows for real-time quantification of base editing activity through fluorescence microscopy or flow-cytometry. After creating a novel base editing reporter system, I hypothesized that I could use the newly designed assay to create more efficient and specific base editors. Using members of the human APOBEC3 family of enzymes, I created a suite of novel base editors. These base editors have advantages over rat APOBEC1-based editors in that the structure of many APOBEC3s are known, allowing for easier structure-guided evolution to improve their editing activity. As a proof of concept, I used this knowledge to evolve APOBEC3H haplotype II into a more efficient base editor by making only a few amino acid substitution mutations. In addition, we were able to create base editors using APOBEC3A and the catalytic domain of APOBEC3B that surpass BE3 in editing efficiency. Taken together, these data contribute positively to the genome engineering field and open new doors for continuing development of this technology.