Expanding the CRISPR-Cas9 toolbox for genome editing

The transformative impact of CRISPR-Cas systems has revolutionized the landscape of genome editing research, making unprecedented strides into clinical settings for treating diseases for which there are no treatments. Nonetheless, the CRISPR toolbox still needs to be advanced to tackle the heterogeneity of genetic diseases and the obstacles of in vivo therapeutic applications. Primary problems when using CRISPR systems concern not only the balance between efficiency and precision, but also size, which becomes critical when targeting specific organs depending on adeno-associated viral (AAV) vectors for delivery. This thesis work aimed to enlarge the CRISPR-based editing toolbox by introducing new Cas9 proteins. Two approaches were exploited: 1) rational molecular engineering to obtain a high-fidelity version of the Streptococcus pyogenes Cas9 (SpCas9), and 2) the identification of new CRISPR-Cas9 loci from metagenomic data. In the first approach, a high-fidelity SpCas9 variant (rCas9HF) was generated starting from the pool of amino acid substitutions resulting from the screening of evoCas9. Differently from evoCas9, rCas9HF is characterized by the K526D substitution, which allows efficient editing through ribonucleoprotein (RNP) electroporation. rCas9HF indels generation and off-target effects are comparable to HiFi Cas9, which is currently the only available high-fidelity Cas9 that can be used as RNP. The two high-fidelity mutants were also tested for homology-directed repair (HDR) and ability to modify the genome in CD34+ cells, showing that the variants assure a benefit compared to wild-type (wt) SpCas9 in terms of precision. Moreover, rCas9HF and HiFi Cas9 are characterized by different editing profiles, providing a valuable new precise tool for genome editing. In the second approach, CoCas9 was identified from a massively expanded microbiome repository as a nuclease characterized by compact molecular size coupled with an editing efficiency on genomic loci slightly similar to SpCas9. CoCas9 was demonstrated to be very precise, compatible with base editing technology, and deliverable in vivo through a single AAV, as demonstrated in transducing experiments targeting the mouse retina. When compared to other <1100 amino acid (aa) Cas9s, this new ortholog showed better or comparable on-target efficiency, with a better specificity profile and broader PAM. Overall, this thesis uncovers and characterizes two Cas9s, further enriching the CRISPR toolbox, and thus improving the ability to efficiently target the variety of pathogenic mutations in the human genome.