RNA-dependent DNA endonuclease Cas9 of the CRISPR system: Holy Grail of genome editing?

Opinion

RNA-dependent DNA endonuclease Cas9 of the CRISPR system: Holy Grail of genome editing? Giedrius Gasiunas and Virginijus Siksnys Institute of Biotechnology, Vilnius University, Graiciuno 8, LT-02241, Vilnius, Lithuania

Tailor-made nucleases for precise genome modification, such as zinc finger or TALE nucleases, currently represent the state-of-the-art for genome editing. These nucleases combine a programmable protein module which guides the enzyme to the target site with a nuclease domain which cuts DNA at the addressed site. Reprogramming of these nucleases to cut genomes at specific locations requires major protein engineering efforts. RNA-guided DNA endonuclease Cas9 of the type II (clustered regularly interspaced short palindromic repeat) CRISPR–Cas system uses CRISPR RNA (crRNA) as a guide to locate the DNA target and the Cas9 protein to cut DNA. Easy programmability of the Cas9 endonuclease using customizable RNAs brings unprecedented flexibility and versatility for targeted genome modification. We highlight the potential of the Cas9 RNA-guided DNA endonuclease as a novel tool for genome surgery, and discuss possible constraints and future prospects.

Engineered, highly specific DNA endonucleases (meganucleases) programmable according to the desired specificity are currently the state-of-the-art tools for gene editing technology. Meganucleases are by definition sequence-specific endonucleases with large (>12 bp) recognition sites [6]. Meganucleases combine a programmable specificity module which guides the enzyme to the target site with a nuclease domain which introduces cleavage at the addressed site. Different protein scaffolds are currently used for the development of meganucleases for gene targeting (Table 1): (i) homing endonucleases (HEases) [6–8]; (ii) zinc finger nucleases (ZFNs) [9]; (iii) TALE nucleases (TALENs) [10,11]; and (iv) restriction enzyme-triple helix forming oligonucleotide conjugates (RE-TFOs) [12,13]. Naturally occurring HEases or their engineered variants, which recognize long DNA sequences (up to 40 base

Gene editing Targeted genome editing technology that enables the generation of site-specific changes in the genomic DNA of cellular organisms is a Holy Grail for genome engineers [1–3]. Currently available genome editing technologies rely on the double-strand break (DSB) repair pathways of the cell. When a DSB occurs in DNA, it triggers a natural process of DNA repair either by an ‘error-prone’ non-homologous end joining (NHEJ; see Glossary) [4] or by homologous recombination (HR) [5]. Therefore, molecular tools that can generate DSBs at specific sites within the genome are at the core of current genome editing technologies. The ideal gene editing tool should meet the following criteria: (i) high frequency of desired sequence changes in the target cell population; (ii) no off-target cleavage; and (iii) rapid and efficient assembly of nucleases that target any site on the genome at low cost [1].

cas: CRISPR-associated genes which are located in the vicinity of CRISPR array and are necessary for the silencing of invading nucleic acid. Cas9t: Cas9–crRNA–tracrRNA ternary complex, which functions as an RNAguided DNA endonuclease and mediates site-specific DNA cleavage. Clustered regularly interspaced short palindromic repeat (CRISPR): an array of short conserved repeat sequences interspaced by unique DNA sequences of similar size called spacers, which often originate from phage or plasmid DNA. CRISPR array together with cas genes form the CRISPR–Cas system, which functions as an adaptive immune system in prokaryotes. CRISPR RNA (crRNA): small RNA molecule generated by transcription and processing of the CRISPR array. crRNA is composed of a conserved repeat fragment(s) and a variable spacer sequence, which matches the complimentary sequence in the invading nucleic acid. Homologous repair (HR): error-free DNA repair pathway that seals the broken DNA molecule using a homologous sequence (template). Non-homologous end joining (NHEJ): a pathway that repairs DNA DSBs in the absence of a homologous template; usually leads to small insertions or deletions. Protospacer adjacent motif (PAM): a short conserved nucleotide stretch located in the vicinity of a protospacer in the target DNA and necessary for DNA cleavage by Cas9t. Protospacer: a fragment in the target DNA, which matches a spacer sequence in the CRISPR array. Single guide RNA (sgRNA): RNA hairpin obtained by connecting crRNA and tracrRNA into a single molecule. Transcription activator-like effector nuclease (TALEN): an artificial nuclease obtained by fusing Xanthomonas transcription activator-like effector (TALE) DNA binding domains to the nonspecific nuclease domain. Trans-acting CRISPR RNA (tracrRNA): trans-encoded small RNA molecule, which forms a duplex with a repeat fragment of crRNA. Triple helix forming oligonucleotide (TFO): an artificial oligodeoxynucleotide, which binds to the polypurine sequences of the double-stranded DNA forming DNA triple helix. Zinc finger nuclease (ZFN): an artificial nuclease created by fusing zinc finger motifs, which serve as DNA recognition modules, to a nonspecific DNA cleavage domain of the FokI restriction endonuclease.

