What's the role or function of the homologous arms in a donor template in a knockout/knock edition via Crispr-cas9?

I have to make an exposition in my university about Crispr-cas9 edition and I have some questions about the method. In the knock out/knock in technique is used a plasmid containing the DNA that codifies for the Cas9 protein and guideRNA and a Donor template that has a gene for puromicine resistance, and a DNA that codifies for a fluorescent protein and "left homologous arm" and "right homologous arms". My questions is What's the fuction of the homologous arms?Furthermore what's the role of the Loxp(look at the image)? The method is described here (page 5-10)

Lung Cancer: Molecular and Cellular Abnormalities


Ras genes function as complex cytoplasmic switches that transmit and amplify external growth signals in a cascade toward the nucleus to regulate gene transcription. Acquired somatic mutations within the Ras gene are observed in a subset of a wide range of human cancers where it locks the protein in a conformation that aberrantly signals growth. Although there are many different members of the Ras gene family, the Ki-ras gene (at chromosome12p12.1) appears to be exclusively targeted for somatic activating mutations in lung cancer. Activating Ki-ras mutations were detected in approximately 30% of non-SCLC tumor samples however, tumors with nonadenocarcinoma histology may have a lower 10–15% incidence. These mutations appear to arise as a late event, similar to colon cancer, in the genesis of non-SCLC. In addition, several studies have confirmed an inferior prognosis in patients with early stage lung tumors that have acquired activating ras mutations. Strikingly, Ki-ras mutations are absent in all SCLC tumors studied to date.


It is a long-standing goal of scientists and breeders to precisely control a gene for studying its function as well as improving crop yield, quality, and tolerance to various environmental stresses. The discovery and modification of CRISPR/Cas system, a nature-occurred gene editing tool, opens an era for studying gene function and precision crop breeding.

Aim of Review

In this review, we first introduce the brief history of CRISPR/Cas discovery followed the mechanism and application of CRISPR/Cas system on gene function study and crop improvement. Currently, CRISPR/Cas genome editing has been becoming a mature cutting-edge biotechnological tool for crop improvement that already used in many different traits in crops, including pathogen resistance, abiotic tolerance, plant development and morphology and even secondary metabolism and fiber development. Finally, we point out the major issues associating with CRISPR/Cas system and the future research directions.

Key Scientific Concepts of Review: CRISPR/Cas9 system is a robust and powerful biotechnological tool for targeting an individual DNA and RNA sequence in the genome. It can be used to target a sequence for gene knockin, knockout and replacement as well as monitoring and regulating gene expression at the genome and epigenome levels by binding a specific sequence. Agrobacterium-mediated method is still the major and efficient method for delivering CRISPR/Cas regents into targeted plant cells. However, other delivery methods, such as virus-mediated method, have been developed and enhanced the application potentials of CRISPR/Cas9-based crop improvement. PAM requirement offers the CRISPR/Cas9-targted genetic loci and also limits the application of CRISPR/Cas9. Discovering new Cas proteins and modifying current Cas enzymes play an important role in CRISPR/Cas9-based genome editing. Developing a better CRISPR/Cas9 system, including the delivery system and the methods eliminating off-target effects, and finding key/master genes for controlling crop growth and development is two major directions for CRISPR/Cas9-based crop improvement.

Access options

Get full journal access for 1 year

All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.

Get time limited or full article access on ReadCube.

All prices are NET prices.

Leaders in Pharmaceutical Business Intelligence (LPBI) Group

CRISPR/Cas9, Familial Amyloid Polyneuropathy (FAP) and Neurodegenerative Disease, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 2: CRISPR for Gene Editing and DNA Repair

CRISPR/Cas9, Familial Amyloid Polyneuropathy ( FAP) and Neurodegenerative Disease

Curator: Larry H. Bernstein, MD, FCAP

CRISPR/Cas9 and Targeted Genome Editing: A New Era in Molecular Biology

The development of efficient and reliable ways to make precise, targeted changes to the genome of living cells is a long-standing goal for biomedical researchers. Recently, a new tool based on a bacterial CRISPR-associated protein-9 nuclease (Cas9) from Streptococcus pyogenes has generated considerable excitement (1). This follows several attempts over the years to manipulate gene function, including homologous recombination (2) and RNA interference (RNAi) (3). RNAi, in particular, became a laboratory staple enabling inexpensive and high-throughput interrogation of gene function (4, 5), but it is hampered by providing only temporary inhibition of gene function and unpredictable off-target effects (6). Other recent approaches to targeted genome modification – zinc-finger nucleases [ZFNs, (7)] and transcription-activator like effector nucleases [TALENs (8)]– enable researchers to generate permanent mutations by introducing doublestranded breaks to activate repair pathways. These approaches are costly and time-consuming to engineer, limiting their widespread use, particularly for large scale, high-throughput studies.

The Biology of Cas9

The functions of CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) and CRISPR-associated (Cas) genes are essential in adaptive immunity in select bacteria and archaea, enabling the organisms to respond to and eliminate invading genetic material. These repeats were initially discovered in the 1980s in E. coli (9), but their function wasn’t confirmed until 2007 by Barrangou and colleagues, who demonstrated that S. thermophilus can acquire resistance against a bacteriophage by integrating a genome fragment of an infectious virus into its CRISPR locus (10).

Three types of CRISPR mechanisms have been identified, of which type II is the most studied. In this case, invading DNA from viruses or plasmids is cut into small fragments and incorporated into a CRISPR locus amidst a series of short repeats (around 20 bps). The loci are transcribed, and transcripts are then processed to generate small RNAs (crRNA – CRISPR RNA), which are used to guide effector endonucleases that target invading DNA based on sequence complementarity (Figure 1) (11).

Figure 1. Cas9 in vivo: Bacterial Adaptive Immunity

In the acquisition phase, foreign DNA is incorporated into the bacterial genome at the CRISPR loci. CRISPR loci is then transcribed and processed into crRNA during crRNA biogenesis. During interference, Cas9 endonuclease complexed with a crRNA and separate tracrRNA cleaves foreign DNA containing a 20-nucleotide crRNA complementary sequence adjacent to the PAM sequence. (Figure not drawn to scale.)

One Cas protein, Cas9 (also known as Csn1), has been shown, through knockdown and rescue experiments to be a key player in certain CRISPR mechanisms (specifically type II CRISPR systems). The type II CRISPR mechanism is unique compared to other CRISPR systems, as only one Cas protein (Cas9) is required for gene silencing (12). In type II systems, Cas9 participates in the processing of crRNAs (12), and is responsible for the destruction of the target DNA (11). Cas9’s function in both of these steps relies on the presence of two nuclease domains, a RuvC-like nuclease domain located at the amino terminus and a HNH-like nuclease domain that resides in the mid-region of the protein (13).

To achieve site-specific DNA recognition and cleavage, Cas9 must be complexed with both a crRNA and a separate trans-activating crRNA (tracrRNA or trRNA), that is partially complementary to the crRNA (11). The tracrRNA is required for crRNA maturation from a primary transcript encoding multiple pre-crRNAs. This occurs in the presence of RNase III and Cas9 (12).

During the destruction of target DNA, the HNH and RuvC-like nuclease domains cut both DNA strands, generating double-stranded breaks (DSBs) at sites defined by a 20-nucleotide target sequence within an associated crRNA transcript (11, 14). The HNH domain cleaves the complementary strand, while the RuvC domain cleaves the noncomplementary strand.

The double-stranded endonuclease activity of Cas9 also requires that a short conserved sequence, (2–5 nts) known as protospacer-associated motif (PAM), follows immediately 3´- of the crRNA complementary sequence (15). In fact, even fully complementary sequences are ignored by Cas9-RNA in the absence of a PAM sequence (16).

Cas9 and CRISPR as a New Tool in Molecular Biology

The simplicity of the type II CRISPR nuclease, with only three required components (Cas9 along with the crRNA and trRNA) makes this system amenable to adaptation for genome editing. This potential was realized in 2012 by the Doudna and Charpentier labs (11). Based on the type II CRISPR system described previously, the authors developed a simplified two-component system by combining trRNA and crRNA into a single synthetic single guide RNA (sgRNA). sgRNAprogrammed Cas9 was shown to be as effective as Cas9 programmed with separate trRNA and crRNA in guiding targeted gene alterations (Figure 2A).

To date, three different variants of the Cas9 nuclease have been adopted in genome-editing protocols. The first is wild-type Cas9, which can site-specifically cleave double-stranded DNA, resulting in the activation of the doublestrand break (DSB) repair machinery. DSBs can be repaired by the cellular Non-Homologous End Joining (NHEJ) pathway (17), resulting in insertions and/or deletions (indels) which disrupt the targeted locus. Alternatively, if a donor template with homology to the targeted locus is supplied, the DSB may be repaired by the homology-directed repair (HDR) pathway allowing for precise replacement mutations to be made (Figure 2A) (17, 18).

Cong and colleagues (1) took the Cas9 system a step further towards increased precision by developing a mutant form, known as Cas9D10A, with only nickase activity. This means it cleaves only one DNA strand, and does not activate NHEJ. Instead, when provided with a homologous repair template, DNA repairs are conducted via the high-fidelity HDR pathway only, resulting in reduced indel mutations (1, 11, 19). Cas9D10A is even more appealing in terms of target specificity when loci are targeted by paired Cas9 complexes designed to generate adjacent DNA nicks (20) (see further details about “paired nickases” in Figure 2B).

The third variant is a nuclease-deficient Cas9 (dCas9, Figure 2C) (21). Mutations H840A in the HNH domain and D10A in the RuvC domain inactivate cleavage activity, but do not prevent DNA binding (11, 22). Therefore, this variant can be used to sequence-specifically target any region of the genome without cleavage. Instead, by fusing with various effector domains, dCas9 can be used either as a gene silencing or activation tool (21, 23–26). Furthermore, it can be used as a visualization tool. For instance, Chen and colleagues used dCas9 fused to Enhanced Green Fluorescent Protein (EGFP) to visualize repetitive DNA sequences with a single sgRNA or nonrepetitive loci using multiple sgRNAs (27).

Figure 2. CRISPR/Cas9 System Applications

  1. Wild-type Cas9 nuclease site specifically cleaves double-stranded DNA activating double-strand break repair machinery. In the absence of a homologous repair template non-homologous end joining can result in indels disrupting the target sequence. Alternatively, precise mutations and knock-ins can be made by providing a homologous repair template and exploiting the homology directed repair pathway.
    B. Mutated Cas9 makes a site specific single-strand nick. Two sgRNA can be used to introduce a staggered double-stranded break which can then undergo homology directed repair.
    C. Nuclease-deficient Cas9 can be fused with various effector domains allowing specific localization. For example, transcriptional activators, repressors, and fluorescent proteins.

Targeting Efficiency and Off-target Mutations

Targeting efficiency, or the percentage of desired mutation achieved, is one of the most important parameters by which to assess a genome-editing tool. The targeting efficiency of Cas9 compares favorably with more established methods, such as TALENs or ZFNs (8). For example, in human cells, custom-designed ZFNs and TALENs could only achieve efficiencies ranging from 1% to 50% (29–31). In contrast, the Cas9 system has been reported to have efficiencies up to >70% in zebrafish (32) and plants (33), and ranging from 2–5% in induced pluripotent stem cells (34). In addition, Zhou and colleagues were able to improve genome targeting up to 78% in one-cell mouse embryos, and achieved effective germline transmission through the use of dual sgRNAs to simultaneously target an individual gene (35).

A widely used method to identify mutations is the T7 Endonuclease I mutation detection assay (36, 37) (Figure 3). This assay detects heteroduplex DNA that results from the annealing of a DNA strand, including desired mutations, with a wildtype DNA strand (37).

Figure 3. T7 Endonuclease I Targeting Efficiency Assay

Genomic DNA is amplified with primers bracketing the modified locus. PCR products are then denatured and re-annealed yielding 3 possible structures. Duplexes containing a mismatch are digested by T7 Endonuclease I. The DNA is then electrophoretically separated and fragment analysis is used to calculate targeting efficiency.

Another important parameter is the incidence of off-target mutations. Such mutations are likely to appear in sites that have differences of only a few nucleotides compared to the original sequence, as long as they are adjacent to a PAM sequence. This occurs as Cas9 can tolerate up to 5 base mismatches within the protospacer region (36) or a single base difference in the PAM sequence (38). Off-target mutations are generally more difficult to detect, requiring whole-genome sequencing to rule them out completely.

Recent improvements to the CRISPR system for reducing off-target mutations have been made through the use of truncated gRNA (truncated within the crRNA-derived sequence) or by adding two extra guanine (G) nucleotides to the 5´ end (28, 37). Another way researchers have attempted to minimize off-target effects is with the use of “paired nickases” (20). This strategy uses D10A Cas9 and two sgRNAs complementary to the adjacent area on opposite strands of the target site (Figure 2B). While this induces DSBs in the target DNA, it is expected to create only single nicks in off-target locations and, therefore, result in minimal off-target mutations.

By leveraging computation to reduce off-target mutations, several groups have developed webbased tools to facilitate the identification of potential CRISPR target sites and assess their potential for off-target cleavage. Examples include the CRISPR Design Tool (38) and the ZiFiT Targeter, Version 4.2 (39, 40).

Applications as a Genome-editing and Genome Targeting Tool

Following its initial demonstration in 2012 (9), the CRISPR/Cas9 system has been widely adopted. This has already been successfully used to target important genes in many cell lines and organisms, including human (34), bacteria (41), zebrafish (32), C. elegans (42), plants (34), Xenopus tropicalis (43), yeast (44), Drosophila (45), monkeys (46), rabbits (47), pigs (42), rats (48) and mice (49). Several groups have now taken advantage of this method to introduce single point mutations (deletions or insertions) in a particular target gene, via a single gRNA (14, 21, 29). Using a pair of gRNA-directed Cas9 nucleases instead, it is also possible to induce large deletions or genomic rearrangements, such as inversions or translocations (50). A recent exciting development is the use of the dCas9 version of the CRISPR/Cas9 system to target protein domains for transcriptional regulation (26, 51, 52), epigenetic modification (25), and microscopic visualization of specific genome loci (27).

The CRISPR/Cas9 system requires only the redesign of the crRNA to change target specificity. This contrasts with other genome editing tools, including zinc finger and TALENs, where redesign of the protein-DNA interface is required. Furthermore, CRISPR/Cas9 enables rapid genome-wide interrogation of gene function by generating large gRNA libraries (51, 53) for genomic screening.

The Future of CRISPR/Cas9

The rapid progress in developing Cas9 into a set of tools for cell and molecular biology research has been remarkable, likely due to the simplicity, high efficiency and versatility of the system. Of the designer nuclease systems currently available for precision genome engineering, the CRISPR/Cas system is by far the most user friendly. It is now also clear that Cas9’s potential reaches beyond DNA cleavage, and its usefulness for genome locus-specific recruitment of proteins will likely only be limited by our imagination.

Scientists urge caution in using new CRISPR technology to treat human genetic disease

The bacterial enzyme Cas9 is the engine of RNA-programmed genome engineering in human cells. (Graphic by Jennifer Doudna/UC Berkeley)

A group of 18 scientists and ethicists today warned that a revolutionary new tool to cut and splice DNA should be used cautiously when attempting to fix human genetic disease, and strongly discouraged any attempts at making changes to the human genome that could be passed on to offspring.

Among the authors of this warning is Jennifer Doudna, the co-inventor of the technology, called CRISPR-Cas9, which is driving a new interest in gene therapy, or “genome engineering.” She and colleagues co-authored a perspective piece that appears in the March 20 issue of Science, based on discussions at a meeting that took place in Napa on Jan. 24. The same issue of Science features a collection of recent research papers, commentary and news articles on CRISPR and its implications. …..

A prudent path forward for genomic engineering and germline gene modification

Correcting genetic defects

Scientists today are changing DNA sequences to correct genetic defects in animals as well as cultured tissues generated from stem cells, strategies that could eventually be used to treat human disease. The technology can also be used to engineer animals with genetic diseases mimicking human disease, which could lead to new insights into previously enigmatic disorders.

The CRISPR-Cas9 tool is still being refined to ensure that genetic changes are precisely targeted, Doudna said. Nevertheless, the authors met “… to initiate an informed discussion of the uses of genome engineering technology, and to identify proactively those areas where current action is essential to prepare for future developments. We recommend taking immediate steps toward ensuring that the application of genome engineering technology is performed safely and ethically.”

Amyloid CRISPR Plasmids and si/shRNA Gene Silencers

Santa Cruz Biotechnology, Inc. offers a broad range of gene silencers in the form of siRNAs, shRNA Plasmids and shRNA Lentiviral Particles as well as CRISPR/Cas9 Knockout and CRISPR Double Nickase plasmids. Amyloid gene silencers are available as Amyloid siRNA, Amyloid shRNA Plasmid, Amyloid shRNA Lentiviral Particles and Amyloid CRISPR/Cas9 Knockout plasmids. Amyloid CRISPR/dCas9 Activation Plasmids and CRISPR Lenti Activation Systems for gene activation are also available. Gene silencers and activators are useful for gene studies in combination with antibodies used for protein detection. Amyloid CRISPR Knockout, HDR and Nickase Knockout Plasmids

CRISPR-Cas9-Based Knockout of the Prion Protein and Its Effect on the Proteome

Mehrabian M, Brethour D, MacIsaac S, Kim JK, Gunawardana C.G, Wang H, et al.
PLoS ONE 2014 9(12): e114594.

The molecular function of the cellular prion protein (PrP C ) and the mechanism by which it may contribute to neurotoxicity in prion diseases and Alzheimer’s disease are only partially understood. Mouse neuroblastoma Neuro2a cells and, more recently, C2C12 myocytes and myotubes have emerged as popular models for investigating the cellular biology of PrP. Mouse epithelial NMuMG cells might become attractive models for studying the possible involvement of PrP in a morphogenetic program underlying epithelial-to-mesenchymal transitions. Here we describe the generation of PrP knockout clones from these cell lines using CRISPR-Cas9 knockout technology. More specifically, knockout clones were generated with two separate guide RNAs targeting recognition sites on opposite strands within the first hundred nucleotides of the Prnp coding sequence. Several PrP knockout clones were isolated and genomic insertions and deletions near the CRISPR-target sites were characterized. Subsequently, deep quantitative global proteome analyses that recorded the relative abundance of>3000 proteins (data deposited to ProteomeXchange Consortium) were undertaken to begin to characterize the molecular consequences of PrP deficiency. The levels of ∼120 proteins were shown to reproducibly correlate with the presence or absence of PrP, with most of these proteins belonging to extracellular components, cell junctions or the cytoskeleton.

Development and Applications of CRISPR-Cas9 for Genome Engineering

Recent advances in genome engineering technologies based on the CRISPR-associated RNA-guided endonuclease Cas9 are enabling the systematic interrogation of mammalian genome function. Analogous to the search function in modern word processors, Cas9 can be guided to specific locations within complex genomes by a short RNA search string. Using this system, DNA sequences within the endogenous genome and their functional outputs are now easily edited or modulated in virtually any organism of choice. Cas9-mediated genetic perturbation is simple and scalable, empowering researchers to elucidate the functional organization of the genome at the systems level and establish causal linkages between genetic variations and biological phenotypes. In this Review, we describe the development and applications of Cas9 for a variety of research or translational applications while highlighting challenges as well as future directions. Derived from a remarkable microbial defense system, Cas9 is driving innovative applications from basic biology to biotechnology and medicine.

