Emerging CRISPR Technologies

Abstract

The discovery and implementation of Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR associated (Cas) systems for genome editing has revolutionized biomedical research and holds great promise for the treatment of human genetic disorders. In addition to the popular CRISPR-Cas9 and CRISPR-Cpf1 systems for genome editing, several additional Class I and Class 2 CRISPR-Cas effectors have been identified and adapted for genome editing and transcriptome modulation. Here we discuss current and emerging CRISPR-based technologies such as Cascade-Cas3, CRISPR-associated transposases (CAST), CRISPR-Cas7–11, and CRISPR-Cas13 for genome and transcriptome modification. These technologies allow for the removal or insertion of large DNA elements, the modulation of gene expression at the transcriptional level, and the editing of RNA transcripts, expanding the capabilities of current technologies.

1. Introduction

Since the discovery of the double helix, scientists have been searching for ways to manipulate genomes. Over the past 15 years, technological advances such as the development of targetable nucleases finally provided a means for introducing specific alterations within a genome of interest. Targetable nucleases function by introducing a DNA double strand break (DSB) at a precise location within a genome which in turn activates cellular DNA repair pathways. By hijacking these pathways, via the coadministration of DNA repair templates, a plethora of genetic modifications ranging from single nucleotide substitutions to chromosomal translocations can be engineered (Figure 1).

Figure 1.

Genetic modifications engineered using conventional CRISPR-Cas systems. Schematic representation of the different types of alleles that can be engineered with a single or a pair of cut sites and the engineering systems used to create them. With one cut site, three types of alleles can be engineered: Targeted random insertion or deletion of genetic material (indels), nucleotide substitutions, and insertion of DNA elements. With two cut sites, three other types of alleles can be engineered: Deletion of DNA elements, inversions, and chromosomal translocations. Introns are shown as light gray boxes, exons as dark gray boxes, start sites as black arrows. Indels are shown as a red box, nucleotide substitutions as a blue box, and insertion of DNA elements as a green box. Black arrows pointing to a specific location within an intron or exon indicate cut sites.

The first implementations of targetable nucleases included zinc finger nucleases (ZFNs) and transcription activator-like effector (TALE) nucleases (TALENs). These enzymes are formed by the combination of a non-specific DNA endonuclease called FokI and DNA binding protein domains derived from the zinc finger or TALE family of transcription factors. These enzymes function as obligate dimers and rely on protein-DNA pairing for target recognition [1, 2]. While these enzymes provide the specificity needed for engineering precise DNA alterations, their programming or reprogramming necessitates the design and synthesis of a new pair of enzymes for each new alteration. The adaptation of the Class 2 type II CRISPR-Cas9 system from Streptococcus pyogenes (CRISPR-SpCas9) for genome editing has drastically changed the way genetic engineering is performed in that it provides the long-awaited simplicity and versatility required for engineering precise DNA alterations. Rather than relying on protein-DNA interactions for target recognition, CRISPR-SpCas9 relies on RNA–DNA base pairing. The simple modification of a short RNA transcript is sufficient to reprogram a nonspecific endonuclease to target other sites.

In bacteria and archaea, CRISPR-Cas systems are RNA-based immune systems that control virus and plasmid invasion [3]. CRISPR-Cas systems are taxonomically classified as Class 1 and Class 2 systems based on the number of components involved in the interference stage of the immune response. With rare exceptions, Class 1 systems, which account for approximately 90% of all CRISPR-Cas systems in prokaryotes, use multiprotein effector complexes whereas Class 2 systems use a single effector. Class 1 and Class 2 systems are further divided, based on signature genes and distinctive gene architectures, into three or more types: Type I, III and IV for Class 1 systems and type II, V and VI for Class 2 systems.

Concurrent with the implementation of CRISPR-Cas9 for genome editing, a variety of Class 1 and Class 2 systems with complementary properties to type II effector Cas9 have been identified and adapted for genome and transcriptome editing. These systems include: Class 1 type I Cascade-Cas3 systems [4, 5, 6], which are RNA-guided DNA shredding systems; Class 2 type V-K effector Cas12k-Tn7-like transposase systems [7], which are RNA-guided DNA transposition systems that allow for unidirectional insertion of large DNA cargos; and Class 2 type VI CRISPR-Cas13 [8, 9] and Class 1 Type III-E effector Cas7–11 systems [10, 11], which are RNA-guided RNA targeting systems.

In this chapter, we describe conventional as well as emerging CRISPR-Cas-based technologies for transcriptome and genome editing. We provide a simplified view of these systems and their operons. When applicable, we describe the modifications made for genome editing in mammalian cells, the sequence and structure of guide RNAs, PAM requirements, and examples of their use for genome and transcriptome editing. For simplicity, we focus on CRISPR-Cas9 from S. pyogenes [12, 13, 14], CRISPR-Cpf1 from Francisella novicida [15], Cascade-Cas3 from Thermobifida fusca [5], CAST from Scytonema hofmannii [7, 16], CRISPR-Cas13 from Leptotrichia shahii [9], and Cas7–11 from Candidatus Scalindua brodae [17] as prototypic systems.