Corresponding author: Siksnys, V. ([email protected]). Keywords: molecular tools; gene targeting; RNA-guided DNA cleavage; clustered regularly interspaced short palindromic repeat (CRISPR). 0966-842X/$ – see front matter ß 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tim.2013.09.001

562

Trends in Microbiology, November 2013, Vol. 21, No. 11

Glossary

Opinion

Trends in Microbiology November 2013, Vol. 21, No. 11

Table 1. Tools for gene editing Tool

Specificity module

Cleavage module Nuclease domain FokI nuclease domain

Target site length, bp 14–40

HEase ZFN

Target recognition domain Zinc finger domains

TALEN

TALE domains

FokI nuclease domain

24–59

TFO conjugate

TFO (+restriction endonuclease)

Restriction endonuclease

4–8 + TFO

Cas9t

RNA (+ PAM)

Cas9

20 + PAM (2–5)

18–36

Reprogramming Complicated: requires protein engineering Complicated: requires domain shuffling, assembly, and protein engineering, from 10 weeks a Relatively easy: requires domain shuffling, assembly, and protein engineering, from 4 weeks b Relatively easy: requires only DNA oligo but includes a complicated chemistry step Easy and fast: requires only sgRNA

Targeting frequency Low

Specificity

Other features

Off-targeting reported Off-targeting reported

High cleavage efficiency Sequence bias, some variants show toxicity

High, nearly every sequence

Systematically not determined

Large protein size

Low, restricted by RE and TFO sequences

Systematically not determined

Slow equilibrium

High, depends on PAM

Off-targeting reported

Multiplexing possibilities

High

a

According to manufacturer’s information (http://www.sigmaaldrich.com/life-science/zinc-finger-nuclease-technology/custom-zfn.html).

b

According to manufacturer’s information (http://www.cellectis-bioresearch.com/products/talen-basic).

pairs), have been applied to gene editing in numerous experimental designs and cell types [6–8]. Artificial nucleases, such as ZFNs or TALENs, have been developed for the purpose of gene targeting. ZFNs are created by fusing zinc finger motifs, which serve as a DNA recognition module, to a nonspecific DNA cleavage domain of the FokI restriction endonuclease [9]. The combination of zinc finger modules specific for different sequences enables targeting of desired sites within the genome. TALENs are based on the fusion of the Xanthomonas transcription activator-like effector (TALE) DNA-binding domain (DBD) to the nuclease [10,11]. The TALE DBD contains repeated motifs that recognize specific nucleotides, and therefore the specificity of TALENs could be programmed by selecting and shuffling repeat segments specific for different nucleotides. Despite the differences in protein scaffolds, specificity of HEases, ZFNs, and TALENs is determined by the DBD of the meganuclease. In an alternative approach, site-specific genome cleavage reagents have been engineered using DNA triple helix forming oligonucleotides (TFOs) as address tags fused to various DNA damaging compounds [14,15] or restriction enzyme nuclease modules [12,13]. HEases, ZFNs, and TALENs are already commercially exploited as molecular tools for the generation of DSBs at specific sites within chromosomes. One major limitation is that HEases, ZFNs, and TALENs must be re-engineered for each new DNA target. Despite progress in the field, engineering of the DBDs of ZFNs and TALENs is time consuming, requires considerable skills, and often produces variants that are poised for off-target cleavage [1,16]. RE-TFO gene targeting technology employs DNA triple helix forming nucleotides to achieve binding specificity [12–15]. In principle, TFO as a specificity module has an advantage because reprogramming requires only oligonucleotide synthesis. However, several drawbacks limit the application of this technology. First, oligonucleotide conjugation to the nuclease module requires complicated chemistry steps [12,13]. Second, triple helix formation is slow and often requires modified oligonucleotides to

