Targeted mutagenesis of multiple and paralogous genes in Xenopus laevis using two pairs of transcription activator-like effector nucleases



Transcription activator-like effector nucleases (TALENs) have been extensively used in genome editing in various organisms. In some cases, however, it is difficult to efficiently disrupt both paralogous genes using a single pair of TALENs in Xenopus laevis because of its polyploidy. Here, we report targeted mutagenesis of multiple and paralogous genes using two pairs of TALENs in X. laevis. First, we show simultaneous targeted mutagenesis of three genes, tyrosinase paralogues (tyra and tyrb) and enhanced green fluorescent protein (egfp) by injection of two TALENs pairs in transgenic embryos carrying egfp. Consistent with the high frequency of both severe phenotypic traits, albinism and loss of GFP fluorescence, frameshift mutation rates of tyr paralogues and egfp reached 40–80%. Next, we show early introduction of TALEN-mediated mutagenesis of these target loci during embryogenesis. Finally, we also demonstrate that two different pairs of TALENs can simultaneously introduce mutations to both paralogues encoding histone chaperone with high efficiency. Our results suggest that targeted mutagenesis of multiple genes using TALENs can be applied to analyze the functions of paralogous genes with redundancy in X. laevis.


Xenopus laevis is a useful model organism for studying developmental biology in vertebrates, because researchers can easily introduce mRNAs and antisense oligonucleotides for genes of interest into fertilized eggs to analyze their functions (Harland & Grainger 2011). Gain- and loss-of-function analyses using these techniques are convenient and effective in early embryogenesis; however, the effects are lost over time during development. In particular, these techniques cannot be applied to post-embryogenesis research such as metamorphosis, regeneration, and reproduction. Therefore, knock-out techniques for precise analysis of gene function are extremely desirable.

Recently, genome editing using transcription activator-like effector nucleases (TALENs) has been applied for various animals (Huang et al. 2011; Sander et al. 2011; Tesson et al. 2011; Liu et al. 2012; Sung et al. 2013). TALEN architecture comprises tandem repeats of DNA-binding modules, N-terminal domain, C-terminal domain, and nuclease domain of FokI (Miller et al. 2011; Scholze & Boch 2011; Joung & Sander 2013). Each DNA-binding module recognizes a single nucleotide (A, T, G, and C) via repeat variable di-residues (RVD). TALENs can be customized by a combination of the DNA-binding modules and induce double strand breaks on the target sequence in the genome, which are then repaired mainly by non-homologous end joining (NHEJ). Because NHEJ is an error-prone repair process, insertions or deletions occur at the target site with high frequency. Consequently, the target gene is disrupted by a frameshift mutation in vivo.

More than 80% of genes in X. laevis are thought to be duplicated by polyploidization (Uno et al. 2013). Therefore, researchers need to simultaneously disrupt both paralogous genes of interest in X. laevis to avoid the issue of functional redundancy. There are two ways to solve this problem by disrupting both paralogues using TALENs. One is to design TALENs that recognize sequences conserved between the paralogues. We previously reported effective targeting and disruption of both pax6 paralogues using a single pair of TALENs, which binds to the conserved sequences (Suzuki et al. 2013). In some cases, however, appropriate sequences for a pair of TALENs that introduces mutations to both duplicated genes cannot be found owing to low homology and polymorphisms between paralogues. The other way is to design two pairs of TALENs to recognize target sequences unique to each paralogous gene. This strategy has an additional advantage in that multiple pairs of TALENs can disrupt multiple genes of interest.

Here we examined the efficiency of two pairs of TALEN injection in disrupting multiple and paralogous target genes in X. laevis embryos. We also checked for the timing of introduction of mutations after injection of TALEN mRNA during early embryogenesis. Our findings demonstrated that TALENs enable targeted mutagenesis of multiple and paralogous genes in vivo for functional analysis in X. laevis, which has a complex genome resulting from polyploidization.