2. Conventional CRISPR-Cas systems for genome editing

Genome editing using conventional CRISPR-Cas systems functions by introducing DNA DSBs at a precise location within a target genome. These breaks, known to be highly recombinogenic, are typically repaired via the nonhomologous end joining (NHEJ) DNA DSB repair pathway and result in the random insertion or deletion of genetic material, often referred to as indels. In actively dividing cells, homologous recombination (HR) can also occur and uses the sister chromatid as repair template, resulting in error free repair of the break. By providing an exogenous repair template, in the form of single or double stranded DNA molecules, a variety of genetic alterations can be engineered, including nucleotide substitutions, insertion of DNA elements, deletion of DNA material, inversion of DNA elements, as well as chromosomal translocations (Figure 1). Two main CRISPR-Cas systems from various species, CRISPR-Cas9 and CRISPR-Cpf1, have been adapted for genome editing and other applications.

2.1 CRISPR-Cas9

The most frequently used CRISPR-Cas system for genome editing in mammalian cells has been derived from the Class 2 type II CRISPR-Cas system from S. pyogenes (Figure 2A) [12, 13, 14]. In this system, a large endonuclease named SpCas9 pairs with a single guide RNA (sgRNA) to target the ribonucleoprotein complex to a specific location within a genome via the formation of an RNA–DNA duplex according to Watson-Crick base pairing (Figure 2B). sgRNAs are 96 nucleotide-long RNA transcripts formed from the fusion of a CRISPR-RNA (crRNA), which provides target specificity, and the trans-acting crRNA (tracrRNA), which bridges the crRNA to the endonuclease SpCas9. The first 20 nucleotides of a sgRNA provides target specificity whereas the other 76 nucleotides contain sequences encoding the tracrRNA (Figure 2C). Binding of SpCas9 to its target site results in the formation of an R-loop, a three-stranded nucleic acid structure composed of a DNA–RNA hybrid and unbound DNA strand referred to as the protospacer element or target site (Figure 2B). Following the R-loop formation, Cas9 – which possesses a RuvC-like and an HNH-like domain – cleaves both DNA strands 3 nucleotides upstream of the protospacer adjacent motif (PAM). PAM sequences are short genomic sequences located at the 3′ end of the protospacer element (for Cas9 nucleases) that must be present for the nuclease to be catalytically active. Although this requirement was initially viewed as a major limitation to the application of CRISPR-SpCas9 for genome editing, its PAM sequence, 5′-NGG-3′, is found approximately every 8 nucleotides in the human genome. Moreover, several CRISPR-Cas9 systems from various bacterial species have been identified, each having distinct PAM requirements (Table 1) [12], expanding the scope of these systems. For genome editing in mammalian cells, the open reading frame of SpCas9 has been optimized for codon usage and nuclear localization signals (NLSs) have also been added.

Figure 2.

CRISPR-Cas9. A) Schematic representation of the CRISPR-Cas9 operon from Streptococcus pyogenes. The operon contains 4 genes, three of which are involved in the adaptation stage of the immune response (Cas1, Cas2 and Csn2) and the maturation and interference stage of the response (Cas9). The locus also contains a tracrRNA and the CRISPR array which comprises repeat elements (gray rectangles) regularly interspaced with spacer elements (colored circles). B) Schematic representation of the CRISPR-SpCas9 system for genome editing and modifications made for its use in mammalian cells. The target genomic DNA is shown in black, with the PAM sequence (5′-NGG-3′) highlighted in green. The crRNA is shown in orange and the tracrRNA is shown in red. Fusion of the crRNA and tracrRNA forms the sgRNA, which is also shown in orange. The RuvC-like and HNH-like activity sites are shown by a black or a blue triangle, respectively. C) Sequence of the CRISPR-spCas9 sgRNA. The SpCas9 sgRNA contains sequences matching the protospacer element followed (N) by a hairpin loop linking the crRNA to the tracrRNA. D-K) schematic representation of various CRISPR_SpCas9 modalities including base editors (D-F), prime editors (G), target specific transcriptional activators or transcriptional repressors (H, I), as well as DNA and RNA tracking devices (J, K). D-F) base editors are formed from the fusion between catalytically impaired SpCas9 and base modifying enzymes such as CD which allows for the conversion of cytidine into a thymine (D), TadA-TadA* heterodimer which allows for the conversion of adenine to guanine (E) and eUNG and APOBEC1 fusion which allows for the transversion of cytidine to guanine (F). G) Prime editors are produced by the combination of catalytically impaired SpCas9 and a reverse transcriptase (RT). Prime editors use a modified sgRNA called pegRNA that not only contains sequences providing target specifificty and Cas9 scaffolding, but also sequences complementary to the target sites and the substitutions to be engineered. Following the cleavage of the protospacer element, a 3′ DNA flap is exposed, which allows the binding of the pegRNA complementary segment to bind target site and serves as template for th RT. the RT will extend the 3′ flap, introducing the designed mutations. H, I) transcriptional modulators usually make use of transcriptional activators such as the viral transcription factor VP64. Transcriptional repressors typically make use of transcriptional repressors or epigenome modifiers such as Kruppel associated box (KRAB) MECP2 or DNMT3a. K) DNA tracking devices are formed from the fusion of a catalytically inactive SpCas9 fused to a fuuorescent marker. Tiling of multiple of these SpCAs9 fusions allows for the visualization of genomic loci in real time imaging. L) RNA tracking devices are formed form the coupling of a catalytically inactive SpCas9 to a fluorescent protein marker. Binding of SpCas9 with a target RNA transcript is promoted by the coadministration of a short oligonucleotide called PAMer that acts as exogenous PAM sequence.