achieve desired target specificity [17]. Therefore, new genome editing tools, which are more affordable and easier and faster to engineer, are still in demand. RNA-dependent DNA endonuclease Cas9 of the CRISPR system CRISPR (clustered regularly interspaced short palindromic repeat) is a recently discovered bacterial and archaeal adaptive immune system, which consists of an array of short conserved repeat sequences interspaced by unique DNA sequences of similar size called spacers, which often originate from phage or plasmid DNA. CRISPR arrays together with cas (CRISPR-associated) genes form the CRISPR–Cas adaptive immune system [18–20]. CRISPR–Cas systems function by incorporating fragments of the invading nucleic acid as spacers into a host genome and later use these spacers as templates to generate small RNA molecules (crRNA) that are combined with Cas proteins into an effector complex which silences foreign nucleic acids in the subsequent rounds of infection. CRISPR–Cas systems are distinct and have been categorized into three main types, based on core element content and sequences [21]. The effector complex that binds crRNA and triggers cleavage of invading nucleic acid differs strikingly between different CRISPR subtypes. In type I systems, crRNAs are incorporated into a multi-subunit effector complex called Cascade (CRISPRassociated complex for antiviral defense), which binds to the target DNA and triggers degradation by an accessory Cas3 protein [22,23]. In type III CRISPR–Cas systems, exemplified by Sulfolobus solfataricus and Pyrococcus furiosus, the Cas RAMP module (Cmr) and crRNA complex recognize and cleave synthetic RNA in vitro [24,25]. Surprisingly, in type II CRISPR–Cas systems only the Cas9 protein (previously named Cas5 or Csn1) is required for DNA interference [19,26–28]. Cas9 is a large, multi-domain protein which contains two nuclease domains, an RuvC-like nuclease domain near the amino terminus and a HNH-like nuclease domain in the middle of the protein [29]. Cas9 forms a ternary complex (Cas9t) with two RNA molecules: crRNA 563

Opinion

Trends in Microbiology November 2013, Vol. 21, No. 11

(A)

Cas9

(B)

DNA target RuvC

DNA target RuvC

PAM

Cas9 mutant

HNH

PAM

HNH

Dual RNA or sgRNA

Dual RNA or sgRNA Dual RNA or sgRNA

Dual RNA or sgRNA

Double-strand break

Single-strand break NHEJ

HR

HR

+

+

TRENDS in Microbiology

Figure 1. Genome editing with the Cas9–dual RNA (–sgRNA) complex. (A) The Cas9 complex guided by a dual tracrRNA–crRNA or mimicking sgRNA (single guide RNA where tracrRNA and crRNA are connected by a hairpin shown as a dotted line) locates the DNA target, and if a correct PAM sequence is present binds to the target site forming an R-loop. Cas9 cuts both target and non-target DNA strands in the R-loop using the HNH and the RuvC active sites, respectively. Cleavage occurs 3 nt from PAM to produce a blunt-ended DNA double-strand break. The double-strand break is repaired by an ‘error-prone’ non-homologous end joining (NHEJ) or by homologous recombination (HR) if a template DNA is present. (B) Cas9 HNH or RuvC mutants (Cas9m) act as nicking enzymes, which introduce single-strand breaks in the target or nontarget strands, as exemplified by the Cas9 RuvC mutant. The single-strand breaks are repaired by HR in the presence of the DNA template. Abbreviations: tracrRNA, transactivating CRISPR RNA; crRNA, CRISPR (clustered regularly interspaced short palindromic repeat) RNA; PAM, protospacer adjacent motif; nt, nucleotide.

