Expanding CRISPR/Cas9 Genome Editing Capacity in Zebrafish Using SaCas9

The type II CRISPR/Cas9 system has been used widely for genome editing in zebrafish. However, the requirement for the 5′-NGG-3′ protospacer-adjacent motif (PAM) of Cas9 from Streptococcus pyogenes (SpCas9) limits its targeting sequences. Here, we report that a Cas9 ortholog from Staphylococcus aureus (SaCas9), and its KKH variant, successfully induced targeted mutagenesis with high frequency in zebrafish. Confirming previous findings, the SpCas9 variant, VQR, can also induce targeted mutations in zebrafish. Bioinformatics analysis of these new Cas targets suggests that the number of available target sites in the zebrafish genome can be greatly expanded. Collectively, the expanded target repertoire of Cas9 in zebrafish should further facilitate the utility of this organism for genetic studies of vertebrate biology.

Targeted genome engineering offers new genetic tools for understanding biological processes, and for developing therapeutics of human diseases. Following the development of Zinc Finger (ZFs) and transcription activator-like effector (TALEs) technology (Doyon et al. 2008;Huang et al. 2011;Meng et al. 2008), the clustered regularly interspaced short palindromic repeat/CRISPR-associated (CRISPR/Cas) system has emerged as the current gene editing tool of choice. CRISPR/Cas system has the advantages of ease of handling, low cost, and universal applicability in different cell types and organisms. CRISPR/Cas can be classified into six types based on the presence of ''signature genes'' (Makarova et al. 2011Shmakov et al. 2015;Wright et al. 2016). Among them, Cas9 from Streptococcus pyogenes (SpCas9), which belongs to the type II CRISPR/Cas system, has been demonstrated to be effective in inducing targeted DNA double strand breaks (DSBs) in a variety of organisms (Chang et al. 2013;Cong et al. 2013;Dickinson et al. 2013;Friedland et al. 2013;Gratz et al. 2013;Hwang et al. 2013;Jinek et al. 2013;Mali et al. 2013;Qin et al. 2015;Shalem et al. 2014;Wang et al. 2014;Yang et al. 2013a). SpCas9 nuclease DNA sequence specificity relies on a guide RNA with a protospacer-adjacent motif (PAM) sequence at the 39 end of a 20-bp target sequence. The most widely used SpCas9 recognizes a short 59-NGG-39 PAM. Since PAM sequences are different in different CRISPR/Cas systems, alternative PAMs would provide more flexibility for targeting strategies such as precise knock-in mutations.
Recently, Cas9 orthologs with distinct DNA binding specificity and PAM recognition, including Neisseria meningitidis (NmCas9), Streptococcus thermophilus1 (St1Cas9), and Staphylococcus aureus (SaCas9) have been applied for genome editing in human cells (Hou et al. 2013;Karvelis et al. 2013;Ran et al. 2015). Among them, Cas9 from S. aureus (SaCas9) is smaller, and has a longer PAM of 59-NNGRRT-39sequence. These features allow easier deliver by viral expression vectors, and higher sequence specificity, which would be more desirable for therapeutic applications. Recently, a SaCas9 variant (KKH SaCas9) with partially relaxed 59-NNNRRT-39 PAM specificities has been demonstrated to show robust genome editing activities in human cells, which further increases the SaCas9 targeting range (Kleinstiver et al. 2015a).
Here, we demonstrate that SaCas9, with its KKH SaCas9 variant, can edit the zebrafish genome with high targeting efficiency. This increases the frequency of available target sites, and expands the utility of CRISPR/Cas9 in zebrafish by targeting those previously inaccessible Cas9 sites in the genome.

Zebrafish husbandry and breeding
Wild type Tu fish and tg(mylpfa:EGFP) transgenic fish strains were raised and maintained at 28.5°in a circulating system. Zebrafish embryos were acquired from in-tank breeding. Development of embryos was staged by standard morphological criteria (Kimmel et al. 1995). All zebrafish experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Peking University. The reference from IACUC of Peking University is LSC-ZhangB-1.
Zebrafish microinjection, T7EI assays, and Sanger sequencing A solution (1-2 nl) containing Cas9 mRNA (300 ng/ml) and gRNA (30 ng/ml) was coinjected into one-cell-stage zebrafish embryos. Injected embryos were incubated at 28.5°for examination of phenotypes. After 2 d post fertilization (dpf), embryos that developed normally were collected for genotyping. Genomic DNA was extracted from pools of six randomly collected embryos by alkaline lysis buffer-based DNA extraction. Targeted genomic loci were amplified from genomic DNA, and then cloned into the pEASY-T1 vector (Transgene) for sequencing. T7E1 assays were performed as previously described for zebrafish (Hwang et al. 2013). The digested samples were analyzed through a 2% agarose gel. Quantification was based on relative band intensity using Quantity One software (Bio-Rad). All experiments were repeated three times.

Imaging
Zebrafish embryos were anesthetized with 0.03% Tricaine (Sigma-Aldrich), and mounted in 4% methylcellulose. Photographs were taken by a Zeiss Axio Imager Z1 microscope, and processed by Adobe Photoshop CC software.

Annotation of CRISPR target sites in coding exons
We searched for all potential CRISPR target sites of NGG, NGA, NNGRRN, NNGRRT, NNNRRN, and NNNRRT on both strands of the zebrafish genome (danRer10), and marked their chromosomal positions. Then, we produced a BED file to show all these PAM sites in exons as annotated by the UCSC browser.