System	Bacterial strain	PAM or PFS
Cas9	Streptococcus pyogenes	NGG
	S. pyogenes (VQR)	NGAG
	S. pyogenes (VRER)	NGCG
	Streptococcus mutans	NGG
	Staphylococcus aureus	NNGGGT; NNGAAT; NNGAGT
	Streptococcus thermophilus (CRISPR3)	NGGNG
	S. thermophilus (CRISPR1)	NNAAAAW
	Campylobacter jejuni	NNNNACA
	Neisseria meningitidis	NNNNGATT
	Pasteurella multocida	GNNNCNNA
	Francisella novicida	NG
	Treponema denticola	NAAAAN
Cpf1	Franciselle novicida	TTN
	Acidaminococcus sp. BV3L6	TTTN
	Moraxella bovoculi 237	(T/C)(T/C)N
Cascade-Cas3	Thermobifida fusca	AAG
	Escherichia coli	ARG
	Pseudomonas aeruginosa	AAG
CRISPR-Cas12k	Scytonema hofmannii (ShCAST)	NGTN
	Anabaena cylindrica (AcCAST)	NGTN
Cascade-Tn6677	Vibrio Clolerae (Tn6677)	CC
CRISPR-Cas13	Leptotrichia shahii	A,U,C (not G)
CRISPR- Cas7–11	Scalindua brodae	N/A

Table 1.

PAM or PFS requirements for various CRISPR-Cas systems.

PAM, protospacer adjacent motif; PFS, protospacer flanking site.

CRISPR-SpCas9 systems have also been developed to introduce a limited number of nucleotide substitutions without introducing DNA DSBs. These systems, called Base Editors and Prime Editors, make use of catalytically impaired SpCas9 fused to various base modifying enzymes like a cytidine deaminase (CD), an uracil DNA N-glycosylase (eUNG), or a modified dimeric tRNA adenine deaminase (TadA*), which catalyze base transversion or base conversion within a precise window upstream of the PAM sequence (Figure 2D-K). The selection of a base editor depends on several criteria including the desired edit, the availability of PAM sequences within the target sequence, the position of the target nucleotide relative to the PAM sequence, the possibility of engineering undesired bystander mutations, and the need to minimize off-target editing [18]. The most recent versions of these systems are BE4max, a cytidine base editor which catalyzes the conversion of a cytidine into a thymine (C- > T) (Figure 2D); BE7.10, an adenine base editor which catalyzes the conversion of adenine to guanine (A- > G) (Figure 2E); CGBE1, a base editor that catalyzes cytidine to guanine (C- > G) base transversion (Figure 2F) [19, 20, 21].

Prime editing, on the other hand, makes use of a catalytically impaired SpCas9 fused to a reverse transcriptase (RT) and a prime editing guide RNA (pegRNA) (Figure 2G) [22]. The pegRNA not only provides target specificity and scaffolding, but also contains sequences that are complementary to the target site and substitutions encoding the desired edits. Following excision of the target strand, a 3′ flap is exposed and the pegRNA complexes with the exposed 3′ flap and serves as primer site for the RT, which extends the 3′ flap and incorporates the desired nucleotide substitutions. Stabilization of the locus is performed by the endogenous endonuclease FEN1 which removes the 5′ flap and allows the hybridization of the edited 3′ flap, resulting in the incorporation of edited bases and conversion of the unmodified allele via the DNA mismatch repair (MMR) pathway.

In addition to providing the framework for various genome editing technologies, CRISPR-SpCas9 has been morphed into DNA and RNA imaging devices [23, 24], epigenetic modifiers [25] as well as transcriptional modulators [26, 27, 28] via the fusion between catalytically inactive but sgRNA competent SpCas9 and transcriptional activators, transcriptional repressors, epigenetic modifiers, fluorescent proteins, and others (Figure 2H-K).