(CRISPR RNA) and tracrRNA (trans-activating CRISPR RNA), which is partially complementary to crRNA [27,30–33]. The 42-nucleotide (nt) crRNA consists of the 22-nt 30 handle originating from the repeat sequence and the 20-nt unique spacer fragment, which guides Cas9t to the matching target DNA. tracrRNA is required for crRNA maturation and cleavage by Cas9t [31,32]. Importantly, a dual tracrRNA–crRNA engineered as a single guide RNA (sgRNA) chimera supports Cas9-mediated double-strand DNA cleavage [31]. The Cas9t binding site in the target DNA is composite and consists of a nucleotide stretch matching crRNA, called a protospacer, and of a short nucleotide sequence, called a protospacer adjacent motif (PAM), in the vicinity of the protospacer (Figure 1A). If both the protospacer and PAM are present, Cas9t binds to the target site generating the R-loop, where one DNA strand is displaced and another is engaged in a heteroduplex with crRNA. The cleavage of the DNA target is executed by two separate active sites within Cas9: RuvC cuts the displaced DNA strand, whereas HNH cuts the DNA strand paired with RNA (Figure 1A) [30,31]. Thus, the Cas9t complex functions as an RNA-guided DNA endonuclease that uses crRNA for target site recognition and Cas9 for DNA cleavage. Cas9: a new tool in a genome editing toolbox Simple modular organization of the Cas9–crRNA– tracrRNA complex, in which the specificity of DNA targets is encoded by crRNA and cleavage machinery consists of a single multi-domain protein, provides a universal platform for engineering of an easy reprogrammable RNA-guided DNA endonuclease [30,31,34]. Not surprisingly, Cas9t has 564

been immediately employed as a tool for genome surgery in bacteria and various eukaryotic cells including human, mouse, zebrafish, Drosophila, worm, plant, and yeast model systems (Table 2). Cas9t-based technology brings several advantages in comparison to ZFNs and TALENs. First, the reprogramming of the Cas9 protein is achieved by using customized RNAs (dual RNAs or sgRNAs); therefore, it is fast and cheap in comparison to other technologies. Indeed, ZFN and TALEN reprogramming requires cloning, protein engineering, selection, and validation, and it takes a minimum of 10 and 4 weeks, respectively (http://www.sigmaaldrich. com/life-science/zinc-finger-nuclease-technology/customzfn.html; http://www.cellectis-bioresearch.com/products/ talen-basic). Furthermore, the Cas9–sgRNA platform enables targeting of two or more targets in one cell (multiplexing), using a single Cas9 protein and multiple RNAs. Multiplexing considerably shortens the time required for generation of transgenic animals containing mutations in multiple genes [35–37]. In addition, two closely situated sites might be targeted to increase the efficiency of gene inactivation or to introduce a programmed deletion. Second, the availability of Cas9 nicking mutants that can introduce nicks either in top or bottom strands of the target [30,31] opens the possibility for DNA repair by HR (Figure 1B) [36,37]. Repair of a single-stranded break by the error-free HR at the nicking site results in gene correction without any footprint, in contrast to an error-prone NHEJ, which frequently creates undesired insertions and deletions [36–38]. The efficiency of desired sequence changes in the target cell population introduced by Cas9t is comparable to that of ZFNs and TALENs (Table 2) [39].

Opinion

Trends in Microbiology November 2013, Vol. 21, No. 11

Table 2. Examples of gene editing by Cas9t Organism Homo sapiens Homo sapiens Homo sapiens Homo sapiens Mus musculus Danio rerio Mus musculus Saccharomyces cerevisiae Caenorhabditis elegans Drosophila melanogaster Arabidopsis thaliana Nicotiana benthamiana Rice (Oryza sativa) Common wheat (Triticum aestivum) Bacteria

Cell line H293K K562 PGP1 iPS H293FT Neuro2A Embryo Embryo BY4733 Embryo Embryo – – – Streptococcus pneumoniae and Escherichia coli