The development of recombinant DNA technology in the 1970s marked the beginning of a new era for biology. For the first time, molecular biologists gained the ability to manipulate DNA molecules, making it possible to study genes and harness them to develop novel medicine and biotechnology. Recent advances in genome engineering technologies are sparking a new revolution in biological research. Rather than studying DNA taken out of the context of the genome, researchers can now directly edit or modulate the function of DNA sequences in their endogenous context in virtually any organism of choice, enabling them to elucidate the functional organization of the genome at the systems level, as well as identify causal genetic variations.

Broadly speaking, genome engineering refers to the process of making targeted modifications to the genome, its contexts (e.g., epigenetic marks), or its outputs (e.g., transcripts). The ability to do so easily and efficiently in eukaryotic and especially mammalian cells holds immense promise to transform basic science, biotechnology, and medicine (Figure 1).

For life sciences research, technologies that can delete, insert, and modify the DNA sequences of cells or organisms enable dissecting the function of specific genes and regulatory elements. Multiplexed editing could further allow the interrogation of gene or protein networks at a larger scale. Similarly, manipulating transcriptional regulation or chromatin states at particular loci can reveal how genetic material is organized and utilized within a cell, illuminating relationships between the architecture of the genome and its functions. In biotechnology, precise manipulation of genetic building blocks and regulatory machinery also facilitates the reverse engineering or reconstruction of useful biological systems, for example, by enhancing biofuel production pathways in industrially relevant organisms or by creating infection-resistant crops. Additionally, genome engineering is stimulating a new generation of drug development processes and medical therapeutics. Perturbation of multiple genes simultaneously could model the additive effects that underlie complex polygenic disorders, leading to new drug targets, while genome editing could directly correct harmful mutations in the context of human gene therapy (Tebas et al., 2014).

Eukaryotic genomes contain billions of DNA bases and are difficult to manipulate. One of the breakthroughs in genome manipulation has been the development of gene targeting by homologous recombination (HR), which integrates exogenous repair templates that contain sequence homology to the donor site (Figure 2A) (Capecchi, 1989). HR-mediated targeting has facilitated the generation of knockin and knockout animal models via manipulation of germline competent stem cells, dramatically advancing many areas of biological research. However, although HR-mediated gene targeting produces highly precise alterations, the desired recombination events occur extremely infrequently (1 in 10 6 –10 9 cells) (Capecchi, 1989), presenting enormous challenges for large-scale applications of gene-targeting experiments.

Genome Editing Technologies Exploit Endogenous DNA Repair Machinery

To overcome these challenges, a series of programmable nuclease-based genome editing technologies have been developed in recent years, enabling targeted and efficient modification of a variety of eukaryotic and particularly mammalian species. Of the current generation of genome editing technologies, the most rapidly developing is the class of RNA-guided endonucleases known as Cas9 from the microbial adaptive immune system CRISPR (clustered regularly interspaced short palindromic repeats), which can be easily targeted to virtually any genomic location of choice by a short RNA guide. Here, we review the development and applications of the CRISPR-associated endonuclease Cas9 as a platform technology for achieving targeted perturbation of endogenous genomic elements and also discuss challenges and future avenues for innovation. ……

Figure 4 Natural Mechanisms of Microbial CRISPR Systems in Adaptive Immunity

…… A key turning point came in 2005, when systematic analysis of the spacer sequences separating the individual direct repeats suggested their extrachromosomal and phage-associated origins (Mojica et al., 2005 Pourcel et al., 2005 Bolotin et al., 2005). This insight was tremendously exciting, especially given previous studies showing that CRISPR loci are transcribed (Tang et al., 2002) and that viruses are unable to infect archaeal cells carrying spacers corresponding to their own genomes (Mojica et al., 2005). Together, these findings led to the speculation that CRISPR arrays serve as an immune memory and defense mechanism, and individual spacers facilitate defense against bacteriophage infection by exploiting Watson-Crick base-pairing between nucleic acids (Mojica et al., 2005 Pourcel et al., 2005). Despite these compelling realizations that CRISPR loci might be involved in microbial immunity, the specific mechanism of how the spacers act to mediate viral defense remained a challenging puzzle. Several hypotheses were raised, including thoughts that CRISPR spacers act as small RNA guides to degrade viral transcripts in a RNAi-like mechanism (Makarova et al., 2006) or that CRISPR spacers direct Cas enzymes to cleave viral DNA at spacer-matching regions (Bolotin et al., 2005). …..

As the pace of CRISPR research accelerated, researchers quickly unraveled many details of each type of CRISPR system (Figure 4). Building on an earlier speculation that protospacer adjacent motifs (PAMs) may direct the type II Cas9 nuclease to cleave DNA (Bolotin et al., 2005), Moineau and colleagues highlighted the importance of PAM sequences by demonstrating that PAM mutations in phage genomes circumvented CRISPR interference (Deveau et al., 2008). Additionally, for types I and II, the lack of PAM within the direct repeat sequence within the CRISPR array prevents self-targeting by the CRISPR system. In type III systems, however, mismatches between the 5′ end of the crRNA and the DNA target are required for plasmid interference (Marraffini and Sontheimer, 2010). …..

In 2013, a pair of studies simultaneously showed how to successfully engineer type II CRISPR systems from Streptococcus thermophilus (Cong et al., 2013) andStreptococcus pyogenes (Cong et al., 2013 Mali et al., 2013a) to accomplish genome editing in mammalian cells. Heterologous expression of mature crRNA-tracrRNA hybrids (Cong et al., 2013) as well as sgRNAs (Cong et al., 2013 Mali et al., 2013a) directs Cas9 cleavage within the mammalian cellular genome to stimulate NHEJ or HDR-mediated genome editing. Multiple guide RNAs can also be used to target several genes at once. Since these initial studies, Cas9 has been used by thousands of laboratories for genome editing applications in a variety of experimental model systems (Sander and Joung, 2014). ……

The majority of CRISPR-based technology development has focused on the signature Cas9 nuclease from type II CRISPR systems. However, there remains a wide diversity of CRISPR types and functions. Cas RAMP module (Cmr) proteins identified in Pyrococcus furiosus and Sulfolobus solfataricus (Hale et al., 2012) constitute an RNA-targeting CRISPR immune system, forming a complex guided by small CRISPR RNAs that target and cleave complementary RNA instead of DNA. Cmr protein homologs can be found throughout bacteria and archaea, typically relying on a 5 ′ site tag sequence on the target-matching crRNA for Cmr-directed cleavage.

Unlike RNAi, which is targeted largely by a 6 nt seed region and to a lesser extent 13 other bases, Cmr crRNAs contain 30–40 nt of target complementarity. Cmr-CRISPR technologies for RNA targeting are thus a promising target for orthogonal engineering and minimal off-target modification. Although the modularity of Cmr systems for RNA-targeting in mammalian cells remains to be investigated, Cmr complexes native to P. furiosus have already been engineered to target novel RNA substrates (Hale et al., 2009, 2012). ……

Although Cas9 has already been widely used as a research tool, a particularly exciting future direction is the development of Cas9 as a therapeutic technology for treating genetic disorders. For a monogenic recessive disorder due to loss-of-function mutations (such as cystic fibrosis, sickle-cell anemia, or Duchenne muscular dystrophy), Cas9 may be used to correct the causative mutation. This has many advantages over traditional methods of gene augmentation that deliver functional genetic copies via viral vector-mediated overexpression—particularly that the newly functional gene is expressed in its natural context. For dominant-negative disorders in which the affected gene is haplosufficient (such as transthyretin-related hereditary amyloidosis or dominant forms of retinitis pigmentosum), it may also be possible to use NHEJ to inactivate the mutated allele to achieve therapeutic benefit. For allele-specific targeting, one could design guide RNAs capable of distinguishing between single-nucleotide polymorphism (SNP) variations in the target gene, such as when the SNP falls within the PAM sequence.

CRISPR/Cas9: a powerful genetic engineering tool for establishing large animal models of neurodegenerative diseases

Zhuchi Tu, Weili Yang, Sen Yan, Xiangyu Guo and Xiao-Jiang Li

Molecular Neurodegeneration 2015 10:35

Animal models are extremely valuable to help us understand the pathogenesis of neurodegenerative disorders and to find treatments for them. Since large animals are more like humans than rodents, they make good models to identify the important pathological events that may be seen in humans but not in small animals large animals are also very important for validating effective treatments or confirming therapeutic targets. Due to the lack of embryonic stem cell lines from large animals, it has been difficult to use traditional gene targeting technology to establish large animal models of neurodegenerative diseases. Recently, CRISPR/Cas9 was used successfully to genetically modify genomes in various species. Here we discuss the use of CRISPR/Cas9 technology to establish large animal models that can more faithfully mimic human neurodegenerative diseases.

Neurodegenerative diseases — Alzheimer’s disease(AD),Parkinson’s disease(PD), amyotrophic lateral sclerosis (ALS), Huntington’s disease (HD), and frontotemporal dementia (FTD) — are characterized by age-dependent and selective neurodegeneration. As the life expectancy of humans lengthens, there is a greater prevalence of these neurodegenerative diseases however, the pathogenesis of most of these neurodegenerative diseases remain unclear, and we lack effective treatments for these important brain disorders.

CRISPR/Cas9, Non-human primates, Neurodegenerative diseases, Animal model

There are a number of excellent reviews covering different types of neurodegenerative diseases and their genetic mouse models [8–12]. Investigations of different mouse models of neurodegenerative diseases have revealed a common pathology shared by these diseases. First, the development of neuropathology and neurological symptoms in genetic mouse models of neurodegenerative diseases is age dependent and progressive. Second, all the mouse models show an accumulation of misfolded or aggregated proteins resulting from the expression of mutant genes. Third, despite the widespread expression of mutant proteins throughout the body and brain, neuronal function appears to be selectively or preferentially affected. All these facts indicate that mouse models of neurodegenerative diseases recapitulate important pathologic features also seen in patients with neurodegenerative diseases.

However, it seems that mouse models can not recapitulate the full range of neuropathology seen in patients with neurodegenerative diseases. Overt neurodegeneration, which is the most important pathological feature in patient brains, is absent in genetic rodent models of AD, PD, and HD. Many rodent models that express transgenic mutant proteins under the control of different promoters do not replicate overt neurodegeneration, which is likely due to their short life spans and the different aging processes of small animals. Also important are the remarkable differences in brain development between rodents and primates. For example, the mouse brain takes 21 days to fully develop, whereas the formation of primate brains requires more than 150 days [13]. The rapid development of the brain in rodents may render neuronal cells resistant to misfolded protein-mediated neurodegeneration. Another difficulty in using rodent models is how to analyze cognitive and emotional abnormalities, which are the early symptoms of most neurodegenerative diseases in humans. Differences in neuronal circuitry, anatomy, and physiology between rodent and primate brains may also account for the behavioral differences between rodent and primate models.

Mitochondrial dynamics–fusion, fission, movement, and mitophagy–in neurodegenerative diseases

Hsiuchen Chen and David C. Chan
Human Molec Gen 2009 18, Review Issue 2 R169–R176

Neurons are metabolically active cells with high energy demands at locations distant from the cell body. As a result, these cells are particularly dependent on mitochondrial function, as reflected by the observation that diseases of mitochondrial dysfunction often have a neurodegenerative component. Recent discoveries have highlighted that neurons are reliant particularly on the dynamic properties of mitochondria. Mitochondria are dynamic organelles by several criteria. They engage in repeated cycles of fusion and fission, which serve to intermix the lipids and contents of a population of mitochondria. In addition, mitochondria are actively recruited to subcellular sites, such as the axonal and dendritic processes of neurons. Finally, the quality of a mitochondrial population is maintained through mitophagy, a form of autophagy in which defective mitochondria are selectively degraded. We review the general features of mitochondrial dynamics, incorporating recent findings on mitochondrial fusion, fission, transport and mitophagy. Defects in these key features are associated with neurodegenerative disease. Charcot-Marie-Tooth type 2A, a peripheral neuropathy, and dominant optic atrophy, an inherited optic neuropathy, result from a primary deficiency of mitochondrial fusion. Moreover, several major neurodegenerative diseases—including Parkinson’s, Alzheimer’s and Huntington’s disease—involve disruption of mitochondrial dynamics. Remarkably, in several disease models, the manipulation of mitochondrial fusion or fission can partially rescue disease phenotypes. We review how mitochondrial dynamics is altered in these neurodegenerative diseases and discuss the reciprocal interactions between mitochondrial fusion, fission, transport and mitophagy.

Applications of CRISPR–Cas systems in Neuroscience

Genome-editing tools, and in particular those based on CRISPR–Cas (clustered regularly interspaced short palindromic repeat (CRISPR)–CRISPR-associated protein) systems, are accelerating the pace of biological research and enabling targeted genetic interrogation in almost any organism and cell type. These tools have opened the door to the development of new model systems for studying the complexity of the nervous system, including animal models and stem cell-derived in vitro models. Precise and efficient gene editing using CRISPR–Cas systems has the potential to advance both basic and translational neuroscience research.
Cellular neuroscience, DNA recombination, Genetic engineering, Molecular neuroscience

Figure 3: In vitro applications of Cas9 in human iPSCs.close

a | Evaluation of disease candidate genes from large-population genome-wide association studies (GWASs). Human primary cells, such as neurons, are not easily available and are difficult to expand in culture. By contrast, induced pluripo…

Molecular Therapy 12 Jan 2016

Scientific Reports 31 Mar 2016

Scientific Reports 12 Nov 2015

Alzheimer’s Disease: Medicine’s Greatest Challenge in the 21st Century

The development of the CRISPR/Cas9 system has made gene editing a relatively simple task. While CRISPR and other gene editing technologies stand to revolutionize biomedical research and offers many promising therapeutic avenues (such as in the treatment of HIV), a great deal of debate exists over whether CRISPR should be used to modify human embryos. As I discussed in my previous Insight article, we lack enough fundamental biological knowledge to enhance many traits like height or intelligence, so we are not near a future with genetically-enhanced super babies. However, scientists have identified a few rare genetic variants that protect against disease. One such protective variant is a mutation in the APP gene that protects against Alzheimer’s disease and cognitive decline in old age. If we can perfect gene editing technologies, is this mutation one that we should be regularly introducing into embryos? In this article, I explore the potential for using gene editing as a way to prevent Alzheimer’s disease in future generations. Alzheimer’s Disease: Medicine’s Greatest Challenge in the 21st Century Can gene editing be the missing piece in the battle against Alzheimer’s? (Source: I chose to assess the benefit of germline gene editing in the context of Alzheimer’s disease because this disease is one of the biggest challenges medicine faces in the 21st century. Alzheimer’s disease is a chronic neurodegenerative disease responsible for the majority of the cases of dementia in the elderly. The disease symptoms begins with short term memory loss and causes more severe symptoms – problems with language, disorientation, mood swings, behavioral issues – as it progresses, eventually leading to the loss of bodily functions and death. Because of the dementia the disease causes, Alzheimer’s patients require a great deal of care, and the world spends

1% of its total GDP on caring for those with Alzheimer’s and related disorders. Because the prevalence of the disease increases with age, the situation will worsen as life expectancies around the globe increase: worldwide cases of Alzheimer’s are expected to grow from 35 million today to over 115 million by 2050.

Despite much research, the exact causes of Alzheimer’s disease remains poorly understood. The disease seems to be related to the accumulation of plaques made of amyloid-β peptides that form on the outside of neurons, as well as the formation of tangles of the protein tau inside of neurons. Although many efforts have been made to target amyloid-β or the enzymes involved in its formation, we have so far been unsuccessful at finding any treatment that stops the disease or reverses its progress. Some researchers believe that most attempts at treating Alzheimer’s have failed because, by the time a patient shows symptoms, the disease has already progressed past the point of no return.

While research towards a cure continues, researchers have sought effective ways to prevent Alzheimer’s disease. Although some studies show that mental and physical exercise may lower ones risk of Alzheimer’s disease, approximately 60-80% of the risk for Alzheimer’s disease appears to be genetic. Thus, if we’re serious about prevention, we may have to act at the genetic level. And because the brain is difficult to access surgically for gene therapy in adults, this means using gene editing on embryos.

Utilising CRISPR to Generate Predictive Disease Models: a Case Study in Neurodegenerative Disorders

Dr. Bhuvaneish.T. Selvaraj – Scottish Centre for Regenerative Medicine

  • Introducing the latest developments in predictive model generation
  • Discover how CRISPR is being used to develop disease models to study and treat neurodegenerative disorders
  • In depth Q&A session to answer your most pressing questions

Turning On Genes, Systematically, with CRISPR/Cas9

Scientists based at MIT assert that they can reliably turn on any gene of their choosing in living cells. [Feng Zhang and Steve Dixon]

With the latest CRISPR/Cas9 advance, the exhortation “turn on, tune in, drop out” comes to mind. The CRISPR/Cas9 gene-editing system was already a well-known means of “tuning in” (inserting new genes) and “dropping out” (knocking out genes). But when it came to “turning on” genes, CRISPR/Cas9 had little potency. That is, it had demonstrated only limited success as a way to activate specific genes.

A new CRISPR/Cas9 approach, however, appears capable of activating genes more effectively than older approaches. The new approach may allow scientists to more easily determine the function of individual genes, according to Feng Zhang, Ph.D., a researcher at MIT and the Broad Institute. Dr. Zhang and colleagues report that the new approach permits multiplexed gene activation and rapid, large-scale studies of gene function.

The new technique was introduced in the December 10 online edition of Nature, in an article entitled, “Genome-scale transcriptional activation by an engineered CRISPR-Cas9 complex.” The article describes how Dr. Zhang, along with the University of Tokyo’s Osamu Nureki, Ph.D., and Hiroshi Nishimasu, Ph.D., overhauled the CRISPR/Cas9 system. The research team based their work on their analysis (published earlier this year) of the structure formed when Cas9 binds to the guide RNA and its target DNA. Specifically, the team used the structure’s 3D shape to rationally improve the system.

In previous efforts to revamp CRISPR/Cas9 for gene activation purposes, scientists had tried to attach the activation domains to either end of the Cas9 protein, with limited success. From their structural studies, the MIT team realized that two small loops of the RNA guide poke out from the Cas9 complex and could be better points of attachment because they allow the activation domains to have more flexibility in recruiting transcription machinery.

Using their revamped system, the researchers activated about a dozen genes that had proven difficult or impossible to turn on using the previous generation of Cas9 activators. Each gene showed at least a twofold boost in transcription, and for many genes, the researchers found multiple orders of magnitude increase in activation.

After investigating single-guide RNA targeting rules for effective transcriptional activation, demonstrating multiplexed activation of 10 genes simultaneously, and upregulating long intergenic noncoding RNA transcripts, the research team decided to undertake a large-scale screen. This screen was designed to identify genes that confer resistance to a melanoma drug called PLX-4720.

“We … synthesized a library consisting of 70,290 guides targeting all human RefSeq coding isoforms to screen for genes that, upon activation, confer resistance to a BRAF inhibitor,” wrote the authors of the Nature paper. “The top hits included genes previously shown to be able to confer resistance, and novel candidates were validated using individual [single-guide RNA] and complementary DNA overexpression.”