2.2 CRISPR-Cpf1

Another popular system for genome editing in mammalian cells has been derived from the Class 2 type V CRISPR-Cpf1 system from F. novicida (CRISPR-FnCpf1) (Figure 3A) [15]. In this system, a large endonuclease called FnCpf1 (or Cas12a) pairs with a 42 nucleotide-long crRNA that contains sequences providing both target specificity and FnCpf1 binding activity (Figure 3B and C). Unlike SpCas9, FnCpf1 contains two RuvC-like activities and introduces scattered DNA DSBs outside of its recognition sequence. More specifically, DNA breaks occur at positions 18 of the non-target strand and 23 of the target strand, leaving a 5′ overhang (Figure 3B). For FnCpf1 to be active, the endonuclease must also recognize a short PAM sequence (5′-TTN-3′) located at the 5′ end of the target sequence (or protospacer element). Similar to Cas9 systems, a variety of Cpf1 systems from diverse bacterial species with distinct PAM requirements have been identified, further expanding targeting possibilities. These include CRSIPR-Cpf1 systems from Acidaminococcus sp. BV3L6, Lachnospiraceae bacterium MA2020, and Moraxella bovoculi 237, among others [15]. Like CRISPR-SpCas9, CRISPR-FnCpf1 has also been converted into RNA-guided epigenetic modifying devices, transcriptional regulators and base editors (Figure 3D-F) [29, 30, 31, 32, 33].

Figure 3.

CRISPR-Cpf1. A) Schematic representation of the CRISPR-Cpf1 operon from Francisella novicida. B) Schematic representation of the CRISPR-FnCpf1 system and the modifications for its usage in mammalian cells. The target genomic DNA is shown in black, with the PAM sequence (5′-TTN-3′) highlighted in green. The crRNA is shown in red, and the two RuvC-like activity sites are shown by black triangles. C) Sequence of the CRISPR-FnCpf1 crRNA. FnCpf1 crRNA. D) Cpf1 base editor. Like Cas9 base editors, Cpf1 base editors are from ed. from the fusion of Cpf1 with. E) Cpf1 transcriptional activator. F) Cpf1 transcriptional repressor.

3. Emerging CRISPR-Cas systems for genome and transcriptome modification

Concomitant with the development of CRISPR-Cas9 and CRISPR-Cpf1 for genome editing, other CRISPR-Cas operons have been adapted for genome and transcriptome modifications. These emerging technologies provide a means to engineer large genomic deletions, large DNA insertions into safe harbor loci, which remains somewhat challenging using conventional CRISPR-Cas systems, or for the modulation of gene expression. These systems include Cascade-Cas3 from T. fusca, an RNA-guided DNA helicase-nuclease capable of removing large DNA segments; CAST from Scytonema hofmanni, Anabaena cylindrica and Vibrio cholerae (Tn6677) which allow for RNA-guided insertion of large DNA cargos to any specified location within a genome, and RNA-guided RNA modifying systems from L. shahii or Candidatus Scalindua brodae for transcriptome silencing or editing.

3.1 Cascade-Cas3

The Type I CRISPR-Cas system from T. fusca, which has been used for genome editing of human cell lines, comprises a multi-subunit protein complex called Cascade that pairs with a short 61-nucleotide long crRNA to recruit a highly processive DNA helicase-nuclease named Cas3 (Figure 4A-C) [5]. The Cascade complex from T. fusca comprises six Cas7 subunits, two Cas11 subunits, as well as Cas5, Cas8 and Cas6 subunits (Figure 4A). Once bound to its target DNA, Cascade undergoes massive conformational changes stabilizing the newly formed R-loop (Figure 4B). This allows for the recruitment of Cas3 which then nicks the non-paired DNA strand and unidirectionally shreds, in a 3′ to 5′ orientation, the target DNA upstream of the PAM sequence (5′-AAG-3′). In this system, the PAM lies 5′ of the protospacer element. Cascade-Cas3 from T. fusca can generate long range deletions, from a few hundred to several thousand nucleotides. For editing in mammalian cells, sequences encoding NLSs were added to Cas3 and Cas7 subunits.

Figure 4.

Cascade-Cas3. A) Schematic representation of the Cascade-Cas3 operon from Thermobifida fusca. B) Schematic representation of the T. fusca Cascade-Cas3 system. The target genomic DNA is shown in black, with the PAM sequence (5′-AAG-3′) highlighted in green. The crRNA is shown in dark blue. The Cas3 subunit is shown in light blue, Cas8 shown in purple, Cas5 shown in tan, Cas6 shown in dark brown, two Cas11 subunits shown in green, and six Cas7 subunits shown in light brown. C) Sequence of the T. fusca Cascade-Cas3 crRNA.

Several other Cascase-Cas3 systems have been developed for genome editing in bacteria and human cell lines. These include Type I-E Cascade-Cas3 from E. coli, and type I-C from Pseudomonas aeruginosa [4, 5, 6]. These systems have different PAM requirements, use slightly different Cas components, as well as crRNAs of different lengths and structures.