Targeted gene AAVS1 AAVS1 AAVS1 EMX1, PPVALB, EMX1 Th etsrp, gata4, or gata5 Tet1, Tet2, Tet3, Sry, Uty CAN1 unc-119, dpy-13, klp-12a, Y61A9LA.1a y1, y2, w1, w2 AtPDS3, AtFLS2, AtRACK1b, and AtRACK1c NbPDS3 OsPDS, OsBADH2, Os02g23823, OsMPK2 TaMLO bgaA, srtA, ermAM, rpsL

Future questions to be addressed The demonstration that the reprogrammable Cas9t endonuclease is a promising tool for precise gene surgery in different model systems brought much excitement into a genome editing field. However, to find out whether the Cas9t endonuclease is the real Holy Grail in the gene editing field, several important issues still have to be addressed. Target site selection The ideal gene correction tool should be easily programmable to address any desired sequence within the genome. In theory, crRNA could be engineered to guide the Cas9t complex to any DNA target complementary to the crRNA. However, in reality, crRNA-guided target selection is restricted by the PAM sequence, which is absolutely required for target DNA binding and cleavage by Cas9t (Table 3). For Streptococcus pyogenes, the Cas9t PAM is NGG; therefore, possible targets are located 1 per 8 bp. The PAM sequences for Streptococcus thermophilus Cas9 proteins are longer, NGGNG and NNAGAAW, respectively. In this case, the possible target frequency becomes 1 per 32 bp and 1 per 256 bp, accordingly, if the GC content of the genome is 50%. However, because systematic PAM analysis shows that the S. pyogenes Cas9 protein in bacteria may also target the NAG PAM, although with lower efficiency, the range for possible target selection may be expanded [40]. Other factors could also affect Cas9t target site selection. S. pyogenes Cas9-mediated cleavage seems to be unaffected by DNA methylation both in vivo and in vitro [41];

Repair pathway HR/NHEJ NHEJ HR/NHEJ HR/NHEJ NHEJ HR/NHEJ HR/NHEJ HR/NHEJ NHEJ NHEJ NHEJ HR/NHEJ HR/NHEJ NHEJ HR/NHEJ

Efficiency, % 10–25 8–13 2–4 Up to 21 Up to 27 Up to 25 Up to 48 Near 100 Up to 80 Up to 88 Up to 5.6 Up to 38.5 Up to 38 28.5 Up to 100

Refs [37] [37] [37] [36] [36] [48] [35] [49] [50] [51] [52] [52] [53] [53] [40]

however, it still has to be established whether a chromatin structure modulates Cas9t cleavage within the cell. Furthermore, because targeting of the Cas9–RNA complex requires hybridization of the crRNA to the target strand, secondary structures in crRNA or crRNA misfolding may affect complex assembly and cleavage efficiency. Off-target cleavage The specificity of Cas9t and off-target cleavage is a key question for gene editing technology. Off-targeting could trigger genomic arrangements or unpredicted mutations, leading to cell death or transformation [39]. Cas9t variants characterized to date (Table 3) recognize 20- or 24-nt sequences programmed by crRNA plus 2- to 5-nt PAM sequences required for Cas9t binding and cleavage. Thus, in theory the specificity of Cas9t covers 22–28-nt sequences, which is unique in most genomes. Unfortunately, Cas9t often tolerates mismatches at certain positions of the DNA target resulting in relaxed specificity [26,31,36,40–43]. For example, according to in vitro analysis, S. pyogenes Cas9t (Sp-Cas9t) can tolerate up to six mismatches [31]. In fact, target positions at the PAM distal end are more prone to mismatch tolerance, whereas mismatches at the PAM proximal 8–12-nt seed sequence are usually not allowed. The mismatch tolerance may restrict Cas9 specificity to shorter sequences determined by seed and PAM sequences. In this case, the availability of Cas9 homologs which recognize different PAMs [33] may be helpful for expanding Cas9t specificity. The longer PAM increases Cas9 specificity; however, at the same time it

