CRISPR-Cas9 genome editing has transformed biotechnology and therapeutics. However, in vivo applications of some Cas9s are hindered by large size (limiting delivery by adeno-associated virus [AAV] vectors), off-target editing, or complex protospacer-adjacent motifs (PAMs) that restrict the density of recognition sequences in target DNA. Here, we exploited natural variation in the PAM-interacting domains (PIDs) of closely related Cas9s to identify a compact ortholog from Neisseria meningitidis-Nme2Cas9-that recognizes a simple dinucleotide PAM (N4CC) that provides for high target site density. All-in-one AAV delivery of Nme2Cas9 with a guide RNA targeting Pcsk9 in adult mouse liver produces efficient genome editing and reduced serum cholesterol with exceptionally high specificity. We further expand our single-AAV platform to pre-implanted zygotes for streamlined generation of genome-edited mice. Nme2Cas9 combines all-in-one AAV compatibility, exceptional editing accuracy within cells, and high target site density for in vivo genome editing applications.

Recent advances with the bacterial CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) defense system as genome editing tools have opened a new avenue for targeting disease-causing mutations. The programmability of the Cas9 endonuclease by RNA makes it a potentially powerful therapeutic tool to correct such mutations. The CRISPR-Cas9 system consists of a Cas9 endonuclease that is guided by RNA (sgRNA) to create double-stranded breaks in a target DNA segment complementary to the guide. This process is dependent on a 2-8 nucleotide sequence (called PAM) that is adjacent to the target and functions as a Cas9 binding signal. Each Cas9 ortholog recognizes a unique PAM. However, factors such as the size of Cas9 or the frequency of its PAM sequence in the genome have hindered its clinical use. The Cas9 from Streptococcus pyogenes (SpyCas9) is commonly used in research because its PAM (NGG, where “N” symbolizes any nucleotide) is present every ~8 bp in the genome, providing robust targeting potential. However, it is too large to fit into typical viral vectors used for in vivo delivery, namely adeno-associated vectors (AAV). While several Cas9 orthologs have been characterized, none satisfied the need for a compact, accurate Cas9 with a short PAM. In this thesis, we use two approaches to identify new compact Cas9 orthologs with small PAMs, one using anti-CRISPR proteins and one by searching through closely related Cas9s. First, we use the presence of anti-CRISPRs (naturally occurring, phage-encoded peptides that inhibit CRISPR-Cas9 described in chapter 2) in a genome as indicators of Cas9s that may be highly active. These orthologs come with the added advantage of having inhibitors that can be used as off-switches. We characterize four Cas9s that are targeted by anti-CRISPR proteins and show that they recognize diverse PAMs in vitro. One of the four Cas9’s, namely HpaCas9 from Haemophilus parainfluenzae, induces efficient genome editing in mammalian cells. However, its long N4GATTT PAM does not satisfy the short PAM criterion. For our second approach, we asked whether closely related Cas9 orthologs with drastically different PAM-interacting domains (PIDs, the domain responsible for PAM recognition) recognize different PAMs, and if so, can be used for genome editing. To this end, we exploited natural variation in the PID of closely related Cas9s to identify a compact ortholog from Neisseria meningitidis (Nme2Cas9). Nme2Cas9 recognizes a simple dinucleotide PAM (N4CC) that provides a high target site density. All-in-one AAV delivery of Nme2Cas9 with a guide RNA into adult mouse liver produces efficient genome editing and reduced serum cholesterol with exceptionally high specificity. We further expand our single-AAV platform to pre-implanted zygotes for streamlined generation of genome-edited mice. Finally, we show preliminary data on how CRISPR-Cas9 can be used for therapeutic genome editing for Amytrophic Lateral Sclerosis. Our new findings promise to accelerate the development of genome editing tools for biomedical and therapeutic applications.

BACKGROUND: The development of CRISPR genome editing has transformed biomedical research. Most applications reported thus far rely upon the Cas9 protein from Streptococcus pyogenes SF370 (SpyCas9). With many RNA guides, wildtype SpyCas9 can induce significant levels of unintended mutations at near-cognate sites, necessitating substantial efforts toward the development of strategies to minimize off-target activity. Although the genome-editing potential of thousands of other Cas9 orthologs remains largely untapped, it is not known how many will require similarly extensive engineering to achieve single-site accuracy within large genomes. In addition to its off-targeting propensity, SpyCas9 is encoded by a relatively large open reading frame, limiting its utility in applications that require size-restricted delivery strategies such as adeno-associated virus vectors. In contrast, some genome-editing-validated Cas9 orthologs are considerably smaller and therefore better suited for viral delivery. RESULTS: Here we show that wildtype NmeCas9, when programmed with guide sequences of the natural length of 24 nucleotides, exhibits a nearly complete absence of unintended editing in human cells, even when targeting sites that are prone to off-target activity with wildtype SpyCas9. We also validate at least six variant protospacer adjacent motifs (PAMs), in addition to the preferred consensus PAM (5'-N4GATT-3'), for NmeCas9 genome editing in human cells. CONCLUSIONS: Our results show that NmeCas9 is a naturally high-fidelity genome-editing enzyme and suggest that additional Cas9 orthologs may prove to exhibit similarly high accuracy, even without extensive engineering.