nprot.2013.143

© 20 13 N at ur e A m er ic a, In c. A ll rig ht s re se rv ed . protocol nature protocols | VOL.8 NO.11 | 2013 | 2281 IntroDuctIon The ability to engineer biological systems and organisms holds enormous potential for applications across basic science, medi- cine and biotechnology. Programmable sequence-specific endo- nucleases that facilitate precise editing of endogenous genomic loci are now enabling systematic interrogation of genetic ele- ments and causal genetic variations1,2 in a broad range of species, including those that have not previously been genetically tracta- ble3–6. A number of genome editing technologies have emerged in recent years, including zinc-finger nucleases (ZFNs)7–10, transcription activator–like effector nucleases (TALENs)10–17 and the RNA-guided CRISPR-Cas nuclease system18–25. The first two technologies use a strategy of tethering endonuclease catalytic domains to modular DNA-binding proteins for inducing targeted DNA double-stranded breaks (DSBs) at spe- cific genomic loci. By contrast, Cas9 is a nuclease guided by small RNAs through Watson-Crick base pairing with target DNA26–28 (Fig. 1), representing a system that is markedly easier to design, highly specific, efficient and well-suited for high- throughput and multiplexed gene editing for a variety of cell types and organisms. Precise genome editing using engineered nucleases Similarly to ZFNs and TALENs, Cas9 promotes genome editing by stimulating a DSB at a target genomic locus29,30. Upon cleavage by Cas9, the target locus typically undergoes one of two major pathways for DNA damage repair (Fig. 2): the error-prone NHEJ or the high-fidelity HDR pathway, both of which can be used to achieve a desired editing outcome. In the absence of a repair tem- plate, DSBs are re-ligated through the NHEJ process, which leaves scars in the form of insertion/deletion (indel) mutations. NHEJ can be harnessed to mediate gene knockouts, as indels occurring within a coding exon can lead to frameshift mutations and prema- ture stop codons31. Multiple DSBs can additionally be exploited to mediate larger deletions in the genome22,32. HDR is an alternative major DNA repair pathway. Although HDR typically occurs at lower and substantially more variable frequencies than NHEJ, it can be leveraged to generate precise, defined modifications at a target locus in the presence of an exo- genously introduced repair template. The repair template can either be in the form of conventional double-stranded DNA targeting constructs with homology arms flanking the insertion sequence, or single-stranded DNA oligonucleotides (ssODNs). The latter provides an effective and simple method for making small edits in the genome, such as the introduction of single- nucleotide mutations for probing causal genetic variations32. Unlike NHEJ, HDR is generally active only in dividing cells, and its efficiency can vary widely depending on the cell type and state, as well as the genomic locus and repair template33. Cas9: an RNA-guided nuclease for genome editing CRISPR-Cas is a microbial adaptive immune system that uses RNA-guided nucleases to cleave foreign genetic elements18–21,26. Three types (I–III) of CRISPR systems have been identified across a wide range of bacterial and archaeal hosts, wherein each system comprises a cluster of CRISPR-associated (Cas) genes, noncoding RNAs and a distinctive array of repetitive elements (direct repeats). These repeats are interspaced by short variable sequences20 derived from exogenous DNA targets known as protospacers, and together they constitute the CRISPR RNA (crRNA) array. Within the DNA target, each protospacer is always associated with a protospacer adjacent motif (PAM), which can vary depending on the specific CRISPR system34–36. The Type II CRISPR system is one of the best characterized26–28,37,38, consisting