3.2 CRISPR-associated transposases

Insertion of large DNA elements using CRISPR-SpCas9 or CRISPR-FnCpf1 technologies has remained a major challenge. One emerging technology that may resolve this issue is CAST, which functions by recruiting Tn7-like transposase components to a specific location within the genome of a cell via guide RNA-target complementarity recognized by a naturally occurring inactive Cas12k variant. The Class 2 Type V-K Tn7-like CRISPR system from Scytonema hofmanni, the best described CAST system, comprises 6 components: Cas12k, a CRISPR-associated protein lacking endonuclease activity, a 216 nucleotide-long tracrRNA, a 34 nucleotide-long crRNA, and Tn7-like transposase subunits encoded by genes tnsB, tnsC, and tniQ (Figure 5A and B) [7, 16]. Similar to CRISPR-Cas9 systems adapted for genome editing, the type V-K tracrRNA and crRNA can be fused together to form a sgRNA. For Cas12k to recognize its target DNA, a 5′-NGTN-3′ PAM sequence must be present at the 5′ end of the protospacer element. Cargo insertion occurs unidirectionally in a 5′ Left End (LE) to 3′ Right End (RE) orientation (Figure 5B,C). A 5 bp integration site is found both 5′ and 3′ of the integration site. Cargos up to 10 kilobases can be introduced using this system and their integration into the host genome occurs 60–66 nucleotides downstream of the PAM sequence (Figure 5C). A variety of CAST systems have been identified in several bacterial species. These include CAST systems from A. cylindrica (AcCAST) which comprises similar components, possess similar PAM requirements (5′-NGTN-3′) and promotes insertion of transposons 49–56 nucleotides downstream of the PAM sequence [7, 16], as well as the Type I-C Cascade Tn7-like transposase system from V. cholerae (Tn6677), which promotes the bidirectional insertion of transposed elements 47–51 nucleotides downstream of PAM sequence [34].

Figure 5.

CRISPR-Cas12k and associated transposase. A) Schematic representation of the CAST-Cas12k operon from Scytonema hofmanni. B) Schematic representation of the S. hofmanni CAST-Cas12 system. The target genomic DNA is shown in black, with the PAM sequence (5′-NGTN-3′) highlighted in green. The sgRNA is shown in purple. The tnsB, tnsC, and tniQ subunits are shown in gray, green, and light blue, respectively. The cargo DNA to be inserted, shown in green, is flanked by LE and RE sequences, shown in yellow and red, respectively. Black arrows indicate the site of integration, which occurs 60–66 bp downstream of the PAM. A 5 bp integration site is found both 5′ and 3′ of the integration site. C) Sequence of the S. hofmanni CAST-Cas12k sgRNA.Nucleotide sequence corresponds to the sgRNA for ShCas12k. Ns represent the CRISPR spacer. D) Sequences of the RE and LE from the S. hofmanni CAST-Cas12 system.

3.3 RNA interference systems

Whereas the vast majority of CRISPR-Cas systems have evolved to protect against invading DNA species, Type VI CRISPR-Cas13 and the newly identified Type III-E CRISPR-Cas7–11 effectors are RNA-guided RNA interfering systems. These systems have been used to silence gene expression at the transcriptional level and have been modified to edit RNA transcripts. Several Type VI and Type III-E systems have been described. For simplicity, we present two of these systems: Type VI CRISPR-Cas13 from L. shahii (Cas13a) [9] and Type III-E CRISPR-Cas7–11 from Candidatus Scalindua brodae (Sb-gRAMP) [17].

Type VI CRISPR-Cas13a from L. shahii (also referred to as CRISPR-LshC2c2) is a single protein effector system that comprises a large ribonuclease containing two Higher Eukaryotes and Prokaryotes Nucleotide (HEPN) binding domains called C2c2 that pairs with a short 54 nucleotide-long crRNA to promote the cleavage of ssRNA transcripts at uracil residues [9] (Figure 6A and B). Of the 54 nucleotides, 28 residues provide target specificity (antisense to the protospacer) and the other 26 residues pair with C2c2 (Figure 6C). Cleavage of the ssRNA is sensitive to the nucleotide composition at the 3′end of the protospacer (also referred to as protospacer-flanking site or PFS). Spacers with a G immediately flanking the 3′ end of the protospacer were cleaved less efficiently than those containing any other nucleotides.

Figure 6.

CRISPR-Cas13. A) Schematic representation of the CRISPR-Cas13a (LshC2c2) operon from Leptotrichia shahii. B) Schematic representation of the CRISPR-Cas13a (LshC2c2) system. The target RNA is shown in black with cleavage sites indicated by black arrows. The crRNA is shown in pink. Unlike DNA targeting systems, RNA targeting systems do not require PAM sequences, but their activity may be influence by a PFS. LshC2c2 cuts its target RNA at accessible uracil residues (red triangles). C) Sequence of the CRISPR-Cas13a (LshC2c2) crRNA.