Table 3. Features of the biochemically characterized Cas9t complexes of type II CRISPR–Cas systems Acronym St1-Cas9 St3-Cas9 Sp-Cas9 Nm-Cas9

Organism Streptococcus thermophilus DGCC7710 Streptococcus thermophilus DGCC7710 Streptococcus pyogenes SF370 Neisseria meningitidis 8013

Loci CRISPR1

Complex composition nd a

crRNA, nt 42

sgRNA nd

PAM (50 -30 ) NNAGAAW

Cleavage position 3 nt upstream PAM

Refs [27,28]

CRISPR3

Cas9–crRNA–tracrRNA

42

+b

NGGNG

3 nt upstream PAM

[30,32]

CRISPR01 –

Cas9–crRNA–tracrRNA Cas9–crRNA–tracrRNA

42 48

+b +b

NGG NNNAGAA

3 nt upstream PAM 3 nt upstream PAM

[27,31] [54,55]

a

nd, not determined.

b

The protein is active with a single guide RNA (sgRNA).

565

Opinion Box 1. Outstanding questions    

Can Cas9 be targeted to any desired DNA sequence in the genome? What is the role of the chromatin state on Cas9 cleavage? How is the PAM recognized by the Cas9 complex? How could Cas9 specificity be improved and off-target cleavage minimized?

restricts the frequency of target selection. Alternatively, establishing a molecular mechanism of PAM recognition may open possibilities to engineer Cas9 variants with a designed PAM sequence. Systematic analysis of Sp-Cas9t targeting specificity in vivo revealed that Sp-Cas9t tolerates up to four mutations relative to the on-target site depending on the locus, base position and identity, and Cas9/sgRNA dosage [41–44]. In this case, a computer-guided prediction of a potential offtarget activity in the design of sgRNA guide sequences and the Cas9t dosage control could minimize off-target cleavage [41,43]. Alternatively, engineering of a pair of Cas9t nicking variants that will require cooperativity to generate a DSB could mitigate off-target cleavage [44]. Further exploration of Cas9 mutants/orthologs with improved specificity and systematic comparative analysis of off-target cleavage activity of ZFNs, TALENs, and Cas9t should reveal whether a Holy Grail quest in the gene targeting field is over or it has to continue. Concluding remarks Despite questions that yet have to be addressed (Box 1), Cas9t emerges as a promising gene editing tool. It represents a completely new family of DNA endonucleases which use RNA molecules for target recognition. General interest in a reprogrammable Cas9 enzyme for gene editing is extremely high, as can be judged from an impressive list of papers that appeared in 2013 in top quality journals. We still need more data to see whether Cas9 is indeed a game changer in the genome editing field (http://www. forbes.com/sites/matthewherper/2013/03/19/the-proteinthat-could-change-biotech-forever/). Recently, iCRISPR technology [45,46] that employs a catalytically defective Cas9 mutant has been developed for gene regulation in vivo, and more Cas9 applications are probably ahead. Indeed, it has turned out that some pathogenic bacteria use the Cas9–scaRNA (small CRISPR–Cas associated RNA) complex to stay infectious by shutting down a gene important for the host immune response [47]. This finding may open novel avenues for Cas9 applications. Acknowledgments We thank Rodolphe Barrangou and Philippe Horvath for discussions. The work on CRISPR–Cas systems in Siksnys’ laboratory is funded by the European Social Fund under the Global Grant measure.

References 1 Perez-Pinera, P. et al. (2012) Advances in targeted genome editing. Curr. Opin. Chem. Biol. 16, 268–277 2 Carroll, D. (2011) Genome engineering with zinc-finger nucleases. Genetics 188, 773–782 3 Urnov, F.D. et al. (2010) Genome editing with engineered zinc finger nucleases. Nat. Rev. Genet. 11, 636–646 4 Lieber, M.R. (2010) The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu. Rev. Biochem. 79, 181–211 566