A gene signature based on the top screening hits, the authors added, correlated with a gene expression signature of BRAF inhibitor resistance in cell lines and patient-derived samples. It was also suggested that large-scale screens such as the one demonstrated in the current study could help researchers discover new cancer drugs that prevent tumors from becoming resistant.

Familial amyloid polyneuropathy type I is an autosomal dominant disorder caused by mutations in the transthyretin (TTR ) gene however, carriers of the same mutation exhibit variability in penetrance and clinical expression. We analyzed alleles of candidate genes encoding non-fibrillar components of TTR amyloid deposits and a molecule metabolically interacting with TTR [retinol-binding protein (RBP)], for possible associations with age of disease onset and/or susceptibility in a Portuguese population sample with the TTR V30M mutation and unrelated controls. We show that the V30M carriers represent a distinct subset of the Portuguese population. Estimates of genetic distance indicated that the controls and the classical onset group were furthest apart, whereas the late-onset group appeared to differ from both. Importantly, the data also indicate that genetic interactions among the multiple loci evaluated, rather than single-locus effects, are more likely to determine differences in the age of disease onset. Multifactor dimensionality reduction indicated that the best genetic model for classical onset group versus controls involved the APCS gene, whereas for late-onset cases, one APCS variant (APCSv1) and two RBP variants (RBPv1 and RBPv2) are involved. Thus, although the TTR V30M mutation is required for the disease in Portuguese patients, different genetic factors may govern the age of onset, as well as the occurrence of anticipation.

Autosomal dominant disorders may vary in expression even within a given kindred. The basis of this variability is uncertain and can be attributed to epigenetic factors, environment or epistasis. We have studied familial amyloid polyneuropathy (FAP), an autosomal dominant disorder characterized by peripheral sensorimotor and autonomic neuropathy. It exhibits variation in cardiac, renal, gastrointestinal and ocular involvement, as well as age of onset. Over 80 missense mutations in the transthyretin gene (TTR ) result in autosomal dominant disease

mjsaraiv/ttrmut.html). The presence of deposits consisting entirely of wild-type TTR molecules in the hearts of 10– 25% of individuals over age 80 reveals its inherent in vivo amyloidogenic potential (1).

FAP was initially described in Portuguese (2) where, until recently, the TTR V30M has been the only pathogenic mutation associated with the disease (3,4). Later reports identified the same mutation in Swedish and Japanese families (5,6). The disorder has since been recognized in other European countries and in North American kindreds in association with V30M, as well as other mutations (7).

TTR V30M produces disease in only 5–10% of Swedish carriers of the allele (8), a much lower degree of penetrance than that seen in Portuguese (80%) (9) or in Japanese with the same mutation. The actual penetrance in Japanese carriers has not been formally established, but appears to resemble that seen in Portuguese. Portuguese and Japanese carriers show considerable variation in the age of clinical onset (10,11). In both populations, the first symptoms had originally been described as typically occurring before age 40 (so-called ‘classical’ or early-onset) however, in recent years, more individuals developing symptoms late in life have been identified (11,12). Hence, present data indicate that the distribution of the age of onset in Portuguese is continuous, but asymmetric with a mean around age 35 and a long tail into the older age group (Fig. 1) (9,13). Further, DNA testing in Portugal has identified asymptomatic carriers over age 70 belonging to a subset of very late-onset kindreds in whose descendants genetic anticipation is frequent. The molecular basis of anticipation in FAP, which is not mediated by trinucleotide repeat expansions in the TTR or any other gene (14), remains elusive.

Variation in penetrance, age of onset and clinical features are hallmarks of many autosomal dominant disorders including the human TTR amyloidoses (7). Some of these clearly reflect specific biological effects of a particular mutation or a class of mutants. However, when such phenotypic variability is seen with a single mutation in the gene encoding the same protein, it suggests an effect of modifying genetic loci and/or environmental factors contributing differentially to the course of disease. We have chosen to examine age of onset as an example of a discrete phenotypic variation in the presence of the particular autosomal dominant disease-associated mutation TTR V30M. Although the role of environmental factors cannot be excluded, the existence of modifier genes involved in TTR amyloidogenesis is an attractive hypothesis to explain the phenotypic variability in FAP. ….

ATTR (TTR amyloid), like all amyloid deposits, contains several molecular components, in addition to the quantitatively dominant fibril-forming amyloid protein, including heparan sulfate proteoglycan 2 (HSPG2 or perlecan), SAP, a plasma glycoprotein of the pentraxin family (encoded by the APCS gene) that undergoes specific calcium-dependent binding to all types of amyloid fibrils, and apolipoprotein E (ApoE), also found in all amyloid deposits (15). The ApoE4 isoform is associated with an increased frequency and earlier onset of Alzheimer’s disease (Ab), the most common form of brain amyloid, whereas the ApoE2 isoform appears to be protective (16). ApoE variants could exert a similar modulatory effect in the onset of FAP, although early studies on a limited number of patients suggested this was not the case (17).

In at least one instance of senile systemic amyloidosis, small amounts of AA-related material were found in TTR deposits (18). These could reflect either a passive co-aggregation or a contributory involvement of protein AA, encoded by the serum amyloid A (SAA ) genes and the main component of secondary (reactive) amyloid fibrils, in the formation of ATTR.

Retinol-binding protein (RBP), the serum carrier of vitamin A, circulates in plasma bound to TTR. Vitamin A-loaded RBP and L-thyroxine, the two natural ligands of TTR, can act alone or synergistically to inhibit the rate and extent of TTR fibrillogenesis in vitro, suggesting that RBP may influence the course of FAP pathology in vivo (19). We have analyzed coding and non-coding sequence polymorphisms in the RBP4 (serum RBP, 10q24), HSPG2 (1p36.1), APCS (1q22), APOE (19q13.2), SAA1 and SAA2 (11p15.1) genes with the goal of identifying chromosomes carrying common and functionally significant variants. At the time these studies were performed, the full human genome sequence was not completed and systematic singlenucleotide polymorphism (SNP) analyses were not available for any of the suspected candidate genes. We identified new SNPs in APCS and RBP4 and utilized polymorphisms in SAA, HSPG2 and APOE that had already been characterized and shown to have potential pathophysiologic significance in other disorders (16,20–22). The genotyping data were analyzed for association with the presence of the V30M amyloidogenic allele (FAP patients versus controls) and with the age of onset (classical- versus late-onset patients). Multilocus analyses were also performed to examine the effects of simultaneous contributions of the six loci for determining the onset of the first symptoms. …..

The potential for different underlying models for classical and late onset is supported by the MDR analysis, which produces two distinct models when comparing each class with the controls. One could view the two onset classes as unique diseases. If this is the case, then the failure to detect a single predictive genetic model is consistent with two related, but different, diseases. This is exactly what would be expected in such a case of genetic heterogeneity (28). Using this approach, a major gene effect can be viewed as a necessary, but not sufficient, condition to explain the course of the disease. Analyzing the cases but omitting from the analysis of phenotype the necessary allele, in this case TTR V30M, can then reveal a variety of important modifiers that are distinct between the phenotypes.

The significant comparisons obtained in our study cohort indicate that the combined effects mainly result from two and three-locus interactions involving all loci except SAA1 and SAA2 for susceptibility to disease. A considerable number of four-site combinations modulate the age of onset with SAA1 appearing in a majority of significant combinations in late-onset disease, perhaps indicating a greater role of the SAA variants in the age of onset of FAP.

The correlation between genotype and phenotype in socalled simple Mendelian disorders is often incomplete, as only a subset of all mutations can reliably predict specific phenotypes (34). This is because non-allelic genetic variations and/or environmental influences underlie these disorders whose phenotypes behave as complex traits. A few examples include the identification of the role of homozygozity for the SAA1.1 allele in conferring the genetic susceptibility to renal amyloidosis in FMF (20) and the association of an insertion/deletion polymorphism in the ACE gene with disease severity in familial hypertrophic cardiomyopathy (35). In these disorders, the phenotypes arise from mutations in MEFV and b-MHC, but are modulated by independently inherited genetic variation. In this report, we show that interactions among multiple genes, whose products are confirmed or putative constituents of ATTR deposits, or metabolically interact with TTR, modulate the onset of the first symptoms and predispose individuals to disease in the presence of the V30M mutation in TTR. The exact nature of the effects identified here requires further study with potential application in the development of genetic screening with prognostic value pertaining to the onset of disease in the TTR V30M carriers.

If the effects of additional single or interacting genes dictate the heterogeneity of phenotype, as reflected in variability of onset and clinical expression (with the same TTR mutation), the products encoded by alleles at such loci could contribute to the process of wild-type TTR deposition in elderly individuals without a mutation (senile systemic amyloidosis), a phenomenon not readily recognized as having a genetic basis because of the insensitivity of family history in the elderly.

Safety and Efficacy of RNAi Therapy for Transthyretin Amyloidosis

Coelho T, Adams D, Silva A, et al.
N Engl J Med 2013369:819-29.

Transthyretin amyloidosis is caused by the deposition of hepatocyte-derived transthyretin amyloid in peripheral nerves and the heart. A therapeutic approach mediated by RNA interference (RNAi) could reduce the production of transthyretin.

Methods We identified a potent antitransthyretin small interfering RNA, which was encapsulated in two distinct first- and second-generation formulations of lipid nanoparticles, generating ALN-TTR01 and ALN-TTR02, respectively. Each formulation was studied in a single-dose, placebo-controlled phase 1 trial to assess safety and effect on transthyretin levels. We first evaluated ALN-TTR01 (at doses of 0.01 to 1.0 mg per kilogram of body weight) in 32 patients with transthyretin amyloidosis and then evaluated ALN-TTR02 (at doses of 0.01 to 0.5 mg per kilogram) in 17 healthy volunteers.

Results Rapid, dose-dependent, and durable lowering of transthyretin levels was observed in the two trials. At a dose of 1.0 mg per kilogram, ALN-TTR01 suppressed transthyretin, with a mean reduction at day 7 of 38%, as compared with placebo (P=0.01) levels of mutant and nonmutant forms of transthyretin were lowered to a similar extent. For ALN-TTR02, the mean reductions in transthyretin levels at doses of 0.15 to 0.3 mg per kilogram ranged from 82.3 to 86.8%, with reductions of 56.6 to 67.1% at 28 days (P<0.001 for all comparisons). These reductions were shown to be RNAi mediated. Mild-to-moderate infusion-related reactions occurred in 20.8% and 7.7% of participants receiving ALN-TTR01 and ALN-TTR02, respectively.

ALN-TTR01 and ALN-TTR02 suppressed the production of both mutant and nonmutant forms of transthyretin, establishing proof of concept for RNAi therapy targeting messenger RNA transcribed from a disease-causing gene.

Alnylam May Seek Approval for TTR Amyloidosis Rx in 2017 as Other Programs Advance

Officials from Alnylam Pharmaceuticals last week provided updates on the two drug candidates from the company’s flagship transthyretin-mediated amyloidosis program, stating that the intravenously delivered agent patisiran is proceeding toward a possible market approval in three years, while a subcutaneously administered version called ALN-TTRsc is poised to enter Phase III testing before the end of the year.

Meanwhile, Alnylam is set to advance a handful of preclinical therapies into human studies in short order, including ones for complement-mediated diseases, hypercholesterolemia, and porphyria.

The officials made their comments during a conference call held to discuss Alnylam’s second-quarter financial results.

ATTR is caused by a mutation in the TTR gene, which normally produces a protein that acts as a carrier for retinol binding protein and is characterized by the accumulation of amyloid deposits in various tissues. Alnylam’s drugs are designed to silence both the mutant and wild-type forms of TTR.

Patisiran, which is delivered using lipid nanoparticles developed by Tekmira Pharmaceuticals, is currently in a Phase III study in patients with a form of ATTR called familial amyloid polyneuropathy (FAP) affecting the peripheral nervous system. Running at over 20 sites in nine countries, that study is set to enroll up to 200 patients and compare treatment to placebo based on improvements in neuropathy symptoms.

According to Alnylam Chief Medical Officer Akshay Vaishnaw, Alnylam expects to have final data from the study in two to three years, which would put patisiran on track for a new drug application filing in 2017.

Meanwhile, ALN-TTRsc, which is under development for a version of ATTR that affects cardiac tissue called familial amyloidotic cardiomyopathy (FAC) and uses Alnylam’s proprietary GalNAc conjugate delivery technology, is set to enter Phase III by year-end as Alnylam holds “active discussions” with US and European regulators on the design of that study, CEO John Maraganore noted during the call.

In the interim, Alnylam continues to enroll patients in a pilot Phase II study of ALN-TTRsc, which is designed to test the drug’s efficacy for FAC or senile systemic amyloidosis (SSA), a condition caused by the idiopathic accumulation of wild-type TTR protein in the heart.

Based on “encouraging” data thus far, Vaishnaw said that Alnylam has upped the expected enrollment in this study to 25 patients from 15. Available data from the trial is slated for release in November, he noted, stressing that “any clinical endpoint result needs to be considered exploratory given the small sample size and the very limited duration of treatment of only six weeks” in the trial.

Vaishnaw added that an open-label extension (OLE) study for patients in the ALN-TTRsc study will kick off in the coming weeks, allowing the company to gather long-term dosing tolerability and clinical activity data on the drug.

Enrollment in an OLE study of patisiran has been completed with 27 patients, he said, and, “as of today, with up to nine months of therapy … there have been no study drug discontinuations.” Clinical endpoint data from approximately 20 patients in this study will be presented at the American Neurological Association meeting in October.

As part of its ATTR efforts, Alnylam has also been conducting natural history of disease studies in both FAP and FAC patients. Data from the 283-patient FAP study was presented earlier this year and showed a rapid progression in neuropathy impairment scores and a high correlation of this measurement with disease severity.

During last week’s conference call, Vaishnaw said that clinical endpoint and biomarker data on about 400 patients with either FAC or SSA have already been collected in a nature history study on cardiac ATTR. Maraganore said that these findings would likely be released sometime next year.

Amyloid disease drug approved

The first medication for a rare and often fatal protein misfolding disorder has been approved in Europe. On November 16, the E gave a green light to Pfizer’s Vyndaqel (tafamidis) for treating transthyretin amyloidosis in adult patients with stage 1 polyneuropathy symptoms. [Jeffery Kelly, La Jolla]

Safety and Efficacy of RNAi Therapy for Transthyretin …

The New England Journal of Medicine

Aug 29, 2013 – Transthyretin amyloidosis is caused by the deposition of hepatocyte-derived transthyretin amyloid in peripheral nerves and the heart.

Alnylam’s RNAi therapy targets amyloid disease

RNA interference’s silencing of target genes could result in potent therapeutics.

The most clinically advanced RNA interference (RNAi) therapeutic achieved a milestone in April when Alnylam Pharmaceuticals in Cambridge, Massachusetts, reported positive results for patisiran, a small interfering RNA (siRNA) oligonucleotide targeting transthyretin for treating familial amyloidotic polyneuropathy (FAP). …

Nature Biotechnology 11 April 2016

Nature Biotechnology 02 March 2014

Nature Biotechnology 04 April 2016

Tafamidis for Transthyretin Familial Amyloid Polyneuropathy: A Randomized, Controlled Trial. … Multiplex Genome Engineering Using CRISPR/Cas Systems.

Is CRISPR a Solution to Familial Amyloid Polyneuropathy?

Author and Curator: Larry H. Bernstein, MD, FCAP

Originally published as

FAP is characterized by the systemic deposition of amyloidogenic variants of the transthyretin protein, especially in the peripheral nervous system, causing a progressive sensory and motor polyneuropathy.

FAP is caused by a mutation of the TTR gene, located on human chromosome 18q12.1-11.2.[5] A replacement of valine by methionine at position 30 (TTR V30M) is the mutation most commonly found in FAP.[1] The variant TTR is mostly produced by the liver.[citation needed] The transthyretin protein is a tetramer. ….

CRISPR-Cas9–mediated generation of cells and animal models

CRISPR-based genome engineering technology has facilitated the rapid generation of alternative in vivo and in vitro disease models. The new alternatives include the following: (i) Genome editing in single-cell embryos via direct injection of sgRNAs and Cas9 mRNA. This approach has been used successfully to generate mouse [ 58], rat [ 59] and monkey models [ 60], thus revealing the full potential of the CRISPR-Cas9 system for efficient and quick creation of genetically modified animals in which one or several genes have been simultaneously altered. (ii) In vivo gene editing, which involves direct delivery of the CRISPR-Cas9 system to specific cells in their native tissues, thus bypassing the need for germline-modified mutant strains. This alternative can be applied to existing disease models and transgenic strains and has promising applicability in gene therapy strategies. In vivo edition has been achieved through the use of viral vectors, mainly adeno-associated viruses (AAVs), with defined tissue-specific tropism [ 61, 62]. (iii) Combination of gene editing with human iPSCs, which enables the generation of models of genetically complex disorders. Using this approach, it is possible to study human genome alterations in various genetic backgrounds. iPSCs from patients can be differentiated in culture to identify disease-affected cells that can be used to recapitulate disease pathogenesis in vitro. CRISPR can be used to revert targeted mutations in iPSCs from individuals with disease, thus elucidating the effect of such mutations and providing proof-of-principle for gene therapy.


Plants play a vital role in human life by offering a variety of plant-based products such as fruits, food grains, vegetables, and medicine. The various traits of plants can be improved by plant breeding and genetic engineering activities [1]. Plant genetic engineering can be achieved utilizing multiple tools such as overexpression, RNA interference, Zinc finger TALEN nuclease, and CRISPR/Cas9 [2,3,4,5]. CRISPR/Cas9 genome editing tool is derived from the bacterial CRISPR system, which is known to involve in the immune system. CRISPR/Cas9 technology has gained tremendous popularity due to its specificity and efficiency in editing the genome. Several CRISPR/Cas9 and its variants have been applied for the genome editing of many desired genes. The CRISPR (clustered regularly interspaced short palindromic repeats) locus and its associated proteins are basically found in few bacteria, and it is related to immunity against phages. The different regions of CRISPR locus transcribed and lead to the formation of CRISPR RNA (crRNA) and trans-activating CRISPR RNA (tracrRNA). The crRNA, tracrRNA, and Cas9 encounter the phage DNA. The guide RNA (gRNA) is a synthetic gene comprised of crRNA and tracrRNA [6].

The Cas9 gene and gRNA under the regulation of the appropriated promoters within any vector, can be delivered into the plant cells. In another approach, the Cas9, also known as RNA guided site-specific nucleases (RGNs) and transcribed gRNA, is assembled and then delivered into plant regenerative tissue. The target site must contain 5′NGG3′ for the action of the CRISPR/Cas9 system, which is also known as Protospacer Adjacent Motif (PAM). Cas9 cleaves both the strands of a target gene or DNA with the help of gRNA. This double-stranded breaks (DSBs) may be restored by either homologous direct repair (HDR) or non-homologous end joining (NHEJ) via a repair mechanism [7]. During DNA repair, the insertion or deletion of nucleotide results in the point mutation or frameshift mutation. These mutations are generally identified by various techniques however, the restriction enzyme site loss assay, AFLP, and Sanger-based sequencing are frequently used [8, 9].

