The laboratory rat has been widely used as an animal model in biomedical science for more than 150 years. Applying zinc-finger nucleases or transcription activator-like effector nucleases to rat embryos via microinjection is an efficient genome editing tool for generating targeted knockout rats. Recently, clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated endonucleases have been used as an effective tool for precise and multiplex genome editing in mice and rats. In this review, the advantages and disadvantages of these site-specific nuclease technologies for genetic analysis and manipulation in rats are discussed.
Genetically modified animals that have been altered using gene targeting technologies are used as experimental models to perform functional analyses or various tests in biomedical research. In particular, knockout (KO) animals can help in understanding how a specific gene functions in vivo. The gene targeting technologies used to produce KO mice using embryonic stem (ES) cells were developed prior to 1990 (Mansour et al. 1988; Capecchi 1989a). Since then, KO mice have become major tools for functional gene analysis. Causative genes for specific human diseases have also been disrupted in mice to mimic human genetic disorders (Capecchi 1989b; Smithies 1993). In the post-genome era, the International Knockout Mouse Consortium (IKMC), which aims to comprehensively disrupt all protein-coding genes in the mouse genome using gene targeting technologies, is now progressing (Skarnes et al. 2004; Nord et al. 2006; Ayadi et al. 2012). Furthermore, knock-in (KI) mice, in which genes are added or modified, or conditional knockout mice with spatial or temporal control of genetic inactivation, are widely used. Gene targeting technologies have become critical tools for understanding gene functions including the genetic basis of human diseases.
Until recently it was difficult to produce mammalian KO animals other than mice using gene targeting technologies, as germline-competent ES cells were available only for mice. However, this situation changed with the availability of newly developed gene targeting technologies, called engineered nucleases or “gene scissors”. These engineered nucleases, such as zinc finger nucleases (ZFNs) or transcription activator-like effector nucleases (TALENs), are very effective, as shown when treating embryos via microinjection to generate targeted KO mice and rats (Geurts et al. 2009; Carbery et al. 2010; Mashimo et al. 2010, 2013; Meyer et al. 2010; Urnov et al. 2010; Cui et al. 2011; Tesson et al. 2011; Sung et al. 2013). In addition to rodents, sea urchins (Ochiai et al. 2010), Drosophila (Beumer et al. 2006, 2008), crickets (Watanabe et al. 2012), killifish (Ansai et al. 2012, 2013), and zebrafish (Doyon et al. 2008; Meng et al. 2008; Huang et al. 2011; Sander et al. 2011), as well as larger animals such as rabbits (Flisikowska et al. 2011; Song et al. 2013), and pigs (Watanabe et al. 2010; Hauschild et al. 2011; Carlson et al. 2012), have been successfully modified using these enzymes. In this review, the advantages and disadvantages of these site-specific nuclease technologies are discussed in relation to genetic analysis and manipulation in animals, especially in rats.
Zinc finger nucleases are chimeric proteins that consist of a specific DNA-binding domain that is made of tandem zinc finger-binding motifs fused to a non-specific cleavage domain of the restriction endonuclease FokI (Bibikova et al. 2001; Porteus & Carroll 2005; Wu et al. 2007) (Fig. 1). As one zinc finger unit binds with 3-bp of DNA, 9–18 bp sequences can be specifically recognized by combining 3–6 different zinc finger units. By designing two zinc finger motifs on either side of 5–6 bp spacer sequences at a target region, the FokI nuclease combined with the zinc finger can introduce a double-strand break (DSB) within the 5–6 bp spacer sequences. Although the DSB is usually repaired via non-homologous end joining (NHEJ), an arbitrary deletion or deletion of base pairs often occurs during the repair process. Consequently, repair by NHEJ is mutagenic and mostly results in a loss-of-function mutation. Moreover, if DNA fragments homologous to the targeted sequences are co-injected with the nucleases, homologous recombination (HR) can occur, enabling insertion of a transgene or replacement of the homologous sequences at the targeted region, which results in KI mutations. Therefore, artificially designed ZFNs can be used to generate KO or KI alleles at the targeted sequences via NHEJ or HR repair, respectively.