of the nuclease Cas9, the crRNA array that encodes the guide RNAs and a required auxiliary trans-activating crRNA (tracrRNA) that facilitates the processing of the crRNA array into discrete units26,28. Each crRNA unit then contains a 20-nt guide sequence and a partial direct repeat, where the former directs Cas9 to a 20-bp DNA target via Watson-Crick base pair- ing (Fig. 1). In the CRISPR-Cas system derived from Streptococcus pyogenes (which is the system used in this protocol), the target DNA must immediately precede a 5′-NGG PAM27, whereas Genome engineering using the CRISPR-Cas9 system F Ann Ran1–5,8, Patrick D Hsu1–5,8, Jason Wright1, Vineeta Agarwala1,6,7, David A Scott1–4 & Feng Zhang1–4 1Broad Institute of Massachusetts Institute of Technology (MIT) and Harvard, Cambridge, Massachusetts, USA. 2McGovern Institute for Brain Research, Cambridge, Massachusetts, USA. 3Department of Brain and Cognitive Sciences, MIT, Cambridge, Massachusetts, USA. 4Department of Biological Engineering, MIT, Cambridge, Massachusetts, USA. 5Department of Molecular and Cellular Biology, Harvard University, Cambridge, Massachusetts, USA. 6Program in Biophysics, Harvard University, MIT, Cambridge, Massachusetts, USA. 7Harvard-MIT Division of Health Sciences and Technology, MIT, Cambridge, Massachusetts, USA. 8These authors contributed equally to this work. Correspondence should be addressed to F.Z. (zhang@broadinstitute.org). Published online 24 October 2013; doi:10.1038/nprot.2013.143 targeted nucleases are powerful tools for mediating genome alteration with high precision. the rna-guided cas9 nuclease from the microbial clustered regularly interspaced short palindromic repeats (crIspr) adaptive immune system can be used to facilitate efficient genome engineering in eukaryotic cells by simply specifying a 20-nt targeting sequence within its guide rna. Here we describe a set of tools for cas9-mediated genome editing via nonhomologous end joining (nHeJ) or homology-directed repair (HDr) in mammalian cells, as well as generation of modified cell lines for downstream functional studies. to minimize off-target cleavage, we further describe a double-nicking strategy using the cas9 nickase mutant with paired guide rnas. this protocol provides experimentally derived guidelines for the selection of target sites, evaluation of cleavage efficiency and analysis of off-target activity. Beginning with target design, gene modifications can be achieved within as little as 1–2 weeks, and modified clonal cell lines can be derived within 2–3 weeks. © 20 13 N at ur e A m er ic a, In c. A ll rig ht s re se rv ed . protocol 2282 | VOL.8 NO.11 | 2013 | nature protocols other Cas9 orthologs may have different PAM requirements, such as those of S. thermophilus (5′-NNAGAA22,26 for CRISPR1 and 5′-NGGNG28,37 for CRISPR3) and Neisseria meningiditis (5′-NNNNGATT)39. The RNA-guided nuclease function of CRISPR-Cas is recon- stituted in mammalian cells through the heterologous expres- sion of human codon–optimized Cas9 and the requisite RNA components22–25. Furthermore, the crRNA and tracrRNA can be fused together to create a chimeric, single-guide RNA (sgRNA)27 (Fig. 1). Cas9 can thus be re-directed toward almost any target of interest in immediate vicinity of the PAM sequence by altering the 20-nt guide sequence within the sgRNA. Given its ease of implementation and multiplexing capacity, Cas9 has been used to generate engineered eukaryotic cells car- rying specific mutations via both NHEJ and HDR22–25,40. Direct injection of sgRNA and mRNA encoding Cas9 into embryos has enabled the rapid generation of transgenic mice with multiple modified alleles41,42. These results hold enormous promise for editing organisms that are otherwise genetically intractable. Cas9 nucleases carry out strand-specific cleavage by using the conserved HNH and