Materials and methods

Construction of TALENs and mRNA synthesis

All TALEN plasmids were constructed according to a previous report (Sakuma et al. 2013a), with various modifications (Sakuma et al. 2013a). Briefly, DNA-binding repeats of TALE harboring non-RVD amino acid variations (Sander et al. 2011) were synthesized and cloned into pBluescript SK vector. Each TALE repeat was assembled into modified final destination vector containing the N- and C-terminal domains of TALE and the FokI nuclease domain (Sakuma et al. 2013a) using the two-step Golden Gate cloning method (Sakuma et al. 2013a). TALEN target sequences of the enhanced green fluorescent protein (egfp) and tyrosinase (oculocutaneous albinism IA; tyr) genes were identical to previous reports (Nakagawa et al. 2013; Sakuma et al. 2013a; Suzuki et al. 2013; Sakuma et al. 2013a). Target sequences of the tyr were located in regions of exon 1 that are conserved between both paralogues, tyra and tyrb. TALEN target sites for the no29 and npm3 genes were designed using the predicted exon 2 sequences from National Center for Biotechnology Information (NCBI) Refseq (Motoi et al. 2011; accession numbers, NM_001088231 and NM_001093806, respectively). After linearization of plasmids by SmaI, mRNAs were synthesized using mMessage mMachine T7 Ultra Kit (Life Technologies, Carlsbad, CA, USA) as described previously (Suzuki et al. 2013).

Manipulation of X. laevis eggs and mRNA microinjection

Fertilized X. laevis eggs were obtained from [CMV:egfp] transgenic or wild-type breeding females and wild-type males injected with human chorionic gonadotropin (Aska Pharmaceutical, Tokyo, Japan). Eggs were dejellied with 2% cysteine and washed in 0.1× Marc's modified ringer (MMR) according to Sive et al. (2000). Washed eggs were transferred into 5% Ficoll (Sigma-Aldrich, St. Louis, MO, USA) in 0.3× MMR and were injected with TALEN mRNAs at the one-cell stage using Nanoject II (Drummond, Broomall, PA, USA). Injected embryos were reared to different developmental stages (Nieuwkoop & Faber 1994) in 0.1× MMR at 20°C. Animals were maintained and used in accordance with the Hiroshima University guidelines for the use and care of experimental animals.

Mutation analysis

Genomic DNA was extracted using a DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). The uninjected and tyr/egfp TALEN-injected embryos were collected at the morula (n = 10), blastula (n = 10), gastrula (n = 5), and the swimming tadpole stages. no29/npm3 TALEN-injected embryos were sampled at the swimming tadpole stage. The target loci were amplified with specific primer sets, described in Table S1, using LA Taq polymerase (Takara Bio, Shiga, Japan). Amplified polymerase chain reaction (PCR) products were purified with a QIAquick PCR Purification Kit (Qiagen) and then subjected to heteroduplex mobility assay (HMA) and restriction fragment length polymorphism (RFLP) analysis. For RFLP analysis, we digested the purified PCR products of the tyra/tyrb and egfp with HinfI and AccII, respectively. PCR products were electrophoresed in 3% agarose gel and stained with ethidium bromide. PCR products were subcloned into pCR2.1/TOPO (Life Technologies) by TA cloning. Colony PCR was performed to select positive clones, which were then subjected to DNA sequencing.


Simultaneous targeted disruption of multiple genes by TALENs

To examine whether TALENs can disrupt multiple target genes, mRNAs of two TALEN pairs against tyr and egfp (250 pg each) were co-injected into the one-cell stage eggs from wild-type males and [CMV:egfp] transgenic females. TALEN-injected embryos showed a drastic loss of pigmentation in the retinal pigment epithelium (RPE) and melanophores (Fig. 1, upper images). Moreover, GFP fluorescence was also attenuated in the TALEN-injected embryos (Fig. 1, lower images).

Figure 1.

Phenotypes of the tyr/egfp transcription activator-like effector nuclease (TALEN)-injected embryos. Bright (upper) and fluorescence (lower) images show pigmentation and green fluorescent protein (GFP) fluorescence, respectively. Uninjected, uninjected embryos. tyr/egfp TALENs, embryos co-injected with each pair of TALEN mRNAs targeting tyr and egfp. Genomic DNA of the uninjected (an asterisk) and the TALEN-injected (i, ii, and iii) embryos were subjected to heteroduplex mobility assay (HMA), restriction fragment length polymorphism (RFLP) analysis, and DNA sequencing (see Fig. 3).

The TALEN-injected embryos were divided into four groups according to loss of pigmentation in the RPE; severe, moderate, weak, and normal (Fig. 2A), as described in the previous report (Suzuki et al. 2013). More than 50% of the TALEN-injected embryos showed severe phenotypes that lacked pigmentation in not only the RPE, but also the melanophores in the trunk region. The survival rate of the TALEN-injected embryos was over 80%. Another phenotypic trait, loss of GFP fluorescence, correlated well with loss of pigmentation (albinism) in the TALEN-injected embryos (Fig. 2B).

Figure 2.

