Engineering and evolving molecular Cas-based therapies

Precision genome engineering has dramatically advanced with the development of CRISPR/Cas base editing systems that include cytosine base editors (CBEs), adenine base editors (ABEs) and cytosine-to-guanine base editors (CGBEs). They facilitate the targeted, processive (bystander) conversion of single nucleobases in a user-defined manner, in which a bespoke guide sequence (sgRNA) enables site-specific cytosine-to-thymine (CBEs), adenine-to-guanine (ABEs) or cytosine-to-guanine (CGBEs) nucleotide conversion.

We first catalogue the current rich diversity of base editors and present a hypothetical taxonomic and phylogenetic framework for the classification of over 200 different DNA base editors. Following evaluation of their in situ activity windows, which were derived by cataloguing their activity in published literature, organisation is done hierarchically with specific base editor signatures being subcategorized according to their on-target activity, or non-specific, genome- or transcriptome-wide activity. We curate a phylogenetic framework, based on protein homology alignment, and then describe a taxonomic structure that clusters base editor variants first on their target chemistry, endonuclease component, identity of their deaminase component, and finally their described properties into discrete taxa. Subsequently, we summarise our findings into a navigable database (ShinyApp in R) that allows users to select through our repository to nominate ideal base editor candidates as a starting point for further testing in their specific application.

Next, to evaluate the efficacy of using an existing array of CBE-variants in a therapeutic sense, we correct a common high-risk allele that is associated with age-related macular degeneration (AMD) (rs1061170; NM_000186.3:c.1204T>C; NP_000177.2:p.His402Tyr) using a transgenic HEK293A cell line. Several different CBE constructs (BE3, SaBE3, SaKKH-BE3, VQR-BE3, and Target-AID) and their respective sgRNA-expression cassettes were screened against this target sequence. Their on-target activities were determined using Sanger sequencing, followed by next-generation deep sequencing of the best performing variant for further assessment. We find that the use of Target-AID and its respective sgRNA demonstrated an appreciable editing efficiency of 21.5% of the total sequencing reads for cytosine-to-thymine conversions. Additionally, the incidence of insertions and deletions (indels) were rare, with virtually no DNA off-target effects observed across the top 11 predicted off-target sites based on their similarity to the target sgRNA.

Encouraged by these preliminary results with CBEs, we decided to further explore base editing technology by developing our own ABE base editing tool in an effort to increase on-target editing efficiency and enhance fidelity using the principles of protein engineering. We compare the editing profile of circularly permuted and domain-inlaid Cas9 base editors, and find that structural permutation of the ABE can affect the frequency of RNA off-target events. With this insight, structure-guided design was used to engineer an SaCas9 ABE variant (microABE I744) that has dramatically improved on-target editing efficiency and a reduced RNA-off target footprint compared to analogous N-terminal linked SaCas9 ABE variants.

Finally, while protein engineering represented a useful avenue for the generation of novel base editing tools, it requires some level of a priori knowledge of protein domains and thus restricts us to optimising only a subset of naturally-occuring deaminases. Therefore, we decided to develop a platform for the directed evolution of biomolecules. We have established a continuous, circuitous platform that allows for the bespoke diversification of protein targets in mammalian cells. By engineering a stably split, recombination-averse Sindbis viral variant, we can facilitate successive propagation and enable robust transgene expression under selective conditions. This allows for the enrichment of biomolecule variants-of-interest for enhanced activity, improved kinetics, or to impart novel functions. Our system supports continuous amplificatory replication in an activity-gated manner, while our inbuilt biosafety measures counter selects against replication-competency as well as transgene dilution.

As proof-in-concept, we use CAESAR (Capsid-complimented Adaptation of Engineered Sindbis virus with Activity-gated Replication) to target an adenosine deaminase variant, TadA-RICFE, as a candidate for directed evolution to develop a de novo ABE variant. Adenosine deaminases have no reported catalytic activity on DNA substrates, and instead natively mediate the conversion of adenosine-to-inosine in RNA. We subjected TadA-RICFE to five rounds of serial passaging under varying conditions of viral transduction (high and low multiplicity-of-infection [MOI]). Under selection, a high selective burden upon variants that fail to acquire DNA-interacting mutations. Following long-read nanopore sequencing, we observed a general enrichment in key mutations at residues that have been previously shown to interact and mediate DNA catalysis in the prototypic ABE homolog. Preliminary characterisation of these mutations however did not support their role in mediating adenine-to-guanine base editing, but curiously showed that there was a 2.2-fold enrichment in cytosine-to-thymine editing and some residual activity mediating cytosine-to-guanine editing at position five of the FANCF loci. This provides us with scope for further in vitro characterisation of our potentially novel ABE variant, and further lays a strong foundation for additional evolution using more robust evolutionary circuits. The burgeoning era of precision medicine places CRISPR/Cas base editing technology as a frontier platform for personalised biologics. This thesis explores the application, optimization and the generation of these tools. Herein, we illustrate the diversity of current generation base editing tools, and further demonstrate that both rational and stochastic design principles can be used to further expand the available repertoire of base editing tools. Finally, after exploring tools for the directed evolution of new base editors, we describe the engineering of a novel system for the directed evolution of proteins in mammalian cells.