RuvC nuclease domains, which can be mutated and exploited for additional function37. An aspartate-to-alanine (D10A) mutation in the RuvC catalytic domain27,28 allows the Cas9 nickase mutant (Cas9n) to nick rather than cleave DNA to yield single-stranded breaks, and the subsequent preferential repair through HDR22 can poten- tially decrease the frequency of unwanted indel mutations from off-target DSBs. Appropriately offset sgRNA pairs can guide Cas9n to simultaneously nick both strands of the target locus to mediate a DSB, thus effectively increasing the specificity of target recognition43. In addition, a Cas9 mutant with both DNA- cleaving catalytic residues mutated has been adapted to enable transcriptional regulation in Escherichia coli44, demonstrating the potential of functionalizing Cas9 for diverse applications, such as recruitment of fluorescent protein labels or chromatin- modifying enzymes to specific genomic loci for reporting or modulating gene function. Here we explain in detail how to use a human codon– optimized, nuclear localization sequence-flanked wild-type (WT) Cas9 nuclease or mutant Cas9 nickase to facilitate eukaryotic gene editing. We describe considerations for design- ing the 20-nt guide sequence, protocols for rapid construction and functional validation of sgRNAs and finally the use of the Cas9 nuclease to mediate both NHEJ- and HDR-based genome modifications in human embryonic kidney (HEK 293FT) and human stem cell (HUES9) lines (Fig. 3). The Cas9 system can similarly be applied to other cell types and organisms, includ- ing humans22,23,25, mice22,41,45, zebrafish45, Drosophila46 and Caenorhabditis elegans47. Comparison with other genome editing technologies As with other designer nuclease technologies such as ZFNs and TALENs, Cas9 can facilitate targeted DNA DSBs at specific loci of interest in the mammalian genome and stimulate genome editing via NHEJ or HDR. Cas9 offers several potential advantages over ZFNs and TALENs, including the ease of customization, higher targeting efficiency and the ability to facilitate multiplex genome editing. As custom ZFNs are often difficult to engineer, we will primarily compare Cas9 with TALEN. Ease of customization. Cas9 can be easily retargeted to new DNA sequences by simply purchasing a pair of oligos encoding the 20-nt guide sequence. In contrast, retargeting of TALEN for a new DNA sequence requires the construction of two new TALEN genes. Although a variety of protocols exist for TALEN con- struction14,17,48,49, it takes substantially more hands-on time to construct a new pair of TALENs. Cleavage pattern. WT S. pyogenes Cas9 (SpCas9) is known to make a blunt cut between the 17th and 18th bases in the target sequence (3 bp 5′ of the PAM)27. Mutating catalytic residues in either the RuvC or the HNH nuclease domain of SpCas9 con- verts the enzyme into a DNA nicking enzyme22,27. In contrast, TALENs cleave nonspecifically in the 12–24-bp linker between the pair of TALEN monomer-binding sites50. • • 5′ PAMTarget (20 bp) sgRNA 5′ 3′ 3′ 5′ Genomic locus 3′ DNA target Cas9 ..AATGGGGAGGACATCGATGTCACCTCCAATGACTAGGGTGGGCAACCAC.. |||||||||||||||||| ||||||||||| ..TTACCCCTCCTGTAGCTACAGTGGAGGTTACTGATCCCACCCGTTGGTG.. |||||||||||||||||||| GTCACCTCCAATGACTAGGGGUUUUAGAGCUAG A A •|||||• |||| GUUCAACUAUUGCCUGAUCGGAAUAAAAUU CGAUA |||| GAA AAAGUGGCACCGA •|||||||G UUUUUUCGUGGCU A A Figure 1 | Schematic of the RNA-guided Cas9 nuclease. The Cas9 nuclease from S. pyogenes (in yellow) is targeted to genomic DNA (shown for example is the human EMX1 locus) by an sgRNA consisting of a 20-nt guide sequence (blue) and a scaffold (red). The guide sequence pairs with the DNA target (blue bar on top strand), directly upstream of a requisite 5′-NGG adjacent motif (PAM; pink). Cas9 mediates a DSB ~3 bp upstream of the PAM (red triangle). NHEJ HDR DSB 5′ 3′ 3′ 