A summary of how to generate targeted KO rats using ZFNs is given in Figure. 1. Briefly, two ZFNs are designed across the spacer domain to recognize the targeted DNA sequences. Messenger RNAs are then transcribed in vitro from the two ZFN plasmids and injected into the male pronuclei of rat zygotes. Pronuclear stage embryos are collected from female rats that were superovulated by equine chorionic gonadotropin and human chorionic gonadotropin injection. The ZFN-injected embryos that differentiate into two cells are then transferred to the oviduct of pseudopregnant females. This method is based on a similar technique used to produce conventional transgenic animals (Palmiter et al. 1982, 1983; Mullins et al. 1990), except that mRNA is used. The procedure for the micromanipulation of embryos is the same for all nucleases, including the below-mentioned TALEN and clustered regularly interspaced short palindromic repeats (CRISPR) enzymes.
The ZFN technology was first reported in the 1990s (Kim et al. 1996; Chandrasegaran & Smith 1999). From the 2000s, ZFNs have been developed for various mammalian cells (Bibikova et al. 2001, 2003; Porteus & Baltimore 2003; Urnov et al. 2005; Hockemeyer et al. 2009), with nematode and zebrafish ZFNs developed in 2006 (Morton et al. 2006) and 2008 (Doyon et al. 2008; Meng et al. 2008) respectively. The first genetic modification (KO) in rats was reported in 2009 (Geurts et al. 2009). In model organisms where ES cells could not be used for gene modifications, this technology has been widely applied for generating genetically modified animals, especially in laboratory animals other than mice. Using this ZFN technology, we developed an interleukin-2 receptor gamma chain (Il2rg) KO rat (X-SCID) to investigate human X-linked severe combined immunodeficiency (X-SCID) (Mashimo et al. 2010). SCID rats that are deficient in the Prkdc gene and FSG (F344-scid Il2rg) rats that are simultaneously deficient in both the Prkdc and Il2rg genes have also been generated using the ZFN technology (Mashimo et al. 2012). In contrast to the “leaky” phenotype of the SCID mouse, where immunoglobulins such as IgG, are detected in the blood, SCID rats did not show such a leaky phenotype (Mashimo et al. 2012). These SCID rats can be used as hosts for xenotransplantation of human stem cells and tissues.
Transcription activator-like effector nucleases
Zinc finger nucleases provide a straightforward strategy for targeted gene disruption in zygotes, resulting in rapid and cost-effective knockouts compared with conventional technology using ES cells. However, there are hurdles in terms of cost and protocols, making it difficult to establish ZFNs as a routine laboratory process. Recently, an artificial nuclease technology similar to ZFN, called TALENs, has been reported (Bogdanove & Voytas 2011; Wood et al. 2011; Mussolino & Cathomen 2012; Joung & Sander 2013). Natural TAL effectors are potent virulence proteins from plant-pathogenic Xanthomonas bacteria that are injected into eukaryotic host cells where they function as transcription factors (Bogdanove & Voytas 2011). As fusions of TAL effectors to the FokI nuclease, TALENs can bind and cleave DNA in pairs (Fig. 1). Although the sequences recognized by ZF domains are limited in ZFNs, TAL effectors can recognize almost any sequence, except T at position 0. Simple and straightforward design and assembly strategies have been developed for rapid construction of TALENs (Carbery et al. 2010; Cermak et al. 2011; Sakuma et al. 2013), providing a cost-effective targeted nuclease platform.
Transcription activator-like effector nucleases technology has also been reported in induced pluripotent stem cells (iPSCs) (Hockemeyer et al. 2011), nematodes (Wood et al. 2011), plants (Li et al. 2012), zebrafish (Huang et al. 2011; Sander et al. 2011), and rats (Tesson et al. 2011). Although TALEN technologies seem to have advantages over ZFNs, there are also some ambiguous points that need to be clarified. For unknown reasons, the system appears to be less effective in rodent embryos. However, we recently showed that combined expression of exonuclease 1 (Exo1) with engineered site-specific TALENs provided highly efficient disruption of the endogenous gene in rat zygotes, and in the production of knockout rats for the albino (Tyr) gene (Mashimo et al. 2013). The microinjection of TALENs with Exo1 is an easy and efficient method of generating gene knockouts using zygotes, which increases the range of gene targeting technologies available to various species.
Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated (Cas) protein
The bacterial CRISPR/Cas system has recently been identified as an efficient gene-targeting technology in mammalian cells (Bassett et al. 2013; Friedland et al. 2013; Hwang et al. 2013; Li et al. 2013a,b; Wang et al. 2013). The system consists of a CRISPR that produces RNA components, along with the CRISPR-associated (Cas) nuclease protein. The CRISPR RNAs (crRNAs), containing short stretches of sequence homologous to specific target DNA, act as guides to direct Cas nucleases to introduce DSBs at the targeted DNA sequences. A synthetic chimeric guide RNA (gRNA) consisting of a fusion between crRNA and trans-activating crRNA (tracrRNA), directs Cas9 to cleave target DNAs that are complementary to the crRNA (Mali et al. 2013). In addition to the ability to easily generate synthetic gRNAs, a significant advantage of the CRISPR/Cas system is that multiplex genes can be targeted simultaneously with multiple targeted gRNAs. Furthermore, studies in mice have shown that homology-directed repair is preferentially activated over the NHEJ pathway when providing donor DNA templates (Wang et al. 2013; Yang et al. 2013). We have constructed the CRISPR/Cas architectures in rats, and applied it to embryos together with single-strand DNA oligonucleotides as donor templates, which efficiently generated targeted KI mutations in rats (Kazuto Yoshimi, Takehito Kaneko and Tomoji Mashimo, unpubl. data, 2013). Over the last decade, the emerging technology of next generation sequencing, and thereby genome wide association studies (Davey et al. 2011; Biesecker & Spinner 2013), has successfully identified numerous common SNPs in the human genome associated with important human diseases. As the functional testing of particular human SNP variants is a challenging proposition, accurate genome editing technologies are required for generating KI rats carrying equivalent mutations to human polymorphisms, rather than the KO models where entire coding genes are deleted. The CRISPR/Cas system provides sophisticated and flexible gene-targeting tools for generating suitable animal models of human diseases.
Advantages of the site-specific nuclease technologies
All of the artificial nuclease ZFN/TALEN/CRISPR technologies share the following advantages as compared with the conventional ES cell technology (Fig. 2). First and foremost, KO rats can be generated in a 4–6-month timeframe and with an efficiency of more than 20%. This is more favorable than the ES cell-based method for mice, which usually takes 12–18 months. Given the high rate of germ line transmission, preliminary phenotypic analysis can be performed on G1 animals after intercrossing the initial G0 founders, thereby saving time and effort. Second, gene targeting with artificial nucleases is not strain dependent (Mashimo et al. 2010), and accordingly can be performed with any inbred strain. This provides a straightforward strategy for directly using targeted gene disruption in existing strains, thereby bypassing tedious and time-consuming backcrossing steps that generally take 2–3 years to complete. Third, the artificial nuclease technologies can be used to induce a wide variety of allelic changes covering small or large deletions or insertions. It is also feasible to use targeted KI technologies that have thus far been inaccessible without rat ES cells. Since the technology does not rely on using species-specific ES cell lines, it may be possible to adapt it to other mammalian species such as pigs, cattle, and monkeys, where it is possible to harvest and manipulate fertilized embryos. The off-target effects are one of the biggest unknowns concerning the use of ZFN/TALEN/CRISPR technologies to modify the targeted genes (Radecke et al. 2010; Fu et al. 2013). It is always important to backcross the mutant lines with multiple generations to eliminate any off-target mutation and/or to validate the phenotypes with at least two independent lines.
The efficient production of inheritable genetically modified animals by artificial nuclease ZFN/TALEN/CRISPR technology will progress very rapidly. These genome-editing techniques will dramatically accelerate the development of advanced medical studies, drug design, and regenerative medicine, among other biomedical research applications, through the use of the huge number of genetically modified rats that are being produced.