Different strategies have been developed to target multiple genes at a time, i.e., multiplexing [10, 11]. This multiplex genome engineering is generally used to target various genes within the genome or distinct target within one gene to increase mutation efficiency. This technique involves the expression of multiple gRNAs under different promoters or single promoter using the polycistronic gRNA unit [11]. Polycistronic gRNA unit is assembled with the help of either tRNA-gRNA or Cys4-gRNA. This single synthetic gene is transcribed by a single promoter. The RNases P and Z enzymes cleave polycistronic tRNA-gRNA. Csy4 (CRISPR system yersinia 4) is an RNA nuclease characterized from Pseudomonas aeruginosa separate polycistronic Cys4-gRNA into individual gRNA [11]. The tRNA processing enzymes are naturally present in almost all living organisms, including plant cells [12]. This technology has been demonstrated and applied in various plants such as Arabidopsis, tobacco, potato, tomato, rice, wheat, and banana [4, 5, 13, 14]. Targeting various genes by employing CRISPR/Cas9 is a more relaxed approach in comparison to the other known genome modification tools. Therefore, it is considered as a most promising tool for metabolic engineering.

The delivery of CRISPR/Cas9 components within rigid plant cells is a tough task. There are three methods of the construct delivery in plant cell: PEG mediated, Agrobacterium-mediated transformation, and bombardment or biolistic transformation. However, we insights the strength and weaknesses of each method of delivery depend upon plant species. We have elaborately discussed two potential methods for CRISPR/Cas9 vector-nanoparticle complex and a novel pollen magnetofection-mediated delivery in plants that would be most useful shortly (Fig. 1).

Existing and potential future CRISPR/Cas9 delivery methods. Different well-known delivery methods such as Agrobacterium-mediated delivery, Bombardment-mediated delivery, PEG-mediated delivery, and floral dip or pollen-tube tube pathway method. Potential pollen magnetofection-mediated delivery and nanoparticle-mediated delivery will be useful in near future to avoid tissue culture

Electroporation-Mediated Knockouts

The primary method for DSB repair in gametes and the early zygote is the NHEJ pathway (Rothkamm et al., 2003). Multiple studies in numerous species have used electroporation to deliver CRISPR Cas9 genome-editing reagents into zygotes to generate knockout embryos and animals. Non-mosaic knockouts have been most efficiently produced in rats and mice (Hashimoto et al., 2016 Chen et al., 2019) targeting a wide range of genes, including LIF (Kim et al., 2020), Rad51 (Iwata et al., 2019), and Rosa26 (Troder et al., 2018).

As previously noted in the poring voltage section, Kaneko et al. (2014) was one of the first to optimize electroporation conditions for rat embryos and successfully generated knockout embryos with a 9% mutation rate. Qin et al. (2015) was able to target 10 different genes in mice and generate 10 different knockout mice with mutation rates from 13 to 100% (Kaneko et al., 2014). Another study published in 2019 utilized Cas12a instead of Cas9 as the nuclease, and targeted three different genes with electroporation. The authors found knockout mutation rates in mouse embryos ranged from 34 to 70% (Dumeau et al., 2019). Unfortunately, mosaicism rates were not studied. More recently, Kaneko explored the possibility of electroporating frozen-warmed pronuclear-stage embryos to generate Tyr knockout mice (Nakagawa et al., 2018) and rats (Kaneko and Nakagawa, 2020) using Cas9 protein and dual sgRNA introduced by electroporation after slow freezing. This same group used a combination with electroporation of Cas9 protein and gRNA into rat oocytes following intracytoplasmic sperm injection (ICSI) of frozen or freeze-dried sperm to produce 56 and 50% genome edited offspring for frozen and freeze-dried sperm, respectively (Nakagawa and Kaneko, 2019).

There are currently only a handful of studies describing the generation of live genome edited livestock following electroporation of editing reagents. To date, only porcine and bovine zygotes have been successfully electroporated to produce knockout live animals. Pig researchers have electroporated zygotes and oocytes to generate genome edited blastocysts and live piglets using Cas9 genome editing reagents. A group led by Tanihara has published six studies describing the electroporation of porcine zygotes and efficient editing of blastocysts with at least an 80% success rate in all six studies. They also produced live knockout piglets in three of the studies. The first of the six studies targeted the MSTN gene using five 1 ms pulses at a voltage of 30 V/mm and generated 10 piglets. Nine of the 10 piglets expressed mutations at the target site, seven of which were mosaic. The next study targeted the TP53 gene using the same parameters which resulted in nine piglets, six of which were genetic knockouts. However, four out of the six mutated piglets were mosaic individuals, a less than ideal outcome if electroporation is to be widely used for the generation of genetically modified livestock (Tanihara et al., 2018). A third study utilized the same parameters again to produce PDX1 knockout blastocysts, and achieved a success rate of up to 94.1%. That same study also attempted to generate PDX1 knockout fetuses, however, only one fetus was collected, and it did not carry genetic mutations at the target site (Tanihara et al., 2019c). A subsequent study re-attempted to generate PDX1 knockout piglets and was successful in producing 10 piglets, nine of which contained the intended knockout. Two of nine piglets with the intended mutations contained no wild-type sequences and another two were mosaic (Tanihara et al., 2020b).

The next porcine study targeted the CD163 gene with slightly different parameters, using 25 V/mm instead of 30 V/mm, and was able to successfully produce edited blastocysts with a 90% success rate as well as eight piglets, one of which showed a mutation at the intended target (Tanihara et al., 2019d). These studies were able to successfully generate edited blastocysts and piglets, however, up to 40% of the CD163 blastocysts, four TP53 piglets, and seven MSTN piglets were mosaic. In 2020, this group successfully knocked out (Le et al., 2020 Tanihara et al., 2020a) MSTN and GGTA-1 using electroporation at 12 hpi with five 1 ms transfer pulses at 25V/mm. Five out of six piglets born in the GGTA1 study carried a bi-allelic mutation in the targeted region of GGTA1, with no off-target events (Tanihara et al., 2020a).

Another study published in 2020 attempted to address the issue of generating mostly mosaic mutants through the co-transfection of a three-prime repair exonuclease (Trex2), an exonuclease known to digest DNA ends with breaks. The authors claim to have increased the production of non-mosaic blastocysts by 70.7% when Trex2 was co-transfected with Cas9. Unfortunately, Trex2 is a known inhibitor of HDR which may result in problems if attempting to generate non-mosaic knock-in animals (Yamashita et al., 2020).

Two studies used electroporation to introduce multiple gRNAs to target more than one gene in porcine zygotes. Double bi-allelic mutations were obtained when targeting two genes, although at a low frequency (0�%) depending upon the gRNA combination (Hirata et al., 2020b). Another study by this group targeted four genes simultaneously. Guides for each gene were first tested independently, and the best guide for each gene was combined to target the four loci. Mutations were observed in one (55.8%) and two genes (20.9%), and no blastocysts had mutations in three or more target genes. This was despite the fact that each guide had independently achieved a rate of at least

20% bi-allelic mutations in blastocysts. The majority of the blastocysts were mosaic. Bi-allelic knockouts were identified in six of the 43 (14%) blastocysts in one of the four genes, and none of these contained edits in a second gene. It is possible that larger than expected deletions or translocations may have occurred that were not detected by the screening methods being used in this study. The authors concluded that the technique to deliver gRNA and Cas9 protein to edit multiple genes will require considerable optimization to improve the success rates (Hirata et al., 2020a).

Miao et al. (2019) published a study describing electroporation of Cas9 protein with gRNA targeting the Nanos2 gene in mice, pigs, and cattle. They were successful in generating knockout embryos for all three species, and pups in mice. They found that the optimal voltage strengths for efficient survival and editing rates were 20 V/mm for bovine and 30 V/mm for mice and porcine. Analysis of mouse embryos and pups found that two cell embryos were 90% mutated and 70% of pups had a Nanos2 mutation. Analysis of bovine and porcine embryos revealed bi-allelic Nanos2 edits at a rate of 82 and 73%, respectively. Some of these knockout Nanos2 bovine embryos were brought to term, and two calves were born alive, and one was stillborn (Ciccarelli et al., 2020). The stillborn and one live calf were bi-allelic knockouts, while the other live bull calf was mosaic containing both wildtype and mutated allele sequences in varying proportions depending upon the tissue analyzed. It should be noted that electroporation in this study was done at 18� hpi.


Genome Editing Technologies

Targeted Mutagenesis Using Molecular Scissors to Form DSBs

Since the discovery of restriction enzymes, the field of biotechnology has entered a new era of molecular engineering facilitated by recombinant DNA technology. Several generations of molecular scissors have been discovered, characterized and developed for DSB-based targeted genome mutagenesis. The technology has been improved from the long recognition sequence homing nucleases to protein-dependent DNA binding nucleases, such as zinc-finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs), and ultimately to the 3rd generation RNA-guided molecular scissors CRISPR /Cas (Fig. 2). With the invention of target-specific synthetic molecular scissors, the specific modification of a gene of interest in a living organism has become possible. Consequently, there are several key factors involved in targeted mutagenesis induced by molecular scissors, including: 1) the ability to specifically recognize and bind to the targeted DNA sequence, 2) effective DSB formation, and 3) error-prone DSB repair.

Four generations of molecular scissors. The first, second and third generations of molecular scissors, Homing nuclease (a) ZFN (b) and TALEN (c), are characterized as nucleases relying on DNA binding domains to recognize DNA target sites. Homing nucleases recognize long DNA sequences of 14–40 bp with their DNA binding domains. A ZFN or TALEN is designed by connecting 3–6 zinc finger motifs or 17–20 TALE modules, respectively, for DNA binding and an endonuclease domain of FokI restriction enzyme for cutting. FokI works only in homodimer form, so usually one has to design pairs of ZFNs or TALENs to target a DNA site. FokI activity usually produces DSB with 4 nt overhangs. The fourth generation, CRISPR/Cas (d), is also the most powerful one it uses guide RNA components to form active complexes, thereby interrogating and searching for target DNA sites based on Watson-Crick base pairing between the guide RNA and targeted strand. The DNA fragments and protein structures are not pictured to scale

The host’s repair of the DSB errors leads to error-free or error-prone outcomes depending on many factors, including the cell cycle state and the availability of homologous DNA templates at the damaged sites. In plant somatic cells, DSB repair by either of the two major pathways, homology-directed repair (HDR) or nonhomologous end joining (NHEJ), usually leads to either error-free or error-prone products. The majority of the error-prone products appear as insertion or deletion (indel) DNA mutations resulting from C-NHEJ or A-NHEJ (Fig. 1). A possibly lower portion of error-prone products may result from SSA repair in the absence of a homologous donor template and from Holliday junction resolution in the last steps of the double-stranded break repair (DSBR) subpathway if the DSB flanking sequences of the sister chromatids are not perfectly matched (Fig. 1). In this section, we briefly summarize the abovementioned molecular scissors. Extensive reviews of the same material can be found elsewhere (Carroll, 2011 Gaj et al., 2013).

Generation 0: Homing Nucleases

Homing nucleases are endonucleases (Mw < 40 kDa) that recognize long DNA sequences (14–40 nt) for their cutting activity (Fig. 2a). Homing nucleases can work alone as monomers or in pairs as homodimers (Chevalier and Stoddard 2001). Members of the LAGLIDADG homing endonucleases family such as I-CreI or I-SceI recognize targeted sequences of 22 bp and 18 bp respectively, thus allowing more specific targeting in the host cells (approximately once every 7 × 10 9 bp) (Jurica et al. 1998 Niu et al. 2008 Chevalier and Stoddard 2001 Jasin 1996). However, this feature also introduces limitations via the scarcity of targetable sites in the genomes of host cells. To compensate for this, researchers have engineered these nucleases for a wider range of binding and cutting sites or combinations of different homing nucleases to recognize multiple sites (Chevalier et al. 2002). Engineered homing nucleases often cleave correct sites as efficiently as wild-type nucleases (Chevalier and Stoddard 2001 Yang et al. 2009 Gao et al. 2010 D'Halluin et al. 2013). However, the engineering of homing nucleases for wider applications is still inefficient, laborious and time consuming.

Generations 1 and 2: Protein-Guided DSB Formation, ZFN and TALENs

ZFNs are derived from the discovery of the zinc finger, a finger-like DNA binding motif found in TFIIIA, a transcription factor from the eggs of Xenopus laevis (Miller et al. 1985). Its structure comprises 30 repetitive amino acid sequences and is stabilized by a zinc ion (Miller et al. 1985 Berg 1988). Berg (1988) suggested that the zinc finger protein structure might play a key role in the recognition of DNA sequences. ZFN was first developed in 1996 by fusing a nonspecific DNA cleavage domain of FokI, a type II-S restriction enzyme, to the C-terminal of the zinc finger motifs (Kim et al. 1996). Typically, three consecutive nucleotides can be specifically recognized by one zinc finger motif, and therefore, several connected zinc finger motifs fused to FokI can bind the target DNA of interest (Kim et al. 1996). ZFN is the first artificial restriction enzyme that recognizes desirable sites in the genome. Due to their binding specificity and dimerization-dependent FokI activity requirement, ZFNs were typically designed in pairs to recognize 9–18 bp using connected 3–6 zinc finger motifs on both the sense and antisense strands of the targeted sequences spaced by 5–7 bp between ZFNs (Kim et al. 1996 Bitinaite et al. 1998 Laity et al. 2001 Urnov et al. 2010) (Fig. 2b). Post cleavage, the DSB sites were recovered by DNA repair mechanisms that showed insertions or deletions at similar rates (Kim et al. 2013). However, for wider application of this technology, one should overcome the limitations of low editing efficiency (0–24%), elevated design and optimization cost, and high off-target possibility. Many efforts have been made to overcome these barriers. For example, to enhance the cleavage activity of the FokI cleavage domain, Gou and coworkers performed direct evolution to optimize a ZFN named ‘Sharkey’. Several approaches were tested to reduce the off-target effect, e.g., extending the recognition length by using more zinc finger modules (Pattanayak et al. 2011 Guo et al. 2010).

TALEN is the second-generation form of molecular scissors, discovered during studies of the plant immune system under attack and hijacking by pathogenic bacteria (Dangl and Jones 2001). AvrBs3, an effector protein secreted by the plant pathogen Xanthomonas campestris, is injected into host cells, thereby binding to the plant UPA-box gene and functioning as a transcription activator to modulate host cell gene expression for its efficient colonization (Kay et al. 2009). The causal agents secreted by Xanthomonas were identified and named transcription activator-like effectors (TALEs). TALEs have 33–35 amino acids that are highly conserved, except for those located at positions 12 and 13. These two hypervariable residues (namely, repeat-variable diresidues (RVDs)) are oriented toward the outside of the protein and play a key role in recognizing a specific nucleotide (Moscou and Bogdanove 2009). Common rules of RVD nucleotide recognition for binding were validated as NG for thymine HD for cytosine NN for guanine or adenine and NI for adenine. The first TALENs were introduced by fusing a DNA binding TAL type III effector with a FokI cleavage domain Fig. 2c (Li et al. 2011). However, unlike ZFN, which recognizes 3 bp per zinc finger module, TALENs allow more precise recognition because each RVD of TALE can recognize only one nucleotide. TALENs were designed in pairs with a 12–21 nt distance between two binding sites for the highest cutting activity (Miller et al. 2011). The combination of the TAL effectors AvrXa7, PthXo1 and FokI was demonstrated to function as molecular scissors for cutting and hence modifying the binding sites of the TALs, subsequently resulting in resistance to rice blight disease (Li et al. 2011). The initial NN RVD repeat recognized either guanine or adenine, raising concerns about its specificity (Moscou and Bogdanove 2009). Ultimately, an NK RVD repeat that recognizes only guanine was discovered, fulfilling the specificity requirement for the TALEN molecular scissors (Miller et al. 2011).

One of the weak points of the TALEN approach is the large size of the binding domain, as every nucleotide requires a repeat of

34 amino acids for binding. Thus, to assure high specificity for one TALEN binding to 20 nt, its DNA binding domain must be 680 amino acids long. In addition, assembly of the highly repeated modules was time consuming and laborious. Thus, a well-designed modular RVD repeat library was in high demand and was eventually developed (Zhang et al. 2011 Cermak et al. 2011 Kim et al. 2013). Another limitation of TALENs for practical applications is that they are sensitive to methylated cytosine, thereby preventing them from binding to the modified nucleotide efficiently. In an attempt to overcome the hurdle, the TALE domain was designed to contain the single asparagine RVD (N*) motif (N* refers to Asn instead of Asn-Gly), a base-recognition domain that could effectively bind to 5 ‘methylated cytosine. The engineered TALENs (N*) showed higher efficacies for genome editing in mammalian cells and rice (Valton et al. 2012 Kaya et al. 2017). TALENs have the advantages of high editing efficiency, low off-target activity and lower design cost than ZFNs and the drawbacks of difficult construction, no activity on methylated cytosines (Kim et al. 2013), and difficult introduction into cells owing to their large size (Kim and Ka 2015).

Generation 3: CRISPR/Cas

Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein (Cas) was shown to be a DNA interference-based defense machinery of prokaryotes such as bacteria and archaea against phage infection (Barrangou et al. 2007 Brouns et al. 2008). CRISPR/Cas systems were classified into two classes according to the number of complexity of their effector modules (Makarova et al. 2011 Makarova et al. 2015). Class 1 systems involve effector complexes formed by multiple subunits, whereas in class 2 systems, single multidomain proteins constitute the effector complexes. Furthermore, each class has been divided into several subtypes (class 1: types I, III and IV and class 2: types II, V and VI) based on their effector architectures with unique signature proteins (Koonin and Makarova 2019). Almost all of the CRISPR/Cas systems used in genome engineering to date are from class 2 due to the simplicity of their effector modules (Additional file 1: Table S1). The most widely used CRISPR/Cas systems are Cas9 and Cas12a (Cpf1).

In the native CRISPR/Cas9 system, phage DNAs were shown to be cleaved by the Cas9 effector complex, which includes the Cas9 protein as a nuclease and a complexed RNA structure formed by a CRISPR RNA (crRNA) and a trans-activating CRISPR RNA (tracrRNA) as a probe. The two-component RNA secondary structure facilitates Cas9 assembly, searching and binding to dsDNA target sites by Watson-Crick complementarity to 19–21 nt of the 5′ end of the crRNA (protospacer) and subsequently cleaving both the strands of the dsDNA at the 3rd nucleotide proximal to a 5′-NGG-3′ protospacer-adjacent motif (PAM) site (Fig. 2d, CRISPR/Cas9). Originally, crRNA:tracrRNA required maturation from a precursor crRNA:tracrRNA by RNase III processing activity, making it more difficult to apply. In the first application of CRISPR/Cas9 for genome editing, the crRNA and tracrRNA were engineered to make a single guide RNA molecule by connecting the 3′ crRNA repeat and 5′ tracrRNA anti-repeat, thereby facilitating the use of the system (Jinek et al. 2012). The Cas9 protein remains inactive until it binds to a guide crRNA:tracrRNA structure. The guide RNA-bound Cas9 complex undergoes conformational changes and then stochastically searches for potential targets by PAM scanning and binding using the PAM-interacting motif. Then, the Cas-sgRNA complex again changes conformation, and the guide RNA sequence is used to pair with the sequence located upstream of the PAM via the Watson-Crick rule (Sternberg et al. 2014 Jinek et al. 2014 Zhu et al. 2019). The gRNA and its seed sequence (10-nucleotide RNA proximal to the NGG PAM) should be fully complemented for R-loop formation and to trigger Cas9 cleavage activities via its endonuclease domains (HNH and RuvC) (Jinek et al. 2012 Jiang et al. 2013 Hsu et al. 2013). The targeted and nontargeted strands of the dsDNA are cleaved by HNH and RuvC, respectively, generating mostly blunt ends (Fig. 2d) (Anders et al. 2014 Nishimasu et al. 2014). Nickase Cas9 (nCas9) that cuts either the targeted strand or nontargeted strand and dead Cas9 (dCas9) were also created by inactivating either the endonuclease domains or both domains for alternative gene editing and regulation.