Not only have Type VI CRISPR-Cas13 systems been developed for the degradation of RNA species, but the fusion of a nuclease-dead Cas13b from Prevotella sp. P5–125 with the adenosine deaminase acting on RNA type 2 (ADAR2) allows for the editing of RNA transcripts without interfering with the genomic sequence [8]. This may have applications for the treatment and understanding of genetic disorders.

The Type III-E effector system Cas7–11 from Candidatus Scalindua brodae, also referred to as giant Repeat-Associated Mysterious Protein (Sb-gRAMP), is a single-protein effector system that comprises a large modular protein containing four Cas7-like and a single Cas11-like domains with intrinsic endoribonuclease activity that pairs with a 47 nucleotide-long crRNA to cleave ssRNA at position 3 and 9 of the spacer [17] (Figure 7A and B). The 5′ most 27 nucleotides of the crRNA encode the direct repeat segment of the crRNA whereas the subsequent 20 nucleotides provide target specificity (Figure 7C). Several Class 1 type III-E systems have been identified, each requiring specific crRNA and endonuclease activity, most of which also introduce ssRNA breaks 6 nucleotides apart, at position 3 and 9 of the spacer [10]. Unlike DNA targeting systems, RNA targeting systems do not require PAM sequences, but their activity may be influenced by a PFS.

Figure 7.

CRISPR-Cas7–11. A) Schematic representation of the Cas7–11 operon from Candidatus Scalindua brodae. B) Schematic representation of the Cas7–11 Sb-gRAMP system. The target RNA is shown in black with cleavage sites indicated by black arrows 6 nucleotides apart, at position 3 and 9 of the spacer. The crRNA is shown in green. Unlike DNA targeting systems, RNA targeting systems do not require PAM sequences, but their activity may be influence by a PFS. C) Sequence of the Cas7–11 Sb-gRAMP crRNA.

4. Discussion

Whereas the vast majority of genetic manipulations can be performed using conventional CRISPR-SpCas9 technology, there are some inherent limitations that may be alleviated by emerging technologies. These include the possibility of introducing DNA DSBs at off-target sites; the possibility of inserting undesired mutations at on-target sites; the requirement for specific PAM sequences, which may somewhat limit the number of target possibilities; the scope of editing; and the delivery of these reagents, particularly for manipulations in vivo or for therapeutic interventions. Table 2 explores the similarities and differences between the systems described in this chapter as well as their advantages and limitations.

System	Source	Mechanism	Advantages	Limitations
CRISPR-Cas9	Streptococcus pyogenes	Cas9 pairs with a sgRNA to introduce a DNA DSB 3 nt upstream of the PAM. DSBs are resolved by NHEJ or HR	A well-established, conventional system; can be used to engineer a plethora of genetic modifications, ranging from nucleotide substitutions to chromosomal translocations; a single effector system which may be easier to use as opposed to multiprotein effector systems	Off-target editing of genomic DNA may occur; uses a long sgRNA; uses a large endonuclease which may be difficult to package in viral delivery systems; introduces DNA DSBs which may have deleterious effects if not resolved properly, particularly for therapeutic applications
CRISPR-Cpf1	Francisella novicida	Cpf1 pairs with a crRNA to introduce scattered DNA breaks at position 18 of the non-target strand and 23 of the target strand, leaving a 5′ overhang. DSBs are resolved by NHEJ or HR	Can be used to engineer a plethora of genetic modifications, ranging from nucleotide substitutions to chromosomal translocations; uses a different PAM than SpCas9, extending the range of possible target sites; a single effector system which may be easier to use as opposed to multiprotein effector systems; uses a short crRNA that may be easier to synthesize and deliver	Off-target editing of genomic DNA may occur; uses a large endonuclease (although smaller than SpCas9) which may be difficult to package in viral delivery systems; introduces DNA DSBs which may have deleterious effects if not resolved properly, particularly for therapeutic applications
Cascade-Cas3	Thermobifida fusca	Cascade pairs with a crRNA to recruit Cas3, a highly processive DNA helicase-nuclease, which nicks the non-paired DNA strand and unidirectionally shreds the target DNA upstream of the PAM sequence in a 3′ to 5′ orientation	Can be used to generate large deletions; uses a short crRNA that may be easier to synthesize and deliver	Off-target editing of genomic DNA may occur; large multiprotein effector which can be difficult to package for viral delivery; limited in the scope of editing; no control over the length of the deletions
CRISPR-CAST	Scytonema hofmanni	Cas12k and Tn7-like transposase components pair with a sgRNA to insert large DNA cargo unidirectionally in a 5′ LE to 3′ RE orientation 60–66 nt downstream of the PAM. A 5 bp integration site is duplicated and found 5′ and 3′ of the cargo integration	Can be used to insert large DNA elements up to ~10Kb	Off-target editing of genomic DNA may occur; insertion results in the integration of flanking sequences and duplication of the integration site; a multiprotein effector which can be difficult to package for viral delivery; uses a long sgRNA; limited in the scope of editing; demonstrated efficacy only in bacterial systems
CRISPR-Cas13	Leptotrichia shahii	C2c2, a HEPN containing protein, pairs with a crRNA to cleave ssRNA transcripts at exposed uracil residues	Targets RNA rather than DNA, can be used to silence gene expression at the translational level instead of permanently altering the DNA; a single effector system which may be easier to use as opposed to multiprotein effector systems; uses a short crRNA that may be easier to synthesize and deliver	Off-target editing of ssRNA transcripts may occur; cleavage of ssRNA transcripts occurs at exposed uracil residues, which are difficult to predict; a large effector protein which can be difficult to package for viral delivery; demonstrated efficacy only in bacterial systems
Cas7–11	Candidatus Scalindua brodae	A protein (gRAMP) with Cas7-like and Cas11-like domains with intrinsic endoribonuclease activity pairs with a crRNA to cleave ssRNA transcripts at positions 3 and 9 of the spacer	Targets RNA rather than DNA, can be used to silence gene expression at the translational level instead of permanently altering the DNA; cleavage of ssRNA transcripts occurs at positions 3 and 9 of the spacer, which are easily predictable compared to CRISPR-Cas13 systems; a single effector system which may be easier to use as opposed to multiprotein effector systems; uses a short crRNA that may be easier to synthesize and deliver	Off-target editing of ssRNA transcripts may occur; a large effector protein which can be difficult to package for viral delivery; demonstrated efficacy only in bacterial systems