5′ sgRNA Cas9 |||||| Indel mutation Premature stop codon Precise gene editing Genomic 5′ DNA 3′5′ 3′ 5′ 3′ 3′ 5′ 3′ 5′ Repair 5′ template 3′ 5′ 3′ 3′ 5′ 3′ 5′ 3′ 5′ Figure 2 | DSB repair promotes gene editing. DSBs induced by Cas9 (yellow) can be repaired in one of two ways. In the error-prone NHEJ pathway, the ends of a DSB are processed by endogenous DNA repair machinery and rejoined, which can result in random indel mutations at the site of junction. Indel mutations occurring within the coding region of a gene can result in frameshifts and the creation of a premature stop codon, resulting in gene knockout. Alternatively, a repair template in the form of a plasmid or ssODN can be supplied to leverage the HDR pathway, which allows high fidelity and precise editing. Single-stranded nicks to the DNA can also induce HDR. © 20 13 N at ur e A m er ic a, In c. A ll rig ht s re se rv ed . protocol nature protocols | VOL.8 NO.11 | 2013 | 2283 Editing efficiency. SpCas9 and TALENs have both been shown to facilitate efficient genome editing in a variety of cell types and organisms. However, owing to the ease of targeting, Cas9 can be used to target multiple genomic loci simultaneously, by co-delivering a combination of sgRNAs to the cells of interest. Limitations of the Cas9 system Cas9 can be targeted to specific genomic loci via a 20-nt guide sequence on the sgRNA. The only requirement for the selec- tion of Cas9 target sites is the presence of a PAM sequence directly 3′ of the 20-bp target sequence. Each Cas9 ortholog has a unique PAM sequence; for example, SpCas9 requires a 5′-NGG PAM sequence. This PAM requirement does not severely limit the targeting range of SpCas9—in the human genome, such target sites can be found on average every 8–12 bp (refs. 22,51). In addition to the targeting range, another possi- ble limitation is the potential for off-target mutagenesis; please see Boxes 1 and 2 for details and strategies on minimizing off- target modifications. Experimental design Target selection for sgRNA. The specificity of the Cas9 nuclease is determined by the 20-nt guide sequence within the sgRNA. For the S. pyogenes system, the target sequence (e.g., 5′-GTC ACCTCCAATGACTAGGG-3′) must immediately precede (i.e., be 5′ to) a 5′-NGG PAM, and the 20-nt guide sequence base pairs with the opposite strand to mediate Cas9 cleavage at ~3 bp upstream of the PAM (Figs. 1 and 4a top strand example). Note • that the PAM sequence is required to immediately follow the tar- get DNA locus, but that it is not a part of the 20-nt guide sequence within the sgRNA. Thus, there are two main considerations in the selection of the 20-nt guide sequence for gene targeting: (i) the 5′-NGG PAM for S. pyogenes Cas9 and (ii) the minimization of off-target activity51,52. We provide an online CRISPR Design Tool (http:// tools.genome-engineering.org) that takes a genomic sequence of interest and identifies suitable target sites. To experimentally assess off-target genomic modifications for each sgRNA, we also provide computationally predicted off-target sites (for a detailed discus- sion, see Box 1) for each intended target, ranked according to our quantitative specificity analysis on the effects of base-pairing mismatch identity, position and distribution. For increased targeting specificity, an alternative strategy using the D10A nick- ase mutant of Cas9 (Cas9n) along with a pair of sgRNAs may be used. The design criteria for orientation and spacing of such sgRNA pairs are described in Box 2. The CRISPR Design Tool provides the sequences for all oligos and primers necessary for (i) preparing the sgRNA constructs, (ii) assaying target modification efficiency and (iii) assessing cleavage at potential off-target sites. It is worth noting that because the U6 RNA polymerase III promoter used to express the sgRNA prefers a guanine (G) nucleotide as the first base of its transcript59, an extra G is appended at the 5′ of the sgRNA where the 20-nt guide sequence does not begin with G (Fig. 4b,c). On rare occasions, certain sgRNAs may not work for reasons yet unknown; therefore, we recommend designing at least two sgRNAs for each locus and testing their efficiencies in the intended cell type. Approaches for sgRNA construction and delivery. Depending on the desired application, sgRNAs can be delivered as either PCR amplicons containing an expression cassette (Fig. 4b) or sgRNA-expressing plasmids (Fig. 4c). PCR-based sgRNA deliv- ery appends the custom sgRNA sequence onto the reverse PCR primer used to amplify a U6 promoter template (Fig. 4b). The resulting amplicon could be co-transfected with a Cas9 expres- sion plasmid pSpCas9. This method is optimal for rapid screen- ing of multiple candidate sgRNAs, as cell transfections for functional testing can be performed shortly after obtaining the sgRNA-encoding primers. Because this simple method obviates the need for plasmid-based cloning and sequence verification, it is well suited for testing or co-transfecting a large number of sgRNAs for generating large knockout libraries or other scale- sensitive applications. Note that the sgRNA-encoding primers are pSpCas9(sgRNA) sgRNA SpCas9 SURVEYOR assay Isolate clonal lines Expand Genotype pSpCas9 (sgRNA) D ay 1 St ep s 1– 4 In s ilic o de sig n D ay s 2– 5 St ep 5 R ea ge nt c on st ru ct io n D ay s 5– 8 St ep s 6– 12 6 Fu nc tio na l v al id at io n D ay s 9– 28 St ep s 54 –7 0 Cl on al e xp an sio n Transfect Repair template (optional) Figure 3 | Timeline and overview of experiments. Steps for reagent design, construction, validation and cell line expansion are depicted. Custom sgRNAs (light blue bars) for each target, as well as genotyping primers, are designed in silico via the CRISPR Design Tool (http://tools.genome- engineering.org). sgRNA guide sequences can be cloned into an expression plasmid bearing both sgRNA scaffold backbone (BB) and Cas9, pSpCas9(BB). The resulting plasmid is annotated as pSpCas9(sgRNA). Completed and sequence-verified pSpCas9(sgRNA) plasmids and optional repair templates for facilitating HDR are then transfected into cells and assayed for their ability to mediate targeted cleavage. Finally, transfected cells can be clonally expanded to derive isogenic cell lines with defined mutations. © 20 13 N at ur e A m er ic a, In c. A ll rig ht s re se rv ed . protocol 2284 | VOL.8 NO.11 | 2013 | nature protocols over 100 bp long, compared with the ~20-bp-long oligos required for plasmid-based sgRNA delivery. Construction of an expression plasmid for sgRNA is also simple and rapid, involving a single cloning step with a pair of partially complementary oligonucleotides. The oligo pairs encoding the 20-nt guide sequences are annealed and ligated into a plasmid (pSpCas9(BB), Fig. 4c) bearing both Cas9 and the remainder of the sgRNA as an invariant scaffold immedi- ately following the oligo cloning site. The transfection plas- mids can also be modified to enable virus production for in vivo delivery. For these approaches, the following plasmids are used within this protocol: Cas9 alone (pSpCas9) or Cas9 with an invariant sgRNA scaffold and cloning sites for inserting a guide sequence (pSpCas9(BB)). For the backbone cloning con- struct, we have also fused 2A-GFP or 2A-Puro to Cas9 to allow screening or selection of transfected cells (pSpCas9(BB)-2A- GFP or pSpCas9(BB)-2A-Puro, respectively). Finally, we provide pSpCas9n(BB), a D10A nickase mutant of Cas9 for HDR and for double-nicking applications (Box 2), along with the 2A- GFP and 2A-Puro fusion constructs (pSpCas9n(BB)-2A-GFP, pSpCas9n(BB)-2A-Puro). In addition to PCR and plasmid-based delivery methods, Cas9 and sgRNAs can be introduced into cells as mRNA and RN