Unlike Cas9, the Cpf1 system does not require a tracrRNA to mature the crRNA and to form an effector complex for its cleavage activity. The Cpf1 protein was also shown to process the precursor crRNA (Zetsche et al. 2015). After assembly, the Cpf1 effector complex recognizes a T-rich PAM for the initiation of binding and searching for target sites. Its seed sequence was illustrated to range from 1 to 10 nt proximal to the PAM (Kim et al. 2016). The Cpf1 protein has a Nuc nuclease domain that cleaves the target strand and a RuvC domain that cleaves the nontargeted strand (Schunder et al. 2013 Makarova and Koonin 2015 Stella et al. 2017). The nuclease domains cut the target dsDNAs at the 18th nt on the nontargeted strand and the 23rd nt on the targeted strand distal to the PAM, generating 5′ overhang ends (Fig. 2d) (Zetsche et al. 2015).

Precision Editing

Base Substitution

It is now well known that the majority of genetic diseases result from point mutations, but the potential DSB-based repair approaches for correcting these mutations are not applicable due to their inaccessibility and the unsuitability of the repair mechanisms (Cox et al. 2015 Hilton and Gersbach 2015). Therefore, a single-base-change technique is highly demanded and has been developed for at least transition fixation (C/G- > T/A or A/T- > G/C): the so-called cytosine base editors (CBEs) or adenosine base editors (ABEs) (Fig. 3a) (Gaudelli et al. 2017 Komor et al. 2016). The basal principle behind the technique is the fusion of dead or nickase Cas9 (d/nCas9) with a cytosine or adenosine deaminase and introduction of the editor complex to the targeted site by the CRISPR guide RNA structure. Deamination of C or A produces U or I, respectively, leading to lesion-by-pass replication and resulting in C/G- > T/A or A/T- > G/C transition, respectively. In addition, the Cas9-based CBEs and ABEs were shown to work in a framed window that was either narrow (13th to 17th nucleotides upstream of the 5′-NGG-3′ PAM) (Komor et al. 2016) or wide (4th to 20th nucleotides upstream of the 5′-NGG-3′ PAM) (Zong et al. 2018) at asymmetric frequency distributions (Additional file 2: Table S2) depending on the types of deaminase used. This fact raises the possibility of controllably and precisely editing every single base of interest by carefully calculating and evaluating the editing frequencies of base editors for a base of a given target. This could also help to avoid the possibility of bystander base changes and unintended off-targets (Gehrke et al. 2018).

Non-DSB precise gene targeting approaches. a Base of approach editing. Cytosine Base Editors (CBEs) and Adenosine Base Editors (ABEs) are the two types of base editors that have been published so far. CBEs: Dead Cas9 (blue) binds to target C (green) via the RNA (pink) guide, which mediates the separation of local DNA strands. A tethered APOBEC1 (green) enzyme by cytosine deamination converts the single-stranded target C to U. The initial G: C is replaced by the A: T base pair at the target location through DNA repair or replication. ABEs: A hypothetical deoxyadenosine deaminase (red) and catalytically impaired nCas9 (Cas9 D10A nickase) bind target DNA in the RNA guide to expose a small bubble of single-stranded DNA that catalyzes the conversion of A to I within this bubble. b Oligonucleotide-directed mutagenesis process. A gene repair oligonucleotide (GRON), which contains designed modifications, is delivered and paired with the target DNA sequence. GRON creates a mismatch at the target site and triggers a DNA repair mechanism. DNA repair enzymes detect the mismatch and repair the target DNA sequence using GRON as a template. Once the repair process is completed during cell division and multiplication, the GRON is removed and degraded. The target sequence is modified with designed changes. The representative DNA fragments and protein structures are not pictured to scale. c Prime editing. Prime editor is a CRISPR/Cas complex developed by fusion of a reverse transcriptase (RT) to a C-terminal of nickase Cas9 (H840A) and a prime editing gRNA (pegRNA) with a 3 ‘extension that could bind to the 3 ‘nicked strands produced by the nCas9. When bound, the 3′-OH free nicked strand is used as a substratum for the RT to copy genetic information from the 3 ‘extension of pegRNA

Oligonucleotide-Directed Mutagenesis (ODM)

Oligonucleotide-directed mutagenesis (ODM) and rapid trait development system (RTDS) are two common names for an oligonucleotide-mediated targeted gene modification technique. This technique uses synthetic oligonucleotides or gene repair oligonucleotides (GRON), which function as a template for endogenous DNA repair to form a heterotriplex with a targeted genome site via homology binding using their sequences, which are identical to the site except at the intentionally modified nucleotide(s), thereby triggering gene conversion and resulting in specific base changes (Fig. 3b). The GRON itself is not inserted into the host genome, and site-directed nucleases or double-strand breaks are not required for this technique. Therefore, ODM was classified as one of the precise gene editing techniques (for review, see Sauer et al. 2016). The changes could be point mutations, multiple base changes, insertions or deletions. The GRON was subsequently degraded during cell divisions, and the modified gene retained its normal pattern of expression and stability within the genome (Sauer et al. 2016).

The first application of synthetic nucleotides was shown in yeast in 1988 (Moerschell et al. 1988) and then in mammalian cells for correction of a faulty human β-globin that causes sickle cell anemia in 1996 (Cole-Strauss 1996 Yoon et al. 1996). In plants, Beetham and coworkers used RNA/DNA chimeric molecules in a work known as the chimeraplasty approach to target tobacco acetoacetate synthase (ALS) (or aceto hydroxyl acid synthesis (AHAS)). Tobacco ALS is a biallelic gene (including alleles ALS1 and ALS2) due to its allotetraploid genome. Therefore, two chimeric ODM oligonucleotides were designed to engineer ALS1 and ALS2 as P196 (CCA) to CAA and to CTA, respectively. Particle bombardment of the oligos and subsequent selection on medium containing 200 ppb of chlorsulfuron revealed one out four ALS alleles with a Pro-196 (CCA) to Thr-196 (ACA) modification. The efficiency was two orders of magnitude higher than that of the control (Beetham et al. 1999). The ODM approach was also conducted in several studies in dicots, such as canola (Gocal 2015) and Arabidopsis (Sauer et al. 2016).

In monocots, ODM was used to target AHAS in maize (Zhu et al. 1999) and rice (Okuzaki and Toriyama 2004). In maize, nucleotide changes were induced at two sites, S621A (AGT to AAT) for imidazolinone and sulfonylurea herbicide resistance and P165A, mimicking the point mutation in tobacco in Beetham’s work (Beetham et al. 1999). The oligonucleotides were transformed into maize cells by bombardment and selected with either 7 μM imazethapyr for S612A or 20 bbp chlorsulfuron for P165A. The mutation frequencies were 1.0 × 10 − 4 to 1.4 × 10 − 4 , approximately three orders of magnitude higher than that of spontaneous mutation and gene targeting by homologous recombination pathway in plants (Tong Zhu et al. 1999). In rice, three chimeric DNA/RNA oligonucleotides for targeted modification of ALS, P171A, W548 L and S627I, were introduced into rice calli by bombardment. Screening by herbicide selection (chlorsulfuron for P171A and bispyribac-sodium for W548 L and S627I) and Sanger sequencing identified independent transformants for both P171A and W548 L but not S627I at a frequency of 1 × 10 − 4 . The ODM approach was also demonstrated in a wheat system using a transient assay with GFP as a reporter. The authors claimed that using 2,4-D in osmotic media boosted the gene targeting efficiency and that the repair of point mutations had a higher frequency than that of single base deletions in immature wheat embryos (Dong et al. 2006).

ODM products have been considered non-GMOs in a number of countries, although not in the EU, due to the targeted point mutation mechanism and transgene-free outcome (Eriksson 2018). In 2011, the UK Advisory Committee of Releases into the Environment (ACRE) suggested that plants being developed by the ODM system should not be regulated as GMOs. Afterwards, the Federal Office of Consumer Protection and Food Safety of Germany decided that ODM products do not constitute GMOs in 2017. Based on its precise modification and the GMO regulation of this technology, ODM has potential for genome editing. However, its low efficiency is the main barrier for its application in research thus, improving the editing frequency is essential. Recently, ODM and SDN have been combined to enhance the efficiency with the range of precise editing from 0.09% to 0.23% in an EPSPS target gene. This study also claimed that the transgene targeting efficiency of CRISPR/Cas9 was nearly 3 times higher than that of TALEN (Sauer et al. 2016).

HR-Based Gene Targeting

In 1988, gene targeting (GT) or HGT was first defined as modification of the host genome achieved by the integration of foreign DNA via the HR pathway (Paszkowski et al. 1988). This method provides a wide range of targeted genome modifications, such as precise insertion, deletion or replacement of a gene or an allele. In fact, HR is an ideal mechanism that can precisely repair DSBs during the S and G2 phases of the cell cycle, while homologous sequences (sister chromatids or donor templates) are available (Tamura et al. 2002). However, its low frequency in higher plants is still a hurdle for practical applications (Puchta 2005). The first application of HGT in a crop was the targeted knockout of the rice Waxy gene using positive/negative selection, which achieved approximately 1% frequency but also left a positive selection marker in the genome (Terada et al. 2002).

Since then, two other important achievements in the plant gene targeting field regarding frequency enhancement have come to light: (1) the key finding of on-target DSB roles (Puchta et al. 1993) and (2) methods to introduce high doses of autonomously homologous donor templates into targeted cells (Baltes et al. 2014). By inducing DSBs at a specific locus using the highly specific restriction enzyme I-Sce I, HDR efficiency can be enhanced from 10 to 100 times (Puchta et al. 1996). To further enhance the efficiency of HR for gene targeting, several approaches have been developed. First, site-specific nucleases such as ZFNs, TALENs and CRISPR/Cas systems are applied to induce double-strand breaks at the target sequence (Belhaj et al. 2013 Voytas 2013). The second approach takes advantage of the virus replicon system to increase the delivery ability and the number of donor templates hence, the HGT efficiency is improved (Baltes et al. 2014). Apart from that, certain studies have demonstrated that overexpression of HR-involved genes or suppression of the NHEJ pathway led to improvement of HGT frequency (Endo et al. 2016 Qi et al. 2013 Shaked et al. 2005).

Prime Editing

RNAs were shown to involve in DSB repairs via non-templated or templated mechanisms in human and yeast cells (for more details, see review by Meers et al. 2016). In addition, Butt et al. (2017) successfully engineered the SpCas9 guide RNA scaffold called chimeric single-guide RNA (cgRNA) for acting as sgRNAs and repair templates in rice protoplast. The HGT rate for the replacement of two nucleotides of OsALS locus was shown to be as high as 16.88% of total mutations when plasmids carrying the CRISPR/Cas9 and cgRNA expression cassettes were transfected to the protoplast. Further, targeted insertion of 3xHA tag at the OsHDT701 locus using a cgRNA showed up to 4.69% of total mutations. However, the HGT rates were much lower when only cgRNA-SpCas9 ribonucleoprotein (RNP) complex was transfected while the mutation rates mediated by NHEJ were much higher (Butt et al. 2017). RNA transcripts were further validated as donor templates for HDR-mediated targeting OsALS locus using CRISPR/Cpf1 ribonucleoprotein (RNP) complex. Nonetheless, the HGT frequency obtained with ssRNA donor templates was at 0.07–0.13%, nearly ten folds lower compared to that of ssDNA donors (Li et al. 2019). To expand the use of RNA as templates for plant HGT approaches, further work needs to be done.

Recently, prime editing using guide RNA extensions for priming reverse transcription-mediated precise editing has been shown to be an excellent precision genome editing technique in mammalian cell lines. It would also be an excellent alternative for HGT with a shorter editing sequence coverage. Anzalone and colleagues tested variations in prime editing methods and demonstrated a wide range of specific genetic modifications, including 19 insertions up to 44 bp 23 deletions up to 80 bp 119 point mutations, including 83 transversions and 18 hybrid edits at 12 human and mouse cell lines without explicit DSBs (Anzalone et al. 2019). The prime editor’s best version used a CRISPR / Cas complex developed by fusion of a reverse transcriptase (RT) to a C-terminal of nickase Cas9 (H840A) and a prime editing gRNA (pegRNA) with a 3 ‘extension that could bind to the 3 ‘nicked strands produced by the nCas9. When bound, the nicked strand’s free 3′-OH is used as a substratum for the RT to copy genetic information from pegRNA’s 3 ‘extension (Fig. 3c). If we design pegRNAs to produce modified nucleotides, they would be inserted into the genome during downstream repair processes. A second nick site present downstream of the first nick would support the retention of the de novo nucleotides introduced. (Fig. 3c) (Anzalone et al. 2019). Although prime editor has not yet been used in plant system, we expect this technology to have a bright future in plant genome editing, as plant HGT is much more challenging.

HGT in Monocots

HDR Mechanisms in Plants

One of the principal questions regarding cell response to DSBs is which repair consequences the cells favor: error-free or error-prone DNA products? In meiosis, error-prone crossing over (CO) or break-induced repair (BIR) (or even NHEJ) is preferred for creating genetic diversity by exchanging genetic information between parental chromosomes, a key factor for adaptation to environmental changes. However, we can expect an opposite situation in mitotic cells, which require genetic stability rather than diversity. In that case, should NHEJ be abolished from mitotic cells? The answer is absolutely not, and one of the key reasons may be the limitation of time, because a single DSB persistence may induce programmed cell death after a certain period of time (Nowsheen and Yang 2012). What can the cells do? NHEJ is so abundant and efficient in mending the broken ends. What can we expect from the bulky HDR apparatus?

HDR has been extensively studied in yeasts and mammals for understanding the mechanisms of genetic diseases caused by DNA DSB damage. Most of the components of the plant HDR pathway are homologs of these known proteins (Schuermann et al. 2005), but the regulation of DSB responses in the kingdoms may be different (Yokota et al. 2005). Unlike in animal systems, HDR efficiency in plant somatic cells is extremely low (Szostak et al. 1983 Puchta et al. 1996) and very much dominated by NHEJ. Plant mitotic HDR is absent in the G1 phase and limited to S/G2, while NHEJ is active throughout the cell cycle (Fig. 4). The HDR pathway is determined by the presence of a sister chromatid as a homologous DNA donor, which is normally produced by replication in the S phase and remains present until the M phase. Even in these favorable cell cycle phases, the HDR pathway has to compete with the predominant NHEJ, and hence, it can be chosen in only certain conditions (Heyer et al. 2010 Voytas 2013 Jasin and Rothstein 2013). Therefore, a comprehensive knowledge of the conditions that favor HDR in plant somatic cells would offer key strategies in plant gene targeting for crop improvement.

Homology-directed repair pathway determination and its favorable cell contexts. Activation of the MRN complex and P53/SOG1 triggers the activation of cell cycle checkpoint proteins such as CDKB1 (G2/M checkpoint) and CYCB1 (S phase checkpoint) or NAC-type transcription factors ANAC044 and ANAC085 (S/G2 checkpoints) or WEE1 kinase for cell cycle arrest

Sensing DSBs and Cell Cycle Arrest

In mammals, DSB formation induces cell cycle arrest, which is necessary to help the cell repair the damage in a reasonable time (Kastan and Bartek 2004). The process is initiated with the conformational changes of the ATM (ataxia telangiectasia mutated) homodimer resulting from sensing a chromatin structure change following DSB formation. Activation of human ATM by autophosphorylation of its serine 1981 disassociates its activated monomers (Bakkenist and Kastan 2003). Monomeric ATM then phosphorylates all the members of the MRE11 (meiotic recombination 11)/RAD50 (Radiation sensitive 50)/NBS1 (Nijmegen breakage syndrome 1) (MRN) complex, a DSB sensor holoenzyme, and is additionally phosphorylated by MRE11 (Lee and Paull 2005 Lamarche et al. 2010). Subsequently, ATM plays a central role in activating cell cycle checkpoint kinases and P53 and indirectly induces the suppression of cyclin-dependent kinases that ultimately leads to cell cycle arrest (Kastan and Bartek 2004 Harper and Elledge 2007 Yata and Esashi 2009). In Arabidopsis, SOG1 (SUPPRESSOR OF GAMMA RESPONSE 1), activated by ATM, is responsible for the regulation of multiple downstream proteins such as CDKB1 (G2/M checkpoint) and CYCB1 (S phase checkpoint), the NAC-type transcription factors ANAC044 and ANAC085 (S/G2 checkpoints) or WEE1 kinase for cell cycle arrest (Yoshiyama et al. 2013 Weimer et al. 2016 Takahashi et al. 2019 De Schutter et al. 2007).

HDR Pathway Determination

Post DSB formation, cell cycle arrest at S/G2 favors the essential condition for the HDR pathway (Fig. 4). In animals, the HDR pathway is determined by recruitment competition between KU70/80-DNA-PK and the MRN complex to the DSB ends and subsequent resection regulation by BRCA1/CtIP and 53BP1/RIF1, which favors HDR and NHEJ, respectively. However, only the Ku complex but not DNA-PK is conserved in plants (West et al. 2002 Tamura et al. 2002), suggesting an alternative regulation of activation by kinases in the plant kingdom. KU70 was shown to colocalize and interact with MRE11 in somatic cells and therefore was proposed to be a key player in the determination of the DSB repair pathway (Goedecke et al. 1999). Because the majority of DNA end binding proteins in a cell are KU70/80 (Gottlieb and Jackson 1993), NHEJ becomes dominant, and hence, HDR efficiency, especially in plant mitotic cells, is extremely low. Recently, it has been increasingly accepted that DSB end resection plays a key role in the determination of NHEJ- or HDR-mediated repair. NHEJ repair keeps the broken end resection in a limited range for its amendment, but HDR requires DSB end resection to produce 3′-protruding ends that are long enough for template annealing and replication of homologous genetic information. NHEJ resection length usually ranges from 0 to 14 bp, although very rare cases can be 25 bp and longer (Lieber 2010).

ATM-dependent phosphorylation of RAD50, NBS1 and MRE11 of the MRN complex plays an important role in DSB end resection and determines the ultimate repair pathway in an MRN-dependent manner. MRE11 acts as an endonuclease that nicks DNA upstream of the break and subsequently resects 3′- > 5′ toward the break, and then, the end is further resected by Endonuclease 1 and Dna 2 (Kijas et al. 2015). CtIP, activated by ATM, acts in concert with the MRN complex to enhance resection and HDR. CtIP physically interacts with the MRN complex and, more importantly, BRCA1 (Limbo et al. 2007), a protein that inactivates 53BP1 by dephosphorylation (Isono et al. 2017), thereby supporting DSB end resection for HDR determination. However, in a recent study, 53BP1 was shown to shield DSB ends from extensive resection, which might result in a strong bias toward RAD52-dependent error-prone SSA (Ochs et al. 2016). Broken end resection is also controlled by phosphorylated MRE11, which protects exonuclease 1 from extensive resection by phosphorylating it (Kijas et al. 2015). In Arabidopsis, PHF11 (plant homeodomain finger 11) plays roles in binding and suppressing RPA, thereby enhancing Exo1 resection (Gong et al. 2017). Furthermore, the resection coordination activity of MRE11 and CtIP/Ctp1 may inactivate KU70/80 and unload it from the broken ends. Meanwhile, a predefined resection length may deactivate the MRN complex and disassociate it from the ends (Langerak et al. 2011).