Table 2.

Mechanisms, advantages, and limitations associated with prototypical CRISPR-Cas systems.

4.1 Off-target mutations

One major limitation associated with the use of CRIPSR-SpCas9 technology is the potential for inserting genetic changes at sites other than the intended ones, also referred to as off-target sites. Off-target cleavage may occur due to the lack of SpCas9 specificity, which stems from the tolerance of the endonuclease for RNA–DNA mismatches, RNA bulges, or DNA bulges [12, 13]. Although there are still no simple and definitive guidelines defining SpCas9 specificity, the number and the position of mismatches relative to the PAM sequence are important. Whereas a single mismatch within the first 13 nucleotides upstream of the PAM sequence can abrogate SpCas9 activity, up to seven mismatches at the 5′ end of the guide sequence can be tolerated [12, 13]. To avoid off-target modifications, various strategies have been established. These include the development of bioinformatic tools to identify highly specific target sequences; the modification of SpCas9 to improve specificity or the duration of its action within cells; and the development of delivery formats and methods to limit the duration of SpCas9 activity [14]. Emerging RNA-guided RNA modifying systems may also help resolves this issue. These systems can modulate gene expression by targeting RNA transcripts rather than modifying genes at the DNA level [9, 11].

To identify highly selective guide sequences, various guide selection applications have been developed. These include Cas-Designer, quick guide-RNA designer for CRISPR/Cas derived RNA guided nucleases (http://www.rgenome.net/); CRISPR Design (http://crispr.mit.edu/);E-CRISP (http://www.e-crisp.org/E-CRISP/); ZiFit (http://zifit.partners.org/ZiFiT/). During the implementation of CRISPR-SpCas9 technology for mouse genome editing, our laboratory also developed a stringent guide selection procedure which makes use of Cas-Designer and Cas-Offinder from CRISPR RGEN Tools (http://www.rgenome.net/). For more details about this guide selection procedure, we recommend reading.

To control the duration of SpCas9 activity within cells, genetically encoded inducible systems have been developed. One of these makes use of split SpCas9 fused to the Magnet Photoactivatable System. Fusion of the split SpCas9 is achieved by illuminating cells with blue light. The fused split SpCas9 is then able to bind a sgRNA and cleave its target site. A second system involves self-cleavable CRIPSR systems, where sequences encoding SpCas9 are targeted by a sgRNA in order to promote its degradation upon expression of the endonuclease. Duration of SpCas9 activity can also be controlled by the delivery of RNA transcripts encoding the various components of the CRISPR-SpCas9 system or via delivery of the ribonucleoprotein complex comprising both the sgRNA and the endonuclease.

In addition to robust guide selection procedures and inducible/self-inactivating systems, various additional strategies have been developed to improve target specificity. These include the use of paired SpCas9 nickases, in which the RuvC-like domains are inactivated, that introduce scattered DNA DSB, guided by a pair of sgRNAs recognizing juxtaposed sequences. The requirement for recognizing these sequences doubles the length of the target sequence and thus increases target specificity. Similarly, catalytically inactive but sgRNA competent pairs of SpCas9 fused to the non-specific endonuclease FokI was shown to reduce off-target activity. Directed evolution has also been used to engineer improved SpCas9 with increased target specificity. These enzymes have been shown to increase on-target over off-target activity by several folds.