Trends in Microbiology November 2013, Vol. 21, No. 11

5 Moynahan, M.E. and Jasin, M. (2010) Mitotic homologous recombination maintains genomic stability and suppresses tumorigenesis. Nat. Rev. Mol. Cell Biol. 11, 196–207 6 Paˆques, F. and Duchateau, P. (2007) Meganucleases and DNA doublestrand break-induced recombination: perspectives for gene therapy. Curr. Gene Ther. 7, 49–66 7 Arnould, S. et al. (2006) Engineering of large numbers of highly specific homing endonucleases that induce recombination on novel DNA targets. J. Mol. Biol. 355, 443–458 8 Grizot, S. et al. (2010) Generation of redesigned homing endonucleases comprising DNA-binding domains derived from two different scaffolds. Nucleic Acids Res. 38, 2006–2018 9 Kim, Y.G. et al. (1996) Hybrid restriction enzymes: zinc finger fusions to Fok I cleavage domain. Proc. Natl. Acad. Sci. U.S.A. 93, 1156–1160 10 Christian, M. et al. (2010) Targeting DNA double-strand breaks with TAL effector nucleases. Genetics 186, 757–761 11 Li, T. et al. (2011) TAL nucleases (TALNs): hybrid proteins composed of TAL effectors and FokI DNA-cleavage domain. Nucleic Acids Res. 39, 359–372 12 Eisenschmidt, K. et al. (2005) Developing a programmed restriction endonuclease for highly specific DNA cleavage. Nucleic Acids Res. 33, 7039–7047 13 Silanskas, A. et al. (2012) Catalytic activity control of restriction endonuclease–triplex forming oligonucleotide conjugates. Bioconjug. Chem. 23, 203–211 14 Arimondo, P.B. et al. (2006) Exploring the cellular activity of camptothecin–triple-helix-forming oligonucleotide conjugates. Mol. Cell. Biol. 26, 324–333 15 Majumdar, A. et al. (2008) Targeted gene knock in and sequence modulation mediated by a psoralen-linked triplex-forming oligonucleotide. J. Biol. Chem. 283, 11244–11252 16 Gaj, T. et al. (2013) ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering. Trends Biotechnol. 31, 397–405 17 Rusling, D. et al. (2005) Combining nucleoside analogues to achieve recognition of oligopurine tracts by triplex-forming oligonucleotides at physiological pH. FEBS Lett. 579, 6616–6620 18 Horvath, P. and Barrangou, R. (2010) CRISPR/Cas, the immune system of bacteria and archaea. Science 327, 167–170 19 Barrangou, R. et al. (2007) CRISPR provides acquired resistance against viruses in prokaryotes. Science 315, 1709–1712 20 Gasiunas, G. et al. (2013) Molecular mechanisms of CRISPR-mediated microbial immunity. Cell Mol. Life Sci. http://dx.doi.org/10.1007/ s00018-013-1438-6 21 Makarova, K.S. et al. (2011) Evolution and classification of the CRISPR–Cas systems. Nat. Rev. Microbiol. 9, 467–477 22 Brouns, S.J.J. et al. (2008) Small CRISPR RNAs guide antiviral defense in prokaryotes. Science 321, 960–964 23 Sinkunas, T. et al. (2013) In vitro reconstitution of Cascade-mediated CRISPR immunity in Streptococcus thermophilus. EMBO J. 32, 385–394 24 Hale, C.R. et al. (2009) RNA-guided RNA cleavage by a CRISPR RNA– Cas protein complex. Cell 139, 945–956 25 Zhang, J. et al. (2012) Structure and mechanism of the CMR complex for CRISPR-mediated antiviral immunity. Mol. Cell 45, 303–313 26 Sapranauskas, R. et al. (2011) The Streptococcus thermophilus CRISPR/Cas system provides immunity in Escherichia coli. Nucleic Acids Res. 39, 9275–9282 27 Deltcheva, E. et al. (2011) CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III. Nature 471, 602–607 28 Garneau, J.E. et al. (2010) The CRISPR/Cas bacterial immune system cleaves bacteriophage and plasmid DNA. Nature 468, 67–71 29 Makarova, K.S. et al. (2006) A putative RNA-interference-based immune system in prokaryotes: computational analysis of the predicted enzymatic machinery, functional analogies with eukaryotic RNAi, and hypothetical mechanisms of action. Biol. Direct 1, 7 30 Gasiunas, G. et al. (2012) Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria. Proc. Natl. Acad. Sci. U.S.A. 109, E2579–E2586 31 Jinek, M. et al. (2012) A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337, 816–821 32 Karvelis, T. et al. (2013) crRNA and tracrRNA guide Cas9-mediated DNA interference in Streptococcus thermophilus. RNA Biol. 10, 841–851 33 Chylinski, K. et al. (2013) The tracrRNA and Cas9 families of type II CRISPR–Cas immunity systems. RNA Biol. 10, 726–737