DSB Amendment by HDR

Once the HDR pathway is determined, in the presence of homologous DNA templates, HDR can occur through gene conversion or synthesis-dependent strand annealing (SDSA), single-stranded annealing (SSA) or crossover (CO, DSB repair (DSBR))/noncrossover (NCO) via double Holliday junction (dHj) formation (Holliday 1977). Only the former HDR subpathway can produce precise sequence products (Fig. 1). In plant somatic cells, SDSA was proven to be the major HDR mechanism to precisely repair damaged DNA (Szostak et al. 1983 Puchta et al. 1996 Voytas 2013). The differentiation of HDR subpathways has been well studied in yeasts and mammals but still remains a matter of investigation in higher plants. In the case of HDR, phosphorylation of H2AX histone protein by ATM or DNA-PKcs is important to open nucleosomes for strand annealing. As one H2AX is present for every 10 nucleosomes, efficient HDR requires relaxing up to thousands of base pairs (Lieber 2010). The resection of broken ends at a controllable length of 3′ ssDNA overhangs would favor RAD51-dependent SDSA repair. RPA binds to the resected ssDNAs to prevent the formation of a secondary loop for RAD51 loading. RAD51 loading is facilitated by BRCA2 through its BRC motif, which plays dual roles as an ssDNA-dsDNA junction binding protein as well as a RAD51 docking site provider (Seeliger et al. 2012 Heyer et al. 2010 Dray et al. 2006). The tight regulation of RAD51 loading and nucleofilament formation has been shown to involve a BRCA2-antagonistic protein called FIDGETIN-LIKE-1 (FIGL1) (Fernandes et al. 2018 Girard et al. 2015 Kumar et al. 2019). Extensive end resection with the involvement of Exonuclease 1 (Exo1) and/or Sgs1-Dna2 would lead to RPA disassociation facilitated by RAD52, which redirects to the error-prone SSA repair pathway (Heyer et al. 2010) (Fig. 1).

In Arabidopsis, INVOLVED IN DE NOVO2 (IDN2) was shown to help RAD51 loading by binding to RPA and unloading it from DSB ends (Liu et al. 2017). RAD51 binds to the resected ssDNA overhang, forming nucleoprotein filaments or presynaptic filaments. The filament structure invades the donor template sequence and then searches for and anneals to the complementary sequence this process is followed by displacement loop (D-loop) formation (Rajanikant et al. 2008). RAD54 binds to and is required for supporting RAD51 strand invasion and annealing and for the disassociation of RAD51 afterward (Klutstein et al. 2008 Osakabe et al. 2006). RAD54 formed DNA repair foci in living Arabidopsis cells depending on ATM-SOG1 signaling and coincided with the formation of phosphorylated H2AX (Hirakawa et al. 2017). Subsequently, the free 3′ OH end of the invaded ssDNA primes donor template-dependent DNA synthesis. This process determines the outcomes of HDR with several subpathways (DSBR, dHJ and SDSA) depending on the type of DNA synthesis and resolution of the final products (Fig. 1). In the later stage of homologous template-dependent synthesis in somatic cells, the D-loop may be processed and reannealed by the activity of RAD5A, REC4Q and MUS81 (Mannuss et al. 2010 Hartung et al. 2006). Only SDSA can generate precise repair products and is favored in mitotic cells (Heyer et al. 2010 Puchta 2005).

HGT in Monocots

Plant gene targeting or HGT was defined by the homology-directed repair (HDR) of an endogenous gene by exogenously introduced homologous DNAs (Paszkowski et al. 1988). Obviously, the initial experiment obtained a very low frequency of homologous recombination (

10 − 4 ), indicating difficulty but feasibility in engineering plant genomes by site-specific gene targeting. Early in the 1990s, a transgenic approach using a preintroduced yeast mitochondrial I-SceI endonuclease as a DSB inducer was adopted in attempts to investigate the mechanisms of DSB repair in plants, especially the HDR pathway in plant somatic cells (Puchta et al. 1993 Fauser et al. 2012 Szostak et al. 1983). It became clear that the HDR pathway employing homologous DNA templates to precisely repair DSB-damaged DNAs occurred mainly via the SDSA mechanism (Fig. 1) with an extremely low efficiency. Nonetheless, the induced DSBs could improve HGT efficiency up to two orders of magnitude (Szostak et al. 1983 Puchta et al. 1996), a large step in plant gene targeting research. Recently, the emerging CRISPR/Cas systems, which have proven to be powerful molecular scissors for in vivo generation of site-specific DSBs, have revolutionized the plant gene targeting approach and brought hope for practical applications in crop improvement.

However, despite the application of flexible approaches (i.e., particle bombardment, protoplast transfection and Agrobacterium-mediated transformation) for the delivery and execution of HGT tools, gene targeting in most major crops is still challenging. As mentioned in the previous sections, most of our knowledge about the principal mechanisms of plant HDR has been taken from yeast and animal research studies, and some of those results are inconsistent with observations in the plant kingdom. Therefore, the plant genome engineering community should continuously focus on research to understand plant-specific factors involved in DSB repair, especially via the HDR pathway, the only approach providing precise gene targeting products. Using this background knowledge, one can propose approaches for improving gene targeting frequency. Two of the most important factors affecting gene targeting efficiency in plant somatic cells are 1) DSB formation at the targeted sites and 2) the number of homologous DNA templates available for the sites of breakage (Puchta et al. 1993 Puchta 2005 Townsend et al. 2009 Endo et al. 2016 Baltes et al. 2014).

Because most of the early studies focused on gene targeting in model dicot plants such as Arabidopsis, tobacco and tomato (for reviews, see (Voytas 2013 Puchta 2005), monocot gene targeting represented a large gap in the early reports, indicating major challenges in monocot gene targeting. In this section, we aim to summarize recent knowledge regarding gene targeting in the monocot plants that represent most of the major food crops for human beings. In addition, we discuss challenges and suggest potential solutions for improving gene targeting frequency in monocots.

HGT without Targeted DSBs

In vivo plant gene targeting without assisted selection was extremely low (Puchta and Hohn 1991 Paszkowski et al. 1988). The first targeted knockout of an endogenous “waxy” allele via HGT was successfully generated in rice at a 0.94% frequency by Terada and coworkers (2002) with an innovative positive (hygromycin phosphotransferase II (HptII)-based)/negative (using diphtheria toxin A (DT-A) subunit) selection method (Table 1). The frequency of the gene-targeted waxy and xyl (b1,2-xylosyltransferase) knockout alleles was further improved by the transformation frequency (Ozawa et al. 2012). The weak point of this strategy is the obligatory use of an associated marker gene hence, the product is subject to GMO categorization. Therefore, Cre/loxP was applied to excise the marker from the gene-targeted allele (Terada et al., 2010 Dang et al. 2013). The approach was later successfully applied to functional genomic studies via tagging endogenous genes with visible marker(s) (Yamauchi et al. 2009 Moritoh et al. 2012 Ono et al. 2012 Tamaki et al. 2015). The positive/negative system using the DT-A subunit might have posed risks to dicots, because it has not been successfully applied in those plants. Therefore, an alternative positive/negative selection system was developed as an alternative, based on a caffeic acid O-methyltransferase (codA) D314A single-mutated version as the negative selection marker (Osakabe et al. 2014) or neomycin phosphotransferase II (NptII) (positive)/RNAi-based anti-NptII (negative) selection at much lower frequencies (Nishizawa-Yokoi et al. 2015b), which might be a result of less efficient G418 selection in rice. Nonetheless, the positive/negative selection strategy was shown to be unsuccessful in barley (Horvath et al. 2016), highlighting its extremely low efficiency in the absence of DSB and the high genome complexity of monocot gene targeting. In an herbicide-selection-based gene targeting experiment, Endo and coworkers successfully replaced the WT allele of rice ALS with the W548 L and S627I alleles and obtained homozygous T2 plants hypertolerant against an herbicide named bispyribac (BS). Under BS selection, gene targeting occurred at both loci at

3% (Endo et al. 2007). The frequency of targeting OsALS for BS tolerance was enhanced to 6% by using the abovementioned HptII/DT-A selection system, and the selection marker was subsequently excised with the piggyBac system, which can remove a marker gene without leaving a DNA scar (Nishizawa-Yokoi et al. 2015a).

Targeted DSB-Based HGT

DSBs induced at the gene targeting sites were shown to dramatically enhance efficiency by several orders of magnitude (Puchta et al. 1993). Since the introduced I-SceI meganuclease-mediated DSBs showed significant enhancement of gene targeting frequency, ectopic recombination was tested in maize and revealed remarkably higher efficiencies than the no-DSB strategies particle bombardment and Agrobacterium-mediated transformation (D'Halluin et al. 2008 Ayar et al. 2013). However, because the preintroduced homing nuclease targets a predefined sequence in the genome of the plant, gene targeting for a native gene/allele of interest in plant genomes still fell far short of expectations, and site-specific molecular scissors were in high demand. Bearing that in mind, researchers engineered ZFNs, zinc finger motifs for DNA binding fused to the type IIS endonuclease FokI, for efficiently and specifically generating DSBs in vivo (Kim et al. 1996) and obtained significant enhancement of gene targeting efficiency at native loci in Drosophila (

1.5%) (Bibikova et al. 2003) and human cells (

18%) (Urnov et al. 2005). Subsequently, ZFNs were applied to plant gene targeting and yielded an average of 17% HGT efficiency with a preintegrated GUS:NPTII reporter system in tobacco protoplasts (Wright et al. 2005). A similar strategy also obtained

10% HGT efficiency in restoring a preintegrated defective herbicide-tolerance gene (Cai et al. 2009). For targeting multiple allelic loci, also acting as an herbicide-tolerance selection marker, the efficiency was several-fold lower at

2% in tobacco (Townsend et al. 2009). In monocots, ZFN-based gene targeting was first shown to be efficient in maize via integrated insertion of an herbicide-tolerance gene as a selection marker into a native inositol-1,3,4,5,6-pentakisphosphate 2-kinase (IPK) gene (Shukla et al. 2009). Although ZFNs offered a great advantage over meganucleases in plant gene targeting, their design, validation and specificity optimization processes were extremely time consuming and laborious (Puchta and Hohn 2010).

TALENs, the second generation of sequence-specific nucleases, also used protein-based DNA binding domains for targeting sites of interest. Their highly specific and modular binding repeats offered an easier alternative for plant gene targeting. The first plant gene targeting events via HGT using TALENs were in tobacco calli regenerated from protoplasts at 3.5% efficiency without any selection marker (Zhang et al. 2013). Overall, 3.5% of calli showed HGT events without antibiotic selection however, it is not clear how many protoplasts were used for transfection. The TALEN approach was first applied in monocots to demonstrate the feasibility of gene targeting and reached 2–3% post bombardment of leaves with TALENs plus donors (Budhagatapalli et al. 2015). In rice, a similar range (1.4–6.3%) of gene targeting frequencies was obtained with the OsALS herbicide-tolerance allele (Li et al. 2016).

With the advent of CRISPR/Cas, which revolutionized molecular scissors for DSB formation, plant gene targeting is in theory applicable to any gene/crop of interest due to the simplicity, flexibility and versatility of the system (Jinek et al. 2012 Zetsche et al. 2015). CRISPR/Cas tools have been adapted for wide use in genome engineering studies in various kingdoms, including Plantae (Jinek et al. 2012 Hsu et al. 2014 Barrangou and Doudna 2016). The first attempt to modify a monocot genome via HGT using CRISPR/Cas9 was shown in 2013 by Shan and coworkers. In a transient experiment, OsPDS was modified by HGT at a 6.9% frequency in rice protoplasts using CRISPR/Cas9 for DSB formation and single-stranded oligos as donor templates (Shan et al. 2013). Gene targeting in maize was shown with an efficiency comparison between Agrobacterium-mediated delivery and particle bombardment and between a meganuclease and CRISPR/Cas9 at two loci, ALS and LIG1 (Svitashev et al. 2015). Several herbicide-tolerant lines were obtained from the bombardment approach only, indicating a very low targeting efficiency in maize and the requirement of a high dose of donor template and editing tools for enhancing it. The herbicide-tolerance ALS allele was also used in another CRISPR/Cas9-based gene targeting work using a short dsDNA donor delivered as linearized or plasmid forms by bombardment or Agrobacterium. With hygromycin and BS herbicide for double selection, the total frequency of HGT events reached 22.5–25%. In detail, most of the HGT lines (42/52) obtained from bombardment showed a range of diversity, with mixtures of perfect W548 L and imperfect S627I. In Agrobacterium-mediated delivery, most HGT lines (30/40) were perfect but heterozygous, and the HGT alleles were co-located at the loci with NHEJ alleles (Sun et al. 2016). The highly chimeric HGT patterns indicate prolonged activity of CRISPR/Cas9 unsynchronized states of the cells used in the experiments and/or predominance of organogenesis during shoot formation post editing. Therefore, to synchronize DSB formation and HDR, Endo et al. (2016) used calli stably expressing SpCas9 and sequential transformation of sgRNAs and repair templates for OsALS gene targeting. The HGT frequency was too low, and it was difficult to obtain the target plants. However, when the DNA ligase 4 (LIG4) gene, a key player in the NHEJ pathway, was knocked out before targeting, up to a 1% frequency of gene targeting among the total herbicide-tolerant calli was observed, indicating competition between the NHEJ and HDR pathways (Endo et al. 2016). In an attempt to modify the nitrate transporter gene NRT1.1B using CRISPR/Cas9-based tools, Li and coworkers obtained 6.72% precise replacement of 4 SNPs in the gene sequence without an additional allele-associated selection marker (Li et al. 2018a). The gene-targeted lines might contain DNA insertions in their genome due to the high frequency of DNA integration of the bombardment system, but this possibility was not examined.

An alternative to the Cas9 system is Cpf1-based molecular scissors. The latter cut dsDNAs using a T-rich PAM for binding initiation and usually form 5′ overhangs at their distal ends relative to the PAM (Zetsche et al. 2015). CRISPR/Cpf1 was also used for gene targeting in monocots and showed precise SDSA-based gene replacement at the OsALS loci at comparable frequencies (0.66–1.22%) (Li et al. 2018b) to those of Cas9 systems (Endo et al. 2016).

Replicon-Based HDR

Because of the highly efficient replication of geminivirus genomes and their single-stranded DNA nature, these genomes have been used as perfect DNA template cargo for gene targeting in plants. Geminiviral genomic DNAs have been reconstructed to overexpress foreign proteins in plants at up to 80-fold higher levels than those of conventional T-DNA systems (Needham et al. 1998 Mor et al. 2003 Zhang and Mason 2006) due to their highly autonomous replication inside host nuclei and the ability to reprogram cells (Gutierrez 1999 Hanley-Bowdoin et al. 2013). Furthermore, Rep/RepA has been reported to promote a cell environment that is permissive for HR to stimulate the replication of viral DNA (Baltes et al. 2014). Interestingly, it has been reported that somatic HR is promoted by geminiviral infection (Richter et al. 2014). The above characteristics of geminiviral replicons have been shown to make them perfect delivery tools for introducing large amounts of homologous donor templates to plant nuclei. Likewise, the movement and coat proteins of a bean yellow dwarf virus (BeYDV)–based replicon were removed and replaced with Cas9 or TALEN to improve gene targeting in plants (Baltes et al. 2014 Butler et al. 2016 Cermak et al. 2015 Dahan-Meir et al. 2018).

In monocots, wheat dwarf virus (WDV) was first engineered for CRISPR/Cas9-based genome editing and gene targeting in wheat (Gil-Humanes et al. 2017). More importantly, this work showed the feasibility of multiplexed gene targeting of multiple homeoalleles of the wheat genome at a 1% frequency. A similar approach using WDV was also applied in rice for targeted insertion of GFP-2A-NPTII to the C terminals of ACT1 and GST genes in a Cas9-overexpressing WT background. The WDV replicon-based tools showed significantly higher targeted knock-in efficiencies than conventional T-DNA tools (Wang et al. 2017).

Present Challenges

Despite higher success rates in gene targeting in plants, most of the abovementioned cases required marker-associated or selectable loci, while the selection and regeneration of HGT events from edited cells are still challenging (Butler et al. 2016 Gil-Humanes et al. 2017 Hummel et al. 2018). The most effective delivery method for HGT tools was reported to be particle bombardment, with relatively high frequencies of gene targeting (see Table 1) due to the high doses of introduced donor DNAs, but it also resulted in multiple DNA integration and/or regeneration difficulties. In addition, compared to other delivery methods, such as Agrobacterium-mediated transformation, particle bombardment requires special equipment and costly consumables that are not widely available in every research laboratory. Agrobacterium-mediated transformation is a very common and cost-effective method for plant gene targeting, but it showed too low frequencies with conventional T-DNA cargos (Table 1). There has been one solution for delivery of high copy numbers of donor DNAs, without facilitating multiple DNA integration, using autonomous DNA replicons (Baltes et al. 2014 Cermak et al. 2015), but this technique is still challenging in monocots if not used in combination with bombardment (Wang et al. 2017) or with a stable Cas9-overexpressing background and selectable marker (Gil-Humanes et al. 2017). The frequencies were dramatically reduced if multiple allelic loci and/or polyploid plants were targeted (Table 1). Furthermore, the effective application of replicon cargos in plant gene targeting has been shown to be limited by their size (Baltes et al. 2014 Suarez-Lopez and Gutierrez 1997 Gil-Humanes et al. 2017). Therefore, plant gene targeting, especially in cases of nonselectable alleles, is still a matter of improvement.

Potential Solutions and Perspectives on Monocot HGT

To improve plant gene targeting frequency, understanding HDR mechanisms and finding optimal conditions for HDR are the most important subjects in the field. The initial data on DSB-based gene targeting led to an important conclusion that in plant somatic cells, the majority of HGT-based products were formed via the SDSA pathway (Fig. 1) (Puchta 1998 Voytas 2013 Vu et al. 2017 D'Halluin et al. 2008). Because it is well known that DSB formation is one of the key factors in gene targeting and that viral replicons are used as efficient delivery systems for HDR donor templates, we will discuss and propose only other factors regarding monocot gene targeting here.

The Role of Homologous Donor Templates

The initial experiments for understanding plant homologous recombination were mostly conducted in a transient manner using newly introduced homologous DNAs/plasmids in plant protoplasts/cells. Baur et al. (1990) reported extrachromosomal homologous recombinations between two plasmids in tobacco mesophyll protoplasts. The most favorable donor plasmids were in linearized forms that obtained 15- to 88-fold higher recombination efficiency and were proportional to homologous zone size. The closer the break sites were to homologous zones, the higher the recombination frequencies were (Table 2) (Baur et al. 1990). Puchta and Hohn also confirmed that the homologous zone sizes (456 bp to 1200 bp) have a direct correlation with extrachromosomal recombination frequencies in Nicotiana plumbaginifolia protoplasts. The frequency was significantly reduced when the homologous zone size was 456 bp or lower (Puchta and Hohn 1991). Single-stranded DNA templates were shown to be efficient substrates for extrachromosomal recombination because they could directly facilitate the initial annealing step between the donor and targeted DNAs. Double-stranded circular DNAs were the least efficient templates for the recombination mode (Bilang et al. 1992 de Groot et al. 1992).