Finally, modifications to the guide RNAs themselves were also used to reduce off-target cleavage. Previous studies have shown that, counterintuitively, shortened guide sequences increase specificity without affecting on-target activity.

4.2 Unintended mutations at the target site

Engineering specific mutations using conventional CRIPSR-SpCas9 technology relies on HR. While HR and NHEJ are both active in most (but not all) dividing cells, NHEJ is usually the sole repair pathway active in postmitotic cells. DNA repair via the NHEJ pathway, as previously mentioned, results in the insertion or deletion of genetic material, and does not allow for the introduction of desired mutations. To get around this, Base Editors and Prime Editors were developed. These systems, as described above, can introduce specific mutations by directly changing the nucleotide composition at the target site, bypassing the need to activate DSB repair pathways. Stabilization of the mutation is performed by the DNA MMR pathway which is present and active in all cells. While these systems allow for the insertion of precise mutations, the window in which they operate is narrow (a few nucleotides), the insertion of mutation(s) depends on the presence of a PAM sequence, stabilization of the mutation is not always complete, and undesired collateral nucleotide substitutions may occur [18, 19, 20, 21, 31, 33]. Moreover, these systems are bulkier than SpCas9 and may not be easily packaged within viral delivery systems. Nevertheless, these systems provide an alternative to conventional CRISPR-SpCas9 systems for engineering mutations in cells that are not amenable to HR or for therapeutic intervention where introducing DNA DSBs may have deleterious effects.

4.3 Editing scope

While conventional CRISPR-Cas9 systems can be used to engineer virtually any kind of mutations in vivo, insertions or deletions of large DNA elements remain somewhat challenging. Cascade-Cas3 and CRISPR-CAST systems may provide alternatives to using conventional CRISPR-Cas9 systems. Cascade-Cas3 systems have been used to engineer large deletions in cultured cells and can potentially be applied to animal models [4, 5, 6]. Deletions range from several hundred to several thousand base pairs. The major drawback of using this technology is the apparent uncontrollable processivity of the helicase-nuclease Cas3. Consequently, deletions of various sizes must be characterized using a large number of primer pairs flanking the potential deletions. Other limitations include the difficulty of packaging multiprotein systems for viral delivery in vivo. CRISPR-CAST systems, on the other hand, allow for targeted integration of large genetic material. In E. coli, up to 10 kilobases of DNA have been successfully inserted. Unlike conventional CRISPR-Cas9 systems, however, which allow for scarless integration of genetic material, insertion of DNA elements using CRISPR-CAST systems results in the integration of flanking sequences (LE and RE) as well as duplication of the integration site. Consequently, this technology cannot be used for precise DNA insertion, but can be used to facilitate insertion of transgenes at safe harbor sites within genomes. Like other multiprotein effector systems, CRISPR-CAST systems may also be difficult to package for viral delivery in vivo. Although CRISPR-CAST systems have only been used in bacteria, the implementation of these systems for their use in mammalian genome engineering is likely and provides an alternative to conventional CRISPR-Cas9 systems.

4.4 Delivery

The vast majority of CRISPR-Cas systems for genome editing make use of either multiprotein effectors or large single effectors. Although delivering these systems, together with their cognate guide RNAs, in cultured cells or zygotes for the generation of animal models does not represent a major hurdle and is routinely performed, delivering these systems in vivo, for therapeutic interventions, does represent a major challenge. More compact CRISPR-Cas9 and Cas3 systems have been identified and these may represent viable alternatives to other larger and more complex CRISPR-Cas system for therapeutic purposes [6].

4.5 PAM requirements

The requirement for CRISPR-Cas systems to recognize short genomic sequences has long been viewed as major disincentive for the use of these technologies for genome engineering. However, most PAM sequences are quite short and are likely present at a high frequency within mammalian genomes. Moreover, the development of several CRISPR-Cas9 and Cpf1 systems with distinct PAM requirements and the generation of engineered SpCas9 endonucleases with altered PAM specificities have expanded the targeting capabilities of CRISPR-Cas systems.

5. Conclusion

Since the discovery of targetable nucleases, more notably CRISPR-Cas systems, the field of genetic and genome engineering has expanded exponentially. In less than a decade, these systems have not only revolutionized how research is performed but have also allowed for a plethora of scientific discoveries and paved the way for novel human therapeutics. Emerging technologies such as Cascade-Cas3, CAST, CRISPR-Cas7–11, and CRISPR-Cas13 provide alternatives to current technologies and may fill a critical technological gap to improve the specificity and scope of genome editing. Moreover, the implementation of these tools as therapeutic agents offers the potential to treat or even cure human genetic diseases.

Acknowledgments

The author would like to thank the Indiana University School of Medicine for the financial support of the Indiana University Genome Editing Center.