Data availability
The authors state that all data necessary for confirming the conclusions presented in the article are represented fully within the article.

RESULTS AND DISCUSSION
Gene editing in zebrafish using SaCas9 First, we engineered SaCas9 to contain a NLS sequence in the pCS2+ expression vector, and used a simple one-step cloning-free PCR method to generate gRNA template for RNA transcription in vitro ( Figure 1A) Quantification is based on T7E1 assay. Error bars represent SEM, n = 6 embryos. Sequencing data of these targeted indels is included in Figure S2. (Bassett et al. 2013). Although SaCas9 can target all 59-NNGRR-39 PAMs, it cleaves the sequence of 59-NNGRRT-39 most efficiently (Ran et al 2015). We therefore selected this type of PAM sequence from the zebrafish genome for initial testing. Since pigmentation defect is a convenient phenotype indication for mutation in the tyrosinase (tyr) gene, which encodes an enzyme that converts tyrosine into melanin (Camp and Lardelli 2001), we designed a gRNA targeting tyr. After coinjection of tyr gRNA and SaCas9 mRNA into one-cellstage zebrafish embryos, pigmentation reduction was observed in 42 embryos (n = 73), some of which totally lacked pigmentation ( Figure 1B). Injected embryos developed normally, no obvious toxicity was observed. We randomly selected six embryos, extracted genomic DNA from pools of embryos and assessed the mutagenesis rates using the T7 endonuclease I (T7EI) assay and further confirmed indels at the expected sites by sequencing ( Figure 1D).
To test if SaCas9/gRNA can be used generally in zebrafish, we constructed eight additional gRNAs, one targeting transgene EGFP and others targeting endogenous genes. When EGFP SaCas9/gRNA was injected into tg (mylpfa:EGFP) zebrafish eggs, loss of EGFP expression was observed in muscle cells ( Figure 1C). For other specific sites, indels were detected in all injected embryos with a frequency of 13-76%. Again, the indels were sequencing confirmed at the expected genomic loci ( Figure 1E and Figure S1). Taken together, these results demonstrate that SaCas9 can effectively edit genome in zebrafish.
Gene editing in zebrafish using KKH SaCas9 Native SaCas9 recognizes a longer 59-NNGRRT-39 PAM, which occurs on average every 32 bps of random DNA. Obtaining variants with relaxed or partially relaxed specificity to expand the repertoire of gene editing tools in zebrafish is desirable. Toward this goal, we investigated two recently described SaCas9 and SpCas9 variants with different specificities ( Figure 2A) (Kleinstiver et al. 2015a(Kleinstiver et al. , 2015b. One SaCas9 variant is called KKH SaCas9, which can target 59-NNNRRT-39 PAM in human genome (Kleinstiver et al. 2015a). We evaluated the activity of KKH SaCas9 by targeting four endogenous genes in zebrafish, and observed indels in all of them ( Figure 2B and Figure S2). Similarly, we evaluated activity of the SpCas9 VQR variant with an NGA PAM, which has been shown previously to induce indels in zebrafish and Caenorhabditis elegans (Bell et al. 2016;Kleinstiver et al. 2015b), and obtained expected indels at the two predicted sites ( Figure 2B and Figure S2). Collectively, these results demonstrate that both KKH SaCas9 and VQR SpCas9 variants are functional in zebrafish. Since 59-NNNRRT-39 represents a relaxed PAM, it is conceivable that the off-target rate might be increased. In zebrafish, this potential problem could be addressed by analyzing the linkage between the targeted mutation, and the specific phenotype in large numbers of individual fish through breeding multiple generations.
The SpCas9 system is an efficient tool for genome editing in many organisms including bacteria, yeast, human cells, C. elegans, zebrafish, and mice (Hsu et al. 2014;Sternberg and Doudna 2015). Due to the restriction of having to use the 59-NGG-39 PAM required by SpCas9, it is challenging to find target sites with high efficiency in a specific genome locus. Improving the capacity to target very specific loci is highly desirable for DNA homology-directed gene repair, which is most efficient when a DSB is placed within 10-20 bp of a desired alteration (Elliott et al. 1998;Findlay et al 2014;Yang et al. 2013b). Here, we demonstrate that using SaCas9 and its variant enable genome editing in a significantly wider range in zebrafish. Since Cas9-based gene editing is most commonly applied to induce or repair mutations in coding exons of a gene, we produced a BED file to annotate six types of target sites (NGG, NGA, NNGRRN, NNGRRT, NNNRRN, and NNNRRT) on the UCSC gene browser.
One of the key issues in using CRISPR/Cas9 system successfully is to assure high activity. First, similar to the zebrafish-codon-optimized SpCas9, which increases gRNA targeting efficiency from 4% to 39% (Liu et al. 2014), we recommend constructing a zebrafish codon-optimized SaCas9 for future study. Second, it is also possible that direct SaCas9 protein injections could further increase efficiency (Burger et al. 2016;Gagnon et al. 2014). Finally, as modifying the SpCas9/gRNA structure by extending the duplex length has been shown to improve knockout efficiency (Dang et al. 2015), we speculate that the same strategy could be applied to SaCas9/gRNA structure in zebrafish.