Positive-Negative Selection

Selection and/or regeneration of gene targeting transformants are critical to the success of the approach. The dual mode of selection strongly enhanced the possibility of obtaining gene targeting events in monocots (see Table 1), even without the involvement of the revolutionary CRISPR/Cas molecular scissors. The positive-negative selection system provides a large advantage in rice HGT and may help us improve crops by HGT (Terada et al. 2002). The hurdles in the removal of the associated positive selection markers have been solved by using a smart transposon-based excision system (Nishizawa-Yokoi et al. 2015a). It is exciting to combine the positive-negative selection system with the high DSB performance of CRISPR/Cas complexes for monocot gene targeting.

Overexpression of Genes Involved in the HDR Pathway

A good number of HDR-related protein homologs have been identified among prokaryotes and eukaryotes. Attempts have also been made to study and/or improve HDR in somatic cells by overexpressing the proteins in targeted organisms. We discuss these approaches in this section, thereby highlighting important points for the improvement of plant gene targeting frequency.

The Escherichia coli RecA protein (EcRecA) was shown to be involved in HR in this bacterium by facilitating ssDNA searching and annealing to its homologous DNA repair templates and subsequently exchanging and displacing the sequence (Radding 1981 Muniyappa et al. 1984 Chen et al. 2008). Overexpression of EcRecA in tobacco protoplasts enhanced the DNA repair efficiency 3-fold upon treatment with interstrand DNA crosslinking agent (mitomycin C) (Table 2). Intrachromosomal HR frequency was also shown to be 10 times higher in cells expressing the protein (Reiss et al. 1996). However, an SpCas9-EcRecA fusion was shown to enhance indel mutation via supporting the SSA repair mode (Fig. 1) and hence to suppress homology-directed gene conversion at 33% in mammalian cells (Lin et al. 2017). In contrast, Cai and coworkers showed a 1.7-fold increase in HGT frequency after cotransfection of the CRISPR/Cas9 complex and EcRecA into human embryonic kidney (HEK) 293FT cells (Cai et al. 2019). In E. coli, EcRecA acted in concert with the RuvC protein to resolve Holliday junctions in the late stage of DNA recombination (Iwasaki et al. 1991). By introducing RuvC into the nuclei of tobacco plants, Shalev and coworkers obtained strong enhancement of somatic crossover (12-fold), intrachromosomal recombination (11-fold), and extrachromosomal recombination (56-fold) (Shalev et al. 1999). This improvement may also be useful and applicable for DSB formation-based plant gene targeting approaches.

Activities of helicases have been shown in the initiation of homologous recombination. A transgenic approach using E. coli RecQ (EcRecQ) revealed positive effects on extrachromosomal recombination of a two-vector system cointroduced into rice leaves. The EcRecQ transient expression driven by a monocot-specific promoter induced a 4-fold increase in extrachromosomal gene targeting. The stimulation was much higher, at 20–40-fold in cases of stable EcRecQ expression (Li et al. 2004). This report confirmed the importance of helicase activities in HDR and suggested another potential approach for the enhancement of monocot HGT frequency.

As discussed earlier, RAD54 plays roles in concert with the activities of RAD51 during the HDR-mediated DSB amendment stage. Overexpression of yeast RAD54 in Arabidopsis was reported to increase gene targeting frequencies up to 27-fold, indicating the importance of strand invasion and/or chromatin remodeling in the HDR pathway (Shaked et al. 2005). The developmental stages of explants used for monocot gene targeting may differentially support the HDR pathway. The largest amount of recombination occurred in embryogenic cells, and this result was explained by the higher expression levels of OsRAD51 mRNA in the cells (Yang et al. 2010). Enhancement of the resection of the broken ends by overexpressing OsRecQl4 (BLM counterpart) and/or OsExo1 (Exo1 homolog) might positively support gene targeting in rice (Kwon et al. 2012).

Knockout of Genes Relating to the HDR Pathway

As discussed above, RAD50 plays a central role in the MRN/MRX complex for the resection of the broken ends of dsDNAs. Knockout mutations of RAD50 led to developmental lethality in mice (Roset et al. 2014) and suppression of gene targeting in moss (Kamisugi et al. 2012). Surprisingly, a homozygous rad50 KO A. thaliana showed hyperrecombination in somatic cells, as it supported 8- to 10-fold higher gene conversion frequencies of an inverted repeat substrate (Table 2) (Gherbi et al. 2001). This led to an important conclusion that MRN/MRX activities are required by NHEJ more than by HR. The data suggest a strategy that transiently suppresses plant RAD50 during a gene targeting experiment to achieve high frequencies.

Sequence divergence between homologous DNA templates and targeted loci has been shown to affect plant HGT frequency. The HGT frequencies were dramatically reduced by 4.1-, 9.6-, 11.7- or 20.3-fold when the levels of sequence divergence were increased by 0.5%, 2%, 4% or 9%, respectively. The sequence divergence might trigger a nucleotide mismatch repair (NMR) mechanism with the involvement of the NMR key protein AtMSH2 and hence disturb the HDR process (Li et al. 2006 Emmanuel et al. 2006). AtMLH1, a homolog of E. coli MutL that is involved in NMR, was shown to be required for homologous recombination and homeologous recombination. AtMHL1 mutation led to strong HDR reduction but a less severe reduction in homeologous recombination (Dion et al. 2007). The data indicate the potential for the regulation of MSH2 and/or MLH1 expression for the enhancement of HGT in monocots, especially when homologous DNA templates with obligate mismatches are used.

Chromosome accessibility is a key factor determining DSB formation and the subsequent repair of the broken DNAs. During replication or transcription, the chromatin is loosened, and the nucleosomes are opened for the assessment of related proteins involved in these processes. The Arabidopsis thaliana chromatin assembly factor 1 (CAF-1) complex involved in nucleosome assembly is formed by AtFAS1, AtFAS2 and AtMSI1 subunits. Endo and coworkers showed that knockout mutations of either AtFAS1 or AtFAS2 led to enhancement of somatic HR, potentially by 40-fold, thanks to the opening of nucleosomes for accessibility, cell cycle synchronization favoring HDR conditions, and high expression of HDR-related genes in the mutant backgrounds (Endo et al. 2006). The data suggest a potential enhancement of gene targeting via transient AtFAS1/2 knockdown by RNAi while introducing editing tools in somatic cells of monocots.

Another approach was tested in several studies that showed positive effects on HGT by suppressing important genes involved in the NHEJ pathway, such as KU70/80 or Lig4 (Nishizawa-Yokoi et al. 2012 Endo et al. 2016). This approach also showed a reduction in stable integration of T-DNA in the KU70/80 and Lig4 suppression conditions, suggesting a mechanism of T-DNA integration in the genome.

Favorable Tissue Culture Conditions for Gene Targeting

Polyamines accumulated in cells with induced DSBs and were subsequently shown to improve HGT by promoting RAD51-mediated DNA strand exchange. During in vitro assays, polyamines facilitated the capture of duplex DNA by the RAD51 presynaptic filament (Lee et al. 2019). Physical support of the substances may be a good approach for enhancing the activity of RAD51, a key protein in the SDSA subpathway for gene targeting in monocot somatic cells. Chemicals that suppress genes involved in the NHEJ pathway were used for testing HGT enhancement effects. Some chemicals inhibited DNA-PK (Robert et al. 2015) or KU70/80 or Lig4 (Table 2) (Chu et al. 2015 Maruyama et al. 2015), thereby enhancing HGT frequency in mammalian cell lines (Yu et al. 2015). It is still not clear whether we can achieve similar gene targeting enhancement in plants. Data obtained from our laboratory showed nearly no effects of SCR7 and/or RS-1 on tomato gene targeting using geminiviral replicons in combination with CRISPR/Cpf1 (unpublished data). Temperature is an important factor enhancing CRISPR/Cas9-based targeted mutagenesis in plants (LeBlanc et al. 2018) and CRISPR/Cpf1-based HDR in zebrafish and Xenopus by controlling genome accessibility (Moreno-Mateos et al. 2017). Recently, we re-engineered geminiviral replicon vectors in combination with CRISPR/Cpf1 and showed enhancement of HGT frequency at high temperatures and under lighting conditions (Vu et al. 2019).

Cell Cycle Synchronization

One of the reasons that HDR is limited to the S-G2 phases is the availability of sister chromatids to be used as donor templates. As a consequence, the majority of HDR genes might have evolved to be specifically expressed in these phases. The ideas are to artificially favor cellular conditions (S and G2 phases) in which HDR is more efficient and that limit NHEJ blocking of the targeted sites in other phases (M and G1), especially in the case of Cas9s, because they cut in the core sequences proximal to their PAMs. To that end, cell cycle synchronization at the S/G2 phase using chemical (hydroxyurea) or molecular approaches could be applied (Tsakraklides et al. 2015 Gutschner et al. 2016). Cas9 fused with the N-terminal (110a.a) end of human Gemini, a replication licensing factor that is a direct target of an M/G1-restricted E3 ubiquitin ligase for proteolysis, synchronized Cas9 expression in the S/G2 phase, thereby enhancing HGT up to 87% compared to only Cas9 (Table 2) (Gutschner et al. 2016).

In planta Gene Targeting

Gene targeting in maize may be performed during fertilization because it provides a permissive environment for sequence exchange by HGT (Djukanovic et al. 2006). In 2012, Fauser and coworkers demonstrated the feasibility of using the pre-integrated target, donor template and homing nuclease (I-SceI) in the planta gene targeting in Arabidopsis to correct the truncated GUS marker in the target with the remaining part located in the donor template. Crossing of the lines carrying homozygous target and donor alleles with a line expressing I-SceI obtained somatic GT events in the F1 generation that could be inherited in the F2 progeny at 6.8 × 10 − 3 frequency (Fauser et al. 2012). Targeted mutagenesis using CRISPR/Cas9 has been shown to be a highly valuable in planta approach for crop improvement (Kelliher et al. 2019). These approaches could also be applied for CRISPR/Cas-based monocot gene targeting, and it would avoid the laborious, time consuming and complex tissue culture process. In planta gene targeting could reduce the mutation rate compared to the tissue culture system, which is accompanied by many mutations. However, the targeting tools should be redesigned to match the conditions (pollen-specific and/or ovule-specific) so that they work within the short time period of pollination and fertilization.

HDR-Based Monocot Events and Regulatory Aspects

Genome-edited crops, including those created by CRISPR/Cas-based targeted mutagenesis and HGT approaches with or without the uses of DNA cargos, are referred to as products of “new breeding techniques (NBTs)” (Laaninen 2016 Lusser et al. 2011) or “new genetic modification techniques (nGMs)” (Eckerstorfer et al. 2019). In most of these genome-editing events, foreign genetic editing tools could be excluded from the organisms after finishing their roles, except that exotic DNA sequence(s) need to be introduced to specific site(s) in their genome(s). Likewise, most of the genome-edited transformants could not be distinguished among other mutated crops generated by conventional mutagens or natural mutations (Friedrichs et al. 2019 Grohmann et al. 2019), and hence, they should not be regulated. The regulatory legislation seems to be more complicated for HGT events because they have been regulated either as non-GMOs or GMOs by the US, EU, Japan, Australia, and others (for an extensive review, see Eckerstorfer et al. 2019). In this section, we summarize and discuss the regulatory aspects of HGT crops, including monocots, as the critical hurdle for the commercialization of HGT crops. We would also attempt to propose a regulatory principle that could be useful for countries during the legislation process.

Current Status in Regulatory Policies for Genome-Edited Crops

The US is the leading country in the commercialization of GM crops to date, with 75 Mha of planted biotech crops in 2018 (ISAAA 2019). In the same year, the US was also the leading country to release policies for the regulation of genome-edited crops. The USDA announced that “Under its biotechnology regulations, USDA does not currently regulate or have any plans to regulate plants that could otherwise have been developed through traditional breeding techniques as long as they are developed without the use of a plant pest as the donor or vector and they are not themselves plant pests” (USDA_Press 2018). This means that genome modifications such as deletions, base substitutions and plant DNA modifications, being similar to those potentially generated by conventional cross-breeding, are all deregulated by USDA policies (NatPlants/Editorial 2018). In Japan, the Ministry of Environment released its final policy on environmental safety on Feb. 8, 2019. According to the decision, creating food items using genome editing is not considered to produce GMOs, under the conditions that any DNA from the nucleases required to edit the target organism are not left within the genome and the resulting gene edits could have also occurred naturally. The Japanese Ministry of Health, Labor and Welfare announced a nearly identical assessment with regard to food safety on March 27, 2019 (USDA/JA9050 2019). Brazil, Argentina, Canada, Chile and Colombia have decided to regulate genome-edited crops at similar levels to the US (Ledford 2019). The Australian government adopted a middle level of regulation because SDN-1 products would not be regulated (Mallapaty 2019). By contrast, on July 25, 2018, the European Court of Justice decided that genome-edited crops would be subject to the same rules as transgenic plants or animals (ECJ 2018). Other governments, including those of the Republic of Korea, China, Russia and India, are still making their determinations of how to regulate this technology.

SDN Declaration

In fact, according to the released ruling policies of the governments except the EU, not all genome-edited transformants are considered non-GM. In principle, the genome-edited crops were initially divided under the classification of the so-called site-directed nucleases (SDN) by the European Food Safety Authority (EFSA) in 2012: “In SDN-1 applications, only the SDNs are introduced into plant cells (stably or transiently), generating site-specific mutations by nonhomologous end-joining (NHEJ). In SDN-2 applications, homologous repair DNA (donor DNA) is introduced together with the SDN complex to create specific nucleotide sequence changes by homologous recombination (HR) or homology-directed repair (HDR). The SDN-2 technique can introduce substantial changes to the nucleotide sequences of the target gene but more precise changes according to the bioengineer’s plan. SDN-2 techniques can provide unlimited SNP alleles that can boost innovative crop breeding. In the SDN-3 technique, a large stretch of donor DNA (up to several kilobases) is introduced together with the SDN complex to target DNA insertion into a predefined genomic locus. The predefined locus may or may not have extensive similarity to the DNA to be inserted. The insertion can take place either by HR or by NHEJ. In the case of insertion by means of NHEJ, the technique is denominated the SDN-3–NHEJ technique” (EFSA 2012). This classification is now generally accepted as the basal information for genome-edited crop regulation. On August 20, 2018, Japan’s Ministry of Environment (MOE) released a draft of its regulatory policies, adding some detailed requirements for SDNs to be excluded from the Cartagena Protocol regulation (USDA/JA8064 2018). The levels of regulation are decided based on the presence/absence of foreign genetic carriers, the levels of modification and the natural existence of the modification in genome-edited organisms. From another point of view, they are assessed on a case-by-case basis (see Table 3). From the released regulations, it is now clear that HGT will be regulated as either non-GMO (some cases of SDN-2) or GMO (some cases of SDN-2 and SDN-3).

The classification and regulatory considerations have created a major challenge for plant gene targeting approaches to be commercialized, even though their efficacy would be enhanced at a practical level. Gene targeting for modifying SNPs is deregulated by “relaxed” governments such as the US but not Australia. HGT products subject to the SDN-3 category, containing inserted sequence(s) that could not potentially form in nature, will all be regulated as transgenic products (Table 3).

A Regulatory Proposal for NBPT Products

Many governments seem to be trying to create sufficient oversight to protect the public interest and at the same time not create new obstacles to technical innovation. Genome-editing-based precision breeding is an innovative technology, but the technologies will evolve continuously. In particular, HDR-based precision breeding technologies are the most cutting-edge technology among genome-editing techniques because they can produce both precise SDN-1 and SDN-2/SDN-3 products. HDR-based precision breeding products are generally classified as SDN-2 or SDN-3, as they use a DNA donor template during the gene editing process and are thus regulated as GMO in Australia and Japan, with potential exceptions. Mechanical classification of HDR-based genome-edited products in the GMO category might pose the most unreasonable obstacle to this plant breeding innovation. In fact, HDR-based precision breeding can fulfill the long-awaited dream of breeders by precisely introducing beneficial gene alleles from crossable relatives without other trait compromises, such as linkage drag (mixing of targeted beneficial traits and unintended undesirable traits by linkage effects). What might be the solution? We must now again remind ourselves of the purpose of the regulation of biotechnology products. Regulations exist to prevent new products from harming human health or the environment. Many various technologies can be used to produce the same or similar, effectively indistinguishable, products to traditional breeding products therefore, the consistent risk-based regulatory approach is to treat similar products identically. In this view, it is worth referring to the Canadian regulatory policy, which regulates only plants with novel traits (PNTs), irrespective of the technologies used (Ellens et al. 2019). According to Canadian regulation, some SDN-1 products or even chemically mutagenized products can be regulated. However, this regulatory policy provides more open opportunities to use various innovative technologies, including genome editing or GMO. In the end, the fruitfulness of NPBT crops will mainly depend on the level of regulation on NPBT products.


Background Information

Targeted protein degradation has resulted in an explosion of new avenues of research, from therapeutic drug discovery and clinical trials to the expansion of E3 ligase studies and phenotypic studies (Chamberlain & Hamann, 2019 Ciulli & Farnaby, 2019 Crews, 2018 Cromm & Crews, 2017 Deshaies, 2015 Schapira et al., 2019 ). Protein loss phenotypic studies using degradation compounds allow the highly precise and temporal loss of a target protein to be tailored a specific desired amount (Buckley et al., 2015 Nabet et al., 2018 Nishimura, Fukagawa, Takisawa, Kakimoto, & Kanemaki, 2009 Sathyan et al., 2019 Tovell et al., 2019 ). This approach is technically different from genetic CRISPR knock-out (Pickar-Oliver & Gersbach, 2019 ) or small interfering RNA (siRNA) approaches (Carthew & Sontheimer, 2009 Jackson & Linsley, 2010 ), which prevent the production of the protein and cannot be used to study essential proteins. To broadly enable targeted degradation studies, either for phenotype or to understand whether a protein could be degraded via the UPP, HaloPROTAC3 (Buckley et al., 2015 ), a small molecule degrader that irreversibly binds to and degrades HaloTag along with its fusion partners in live cells, was developed (Fig. 1). HaloPROTAC3 elicits degradation through formation of a ternary complex with a target HaloTag fusion protein and VHL, an E3 ligase component (Fig. 1). This results in ubiquitination of HaloTag and HaloTag target fusions and subsequent degradation by the proteasome (Fig. 1).