Opinion 34 Barrangou, R. (2012) RNA-mediated programmable DNA cleavage. Nat. Biotechnol. 30, 836–838 35 Wang, H. et al. (2013) One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell 153, 910–918 36 Cong, L. et al. (2013) Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823 37 Mali, P. et al. (2013) RNA-guided human genome engineering via Cas9. Science 339, 823–826 38 Kim, E. et al. (2012) Precision genome engineering with programmable DNA-nicking enzymes. Genome Res. 22, 1327–1333 39 Mussolino, C. and Cathomen, T. (2013) RNA guides genome engineering. Nat. Biotechnol. 31, 208–209 40 Jiang, W. et al. (2013) RNA-guided editing of bacterial genomes using CRISPR–Cas systems. Nat. Biotechnol. 31, 233–239 41 Hsu, P.D. et al. (2013) DNA targeting specificity of RNA-guided Cas9 nucleases. Nat. Biotechnol. http://dx.doi.org/10.1038/nbt.2647 42 Fu, Y. et al. (2013) High-frequency off-target mutagenesis induced by CRISPR–Cas nucleases in human cells. Nat. Biotechnol. http:// dx.doi.org/10.1038/nbt.2623 43 Pattanayak, V. et al. (2013) High-throughput profiling of off-target DNA cleavage reveals RNA-programmed Cas9 nuclease specificity. Nat. Biotechnol. http://dx.doi.org/10.1038/nbt.2673 44 Mali, P. et al. (2013) CAS9 transcriptional activators for target specificity screening and paired nickases for cooperative genome engineering. Nat. Biotechnol. http://dx.doi.org/10.1038/nbt.2675

Trends in Microbiology November 2013, Vol. 21, No. 11

45 Qi, L.S. et al. (2013) Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression. Cell 152, 1173–1183 46 Gilbert, L.A. et al. (2013) CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes. Cell 154, 442–451 47 Sampson, T.R. et al. (2013) A CRISPR/Cas system mediates bacterial innate immune evasion and virulence. Nature 497, 254–257 48 Chang, N. et al. (2013) Genome editing with RNA-guided Cas9 nuclease in zebrafish embryos. Cell Res. 23, 465–472 49 Dicarlo, J.E. et al. (2013) Genome engineering in Saccharomyces cerevisiae using CRISPR–Cas systems. Nucleic Acids Res. 41, 4336–4343 50 Friedland, A.E. et al. (2013) Heritable genome editing in C. elegans via a CRISPR–Cas9 system. Nat. Methods 10, 741–743 51 Bassett, A.R. et al. (2013) Highly efficient targeted mutagenesis of Drosophila with the CRISPR/Cas9 system. Cell Rep. 4, 220–228 52 Li, J.-F. et al. (2013) Multiplex and homologous recombinationmediated genome editing in Arabidopsis and Nicotiana benthamiana using guide RNA and Cas9. Nat. Biotechnol. 31, 688–691 53 Shan, Q. et al. (2013) Targeted genome modification of crop plants using a CRISPR–Cas system. Nat. Biotechnol. 31, 686–688 54 Hou, Z. et al. (2013) Efficient genome engineering in human pluripotent stem cells using Cas9 from Neisseria meningitidis. Proc. Natl. Acad. Sci. U.S.A. http://dx.doi.org/10.1073/pnas.1313587110 55 Zhang, Y. et al. (2013) Processing-independent CRISPR RNAs limit natural transformation in Neisseria meningitidis. Mol. Cell 50, 488–503

567