Several targeted protein degradation methods that utilize fusion tags or short degron epitopes have been developed (Buckley et al., 2015 Nabet et al., 2018 Nishimura et al., 2009 Sathyan et al., 2019 Tovell et al., 2019 ). These systems overcome the challenges involved in the design and development of successful target-specific degradation compounds (Crews, 2018 ) and provide broad and robust applicability for numerous targets. The chosen target for fusion-tag degradation studies must be accessible for recruitment into the UPP for degradation and expressed as a fusion tag protein (Buckley et al., 2015 Nabet et al., 2018 Nishimura et al., 2009 Sathyan et al., 2019 Tovell et al., 2019 ). In addition to HaloTag and HaloPROTAC3, discussed above, other systems include the auxin-inducible degron (AID Nishimura et al., 2009 Sathyan et al., 2019 ) and the FKBP F36V dTAG (Nabet et al., 2018 ) systems. The AID system utilizes a degron tag appended to the target protein yet requires exogenous expression of a non-native plant F-box protein as well as addition of auxin to induce degradation (Nishimura et al., 2009 ). Although this has been shown to successfully degrade several proteins, it has proved to be challenging for in vivo models. The dTAG system (Nabet et al., 2018 ) conceptually is more similar to HaloPROTAC3. Target proteins are expressed as FKBP F36V fusion proteins and recruited into a ternary complex with dTAG PROTAC and CRBN. A challenge with the dTAG system is the limited functionality of the FKBP F36V protein, which cannot be labeled fluorescently to enrich for CRISPR insertion or be used to further understand protein function. Both dTAG and HaloPROTAC3 work well for in vivo studies (BasuRay, Wang, Smagris, Cohen, & Hobbs, 2019 Nabet et al., 2018 ). In mice, target HaloTag fusions have been introduced genetically or via xenographs into the organism and then efficiently degraded through HaloPROTAC3 injection (BasuRay et al., 2019 ). The HaloTag fusions can additionally be fluorescently labeled to study protein localization as well as isolated using a HaloTag resin to study protein interactions (Urh & Rosenberg, 2012 ).

Critical Parameters and Troubleshooting

Basic Protocol 1 : CRISPR/Cas9 insertion of HaloTag or HiBiT-HaloTag

The use of CRISPR/Cas9 application for insertion of sequences >500 bp requires use of dsDNA donor vectors and longer homology arms (300-500 bp each) as compared to smaller insertions, for which single-stranded oligodeoxynucleotide (ssODN) synthesis is possible and homology arms (30-50bp each) are shorter. The use of dsDNA donor vectors in general results in low insertion efficiency, and expected pool percentage of edited cells with this approach could range between 0.1% and 15%. For insertion of HaloTag or HiBiT-HaloTag, the use of donor vector without a promoter is important to promote specific, on-target insertion and minimize random integration of vector, which could then result in expression of the tag alone. To identify CRISPR HaloTag target-edited cells from random integration, it is important to assess that the proper-sized fusion is made by amplifying the genomic region by PCR or visualizing protein size on a protein gel with the HaloTag TMR ligand (Los et al., 2008 ). Sequencing is also necessary to assess zygosity and ensure the proper genomic insertion. The fluorescent HaloTag ligands used for FACS enrichment can aid these protein functional characterizations as well by assessing proper localization in the cell by confocal microscopy (Los et al., 2008 ).

To further optimize the success of a HaloTag CRISPR insertion, it is advisable to test multiple guide RNAs and to choose a cell line that has high electroporation or nucleofection efficiency, if possible. If inserting the HiBiT-HaloTag sequence, the resultant pools from varying guides or electroporation settings can readily be screened for luminescence to see which conditions are most favorable for further optimization. If there is no preference for choice in termini, oftentimes insertion of tags at the extreme C terminus is easier in design, though N-terminal insertion can also be done. With N-terminal insertions it is important to place the tag directly following the starting ATG codon of the target protein, whereas with C-terminal insertions, the tag should be placed immediately preceding the endogenous stop codon. Even with insertion efficiency as low as 1%, the ability to fluorescently label HaloTag positive cells with JF646 HaloTag ligand coupled with FACS enables a path to enriched pool populations or single-cell clones that would otherwise be very challenging to achieve with blind sorting or limiting dilution approaches.

Ideally, for HaloPROTAC3 phenotypic studies, HaloPROTAC3 should degrade all copies of the target protein in the cell. For this reason, homozygous allelic HaloTag insertion is desired. However, due to the low efficiency, homozygous allelic insertions are rare as compared to heterozygous insertions, and many clones need to be screened to identify a homozygous insertion. Droplet digital PCR can be applied after FACS to identify heterozygous versus homozygous insertions before Sanger sequencing for final confirmation. Alternatively, heterozygous clones, particularly those with the insertion targeted to the N terminus of the protein, often contain small insertions or deletions (INDELs) on untagged alleles, resulting in a knockout of the untagged protein copies. As a result, the entire target protein pool in these clones is expressed as a fusion to HaloTag therefore, these clones are sufficient for phenotypic experiments with HaloPROTAC3.

Basic Protocol 2 : HaloPROTAC3 degradation of endogenous HaloTag fusions

HaloPROTAC3 will degrade HaloTag target fusion proteins that are recruited to the VHL E3 ligase component, incorporated into active E2/E3 ligase complexes for ubiquitination, and then trafficked to the proteasome. VHL is expressed throughout the cytoplasm and nucleus, as well as in numerous cell types (Buckley et al., 2012 ). When performing studies with HaloPROTAC3, it is important to be certain the cell type used expresses VHL (Buckley et al., 2012 ), and the HaloTag on the fusion protein is presented to the cytoplasm or nucleus. HaloTag itself contains numerous lysine residues (Encell et al., 2012 ) that can be ubiquitinated therefore, it is not necessary that the target protein contain lysines available for ubiquitination. Single-pass transmembrane (TM) proteins are degraded with PROTAC compounds (Burslem et al., 2018 Huang et al., 2018 ), and recently the first example of a multipass mitochondrial TM protein was shown as well (Bensimon et al., 2020 ). It is possible that certain HaloTag target fusion proteins are in higher-order complexes, and these structures may prevent engagement with VHL or incorporation into the productive ubiquitination complexes required to drive degradation of the HaloTag fusion. In these cases, changing the terminus where HaloTag is located may help, but there may be some situations where structural incompatibility may result in ineffective target degradation.

As endogenous target proteins have highly variable expression levels that will not be significantly altered by HaloTag insertion, varying concentrations of HaloPROTAC3 may be necessary for successful and maximal degradation. Because HaloPROTAC3 binds irreversibly to HaloTag (Buckley et al., 2015 ), it is not likely to act catalytically therefore, the HaloPROTAC3 concentration needed is directly proportional to the amount of target protein. The recommended initial concentration to test is 300 nM, but this could vary from low nanomolar to low micromolar depending upon the target. Dose-response series and different time points or kinetic degradation runs can be performed to determine the optimal dose of HaloPROTAC3 for the HaloTag fusion. In vivo studies would require multiple routes, doses, and time to be tested before an optimal HaloPRTOAC3 concentration could be chosen for downstream applications. As HaloPROTAC3 is nontoxic at these concentrations, cell viability is not impacted. The other parameter for optimization of HaloPROTAC3 degradation is time. Among the endogenous HaloTag fusions tested, all showed rapid and sustained degradation. Depending on the goals of any particular degradation study, the desired amount of degradation can be regulated by both the concentration and the duration of treatment, and these parameters can be well defined by monitoring the kinetics of degradation via luminescence with HiBiT. It is also important to note that if the HaloTag CRISPR insertion is heterozygous, treatment with the HaloPROTAC3 compound will degrade only the tagged endogenous protein. If full target knockout is required for phenotypic studies, homozygous CRISPR insertion or a heterozygous insertion with an INDEL in the untagged copy will be necessary.

With any PROTAC degradation study, it is important to be certain that observed target protein loss is due to the specific PROTAC mechanism. For endogenous HaloTag proteins, this can be achieved using ent-HaloPROTAC3 (Buckley et al., 2015 ). ent-HaloPROTAC3 has significantly reduced affinity for VHL engagement (Buckley et al., 2015 ), but this low affinity can be overcome with very high concentrations therefore, it is not recommended for use as a negative control beyond 1 µM. It is also possible that degradation of the target protein with HaloPROTAC3 may induce cell death, as may be the case for essential endogenous HaloTag target fusions. If so, orthogonal cell viability assays, such as CellTiter Glo (Promega, cat. no. G7570) or CellTox Green (Promega, cat. no. G8741), will be important to deconvolute protein loss from cell death. In addition, as these are live-cell assays, it is recommended that the DMSO concentration after addition of HaloPROTAC3 be maintained below 0.5% by volume after addition to the cells.

For the alternate procedures utilizing HiBiT-HaloTag fusion proteins, protein levels are easily measured with luminescence (Daniels et al., 2019 Riching et al., 2018 Schwinn et al., 2018 ). If low or no luminescence results with HiBiT lytic assays, it is critical to confirm by sequencing and blotting techniques that the cells carry the HiBiT-HaloTag CRISPR insertion with 100% sequence conformity and express the full protein fusion. If low or no luminescence is measured in the live-cell assay, be certain that the cells are expressing the LgBiT protein. Performing a HiBiT lytic endpoint assay (Nano-Glo HiBiT Lytic Detection System, Promega, cat. no. N3030) will determine whether LgBiT is present. Purified LgBiT protein could be added to ensure the HiBiT and luminescence substrate are functional. This can be separately confirmed with parental cells expressing the LgBiT vectors and then addition of a purified HiBiT protein as a control.

Understanding Results

In the schematic shown in Figure 1, the first step towards the degradation of endogenously tagged HaloTag fusions is the introduction of HaloTag or HiBiT-HaloTag into the target genomic locus through CRISPR/Cas9. In Basic Protocol 1, it is advisable to test several guide RNAs during the initial electroporation step to increase the chances of successful insertion at a chosen target terminus. Shown in Figure 2A are signal-to-background (S/B) ratio differences due to varying insertion efficiencies of multiple crRNAs chosen for tagging of the N terminus of elongin BC and Polycomb repressive complex 2–associated protein (EPOP) with HiBiT-HaloTag in HEK293 cells with the same dsDNA vector. The crRNAs have distinct sequences, which will guide Cas9 to cut at various locations around the desired insertion site on the genome. A HiBiT lytic luminescence assay was performed on the resultant CRISPR pools and the un-edited parental cell line. The S/B ratio was determined and the results illustrate how the choice of crRNA can significantly influence the efficiency of insertion at a chosen target site (Fig. 2A). CRISPR pools can then be further enriched for HaloTag-positive cells using live-cell fluorescence labeling of HaloTag with the JF646 ligand followed by FACS. Shown in Figure 2B is an example of an expected FACS histogram overlay of unedited parental HEK293 cells and CRISPR pools of HEK293 β-catenin-HaloTag labeled with JF646. As expected, due to the low efficiency of CRISPR insertion of HaloTag within a CRISPR pool (<5% edited cells in the total cell population), many of the cells in the HEK293 β-catenin-HaloTag CRISPR pools are unedited, overlapping with the parental HEK293 cell line (Fig. 2B). However, the HaloTag-positive cells, though only a small fraction of the total population, could easily be identified and separated from the unedited population, allowing the enrichment of those with the HaloTag insertion (Fig. 2B). This approach can be used to generate enriched mini-pools of HaloTag-positive edited cells or single-cell clones even in cases in which only 1% of the cell population contains the HaloTag insert.

As an example of HaloPROTAC3 degradation and phenotype studies of a target with no available degraders or specific PROTACs, a homozygous CRISPR clone of β-catenin-HaloTag-HiBiT was generated in HEK293 cells stably expressing LgBiT. In Figure 3A, we found that a HaloPROTAC3 concentration of 333 nM degraded ∼60% of the β-catenin by 3 hr and ∼80% by 24 hr using the optional HiBiT lytic detection protocol outline in Basic Protocol 2 (Fig. 3A). Degradation was not observed with ent-HaloPROTAC3 at either 3 or 24 hr (Fig. 3B), demonstrating that the protein loss is by the PROTAC-mediated mechanism. As HiBiT can also be used for imaging upon live-cell complementation with the expressed LgBiT protein, luminescence imaging was performed before and after 1 µM HaloPROTAC3 treatment for 2 hr (Fig. 3C). These results showed that the endogenous β-catenin-HaloTag-HiBiT is properly localized at the plasma membrane, cytoplasm, and nucleus (Clevers & Nusse, 2012 Nusse, 2012 ) and degradation appears to occur of all populations of β-catenin-HaloTag-HiBiT (Fig. 3C). Because the HaloTag-HiBiT tag is added to β-catenin upon synthesis, all populations of the protein will be labeled and therefore degraded by HaloPROTAC3, which binds all available HaloTag regardless of location. To study β-catenin phenotypic response to Wnt signaling and activation of the canonical Wnt pathway (Clevers & Nusse, 2012 Nusse, 2012 ), an orthogonal firefly Tcf reporter was transfected into the CRISPR β-catenin-HaloTag-HiBiT HEK293 cells (Figs. 3D and E). Upon Wnt3a treatment, β-catenin accumulates and enters the nucleus, binds to and initiates transcription of TCF/Lef genes (Clevers & Nusse, 2012 Nusse, 2012 ), including the firefly TCF luminescent reporter (Fig. 3D and E). Treatment of Wnt3a in the presence of a constant 1 µM concentration of HaloPROTAC3 results in a muted response of Tcf reporter activation (Fig. 3D) and loss of β-catenin even in the presence of Wnt3a stimulation (Fig. 3E).

To demonstrate the ability to degrade endogenous HaloTag proteins at a variety of cellular locations, as well as to understand the quantitative parameters of HaloPROTAC3 degradation, the HiBiT kinetic studies outlined in the optional live-cell luminescence degradation detection of HiBiT-HaloTag CRISPR insertion protocol steps were performed (Fig. 4). For these studies, three different HaloTag-HiBiT CRISPR clones were generated to different targets at different locations: nuclear (Fig. 4A and B), mitochondrial membrane (Fig. 4C and D), and cytoplasmic (Fig. 4E and F). The nuclear and cytoplasmic fusions were created in HEK293 cells that stably express LgBiT, and the mitochondrial membrane fusion was transfected with the LgBiT vector all cells were then treated with an 8-point dilution series of HaloPROTAC3 or the negative control ent-HaloPROTAC3 (Fig. 4A-F) to produce a dose-response curve. Luminescence was continuously monitored for 24 hr, and fractional RLU, normalized to the DMSO control, was calculated to generate full degradation profiles (Fig. 4A-F). All three target proteins from the different cellular compartments showed degradation by HaloPROTAC3 (Fig. 4A, C, and E), including the single-pass mitochondrial membrane protein (Fig. 4C). This is now the second example of a transmembrane mitochondrial protein showing degradation (Bensimon et al., 2020 ), suggesting that this class of membrane proteins are amenable to degradation via PROTACs. The lack of degradation with the ent-HaloPROTAC3 shows that protein loss of each of these targets is specific to HaloPROTAC3. Each of the HaloPROTAC3 degradation profiles was then used to calculate the degradation rate (Fig. 4G), as well as the degradation maximum (Dmax) and Dmax50, the concentration of HaloPROTAC3 that gave half the degradation maximum (Fig. 4H), for all three endogenous HiBiT-HaloTag targets. The degradation rate is calculated by fitting a single-component exponential decay model to each curve until the data reaches a plateau (Riching et al., 2018 ). The similarity of the rates and Dmax50 values for each of these targets indicates that it is primarily HaloTag which is driving the degradation, with minimal influence by the different targets. This is desirable for a fusion tag PROTAC system as it needs to be broadly applicable to numerous targets. The kinetic analysis also shows how the optimal dose of HaloPROTAC3 and time of treatment to achieve degradation can be clearly understood from the profiles, saving significant time and yielding more detailed information as compared to western blot analysis.

To quantitatively compare HaloPROTAC3 degradation with the dTAG PROTAC system, CRISPR clones of the EPOP target protein were generated by insertion of either HaloTag-HiBiT or FKBP12 F36V -HiBiT at the C terminus in HEK293 cells (Fig. 5). These cells were transfected with the LgBiT vector to enable HiBiT kinetic degradation detection, and dose-response curves for HaloPROTAC3 (Fig. 5A) or dTAG-13 (Fig. 5B) were obtained. The resultant degradation profiles showed that both HaloPROTAC3, which recruits VHL (Buckley et al., 2015 ), and dTAG-13, which recruits CRBN (Nabet et al., 2018 ), showed rapid and robust degradation of the endogenous EPOP target. The dTAG-13 showed a hook effect (whereby the rate slows and degradation decreases at higher concentrations) at concentrations >100 nM (Fig. 5B), which was not observed with HaloPROTAC3 (Fig. 5A). A hook effect can occur when unfavorable binary complexes can form due to the high amount of PROTAC present or affinity differences between the E3 ligase and target binding ligands (Ciulli & Farnaby, 2019 ). This results in a slowing of the degradation rate, as observed with dTag-13 (Fig. 5B). The HaloPROTAC3 degraded the endogenous target at a faster initial rate, and both showed nearly identical Dmax at the optimal concentrations for each PROTAC (Fig. 5C). Together, we conclude these approaches are highly similar in ability to degrade their targets yet differ significantly in ease of CRISPR editing, as the FKBP12 F36V has low efficiency of insertion and edited cells cannot easily be identified or fluorescently enriched from the CRISPR pool.

Time Considerations

Basic Protocol 1: Depending upon the target, the time required for computational design of guide RNAs and donor DNAs, including homology arms, is ∼1-2 hr. Several commercial programs exist to aid with this development. Including preparation of cells, the initial CRISPR electroporation requires 2-3 hr of laboratory time. The cell recovery from the electroporation varies greatly depending on the cell type and its growth rate, but this commonly requires 2-3 weeks of passaging and incubation. Sorting of the cells via FACS requires 3-4 hr, dependent on the instrumentation (how fast the instrument can sort out positive cells) as well as the concentration of the cells. Cell recovery post-FACS again is highly dependent upon cell type and growth rates therefore, this can vary between 2 and 8 weeks. This length of time will be shorter as higher numbers of cells are sorted into each well (mini-pools) as compared to single cell clones.

Basic Protocol 2: HaloPROTAC3 degradation is rapid and robust, and the time required depends on the target, as some are degraded to near completion within 5-6 hr, and other require 24-30 hr of treatment. Day 1 involves plating the cells for the assay. On day 2, the cells are treated, which can take up to 1 hr. They are then incubated for the time needed for degradation (typically 6-30 hr) and then protein levels are assessed. If performing a HiBiT lytic assay to determine endogenous protein levels, this takes ∼1 hr. If using antibodies and western blot analysis, typically this requires 1-2 days and a target-specific antibody.

Optional HiBiT Lytic and kinetic protocols: The HiBiT lytic assay to determine insertion efficiency within CRISPR pools or measure degradation of HiBiT-HaloTag target proteins takes 30-60 min. The kinetic degradation assay takes 3 days to perform including plating, Endurazine equilibration, HaloPROTAC3 degradation, and data analysis.


E.A.C., S.D.M., R.L.J., A.N.N., N.L., K.M.R., M.U., and D.L.D. are supported and funded by Promega Corporation. C.R.W. is supported and funded by Boehringer Ingelheim Pharmaceuticals. Promega Corporation is the commercial owner by assignment of patents or licenses of the HiBiT, HaloTag, and HaloPROTAC3 technologies.

Author Contributions

Elizabeth A. Caine: Data curation formal analysis investigation writing-original draft writing-review & editing. Sarah D. Mahan: Data curation methodology. Rebecca L. Johnson: Methodology. Amanda N. Nieman: Methodology. Ngan Lam: Methodology supervision. Curtis R. Warren: Conceptualization supervision. Kristin M. Riching: Conceptualization formal analysis investigation. Marjeta Urh: Conceptualization supervision writing-original draft writing-review & editing. Danette L. Daniels: Conceptualization data curation formal analysis investigation project administration.