Phenotype frequency and toxicity of the tyr/egfp transcription activator-like effector nuclease (TALEN)-injected embryos. (A) Percentage of phenotypes in the uninjected embryos (Uninjected) and the tyr/egfp TALEN-injected embryos (tyr/egfp TALENs). Except for abnormally developed embryos, phenotypes were divided into four groups (severe, moderate, weak, and normal) according to the extent of loss of pigments in the RPE. Total numbers of individuals are shown at the top of each graph. (B) Correlation between albinism and loss of green fluorescent protein (GFP) fluorescence in the TALEN-injected embryos. Corresponding images in each embryo show phenotypes of pigmentation (severe, moderate, weak, and normal; upper images) and GFP fluorescence (lower images).

Next, we examined the frequency and rate of mutations introduced by TALENs using HMA, RFLP analysis, and DNA sequencing. At the swimming tadpole stage, genomic DNA was individually extracted from each embryo in Figure 1, and then was subjected to different analyses. HMA detects introduced mutations in target genes as the up-shifted bands that represent heteroduplex forms of wild and mutant alleles in genomic DNA PCR (Nakagawa et al. 2013; Ota et al. 2013; Suzuki et al. 2013). Up-shifted bands were detected in PCR products of tyra, tyrb, and egfp from the TALEN-injected embryos, whereas these bands were not observed in the uninjected embryo (Fig. 3A; arrows in upper images). In RFLP analysis, PCR products from the uninjected embryo were almost completely cut by restriction enzymes, because the spacer sequences of tyra/tyrb and egfp contain HinfI and AccII recognition site, respectively (Fig. 3A; cut bands, closed arrowheads in lower images). In contrast, most PCR products from the TALEN-injected embryos could no longer be cut by these enzymes, due to alteration of the spacer sequences (Fig. 3A; uncut bands, open arrowheads in lower images). As the result of DNA sequencing analysis, we found that all 10 clones of tyra and tyrb carried deletions: 50% and 80% of total clones were observed as frameshift mutations in tyra and tyrb, respectively (Fig. 3B). Similarly, seven clones of egfp carried deletions, of which four had frameshift mutations (40% of total clones). The mutation rate of egfp was somewhat lower, probably due to multiple copies of egfp integrated in the transgenic embryos generated by sperm nuclear transfer (Kroll & Amaya 1996).

Figure 3.

Detection of transcription activator-like effector nuclease (TALEN)-mediated mutations in the tyr/egfp TALEN-injected embryos. (A) Detection of mutations by heteroduplex mobility assay (HMA) and restriction fragment length polymorphism (RFLP) analysis. The uninjected (control, C) and TALEN-injected (i, ii, and iii) embryos correspond to the phenotypes in Figure 1. Upper gel images show the results of HMA. Heteroduplex polymerase chain reaction (PCR) products are visible as up-shifted bands (arrows). Lower gel images indicate the results of RFLP analysis. PCR products of tyra/tyrb and egfp were digested by HinfI and AccII, respectively. PCR products with introduced mutations are observed as uncut bands. Open and closed arrowheads represent uncut and cut PCR products, respectively. (B) Sequences of TALEN target sites in tyra, tyrb, and egfp. PCR products of each target site from the tyr/egfp TALEN-injected embryo (i in Fig. 1) were subcloned and sequenced. Wild-type sequences are shown at the top with the target sites indicated as capital and highlighted letters. Sequences are aligned to the wild-type. Deletions and insertions are shown by dashed lines and blue characters, respectively. Types of mutations and each frequency are shown at the right of the panel.

Timing of introduction of mutations by TALENs during early development

To examine when mutations occur in the target loci after the injection of TALEN mRNA, we performed HMA and RFLP analysis using genomic PCR products from the uninjected and tyr/egfp TALEN-injected embryos at the morula, blastula, and gastrula stages. As the results of HMA and RFLP analysis, up-shifted bands and uncut bands were already detectable in all target genes from the TALEN-injected embryos at the morula stage (Fig. 4). PCR products of tyra and tyrb from the TALEN-injected embryos were almost completely uncut at the gastrula stage. In the case of egfp, although cut PCR bands were slightly observed at the gastrula stage, they finally became almost undetectable at the swimming tadpole stage (see Fig. 3A). In contrast, both up-shifted and uncut bands were not detected in PCR products of the three genes from the uninjected embryos throughout all stages indicated.

Figure 4.

Timing of introduction of mutations after transcription activator-like effector nucleases (TALENs) mRNA injection. TALEN-mediated mutations in tyra, tyrb, and egfp at the early stages of embryogenesis were detected by heteroduplex mobility assay (HMA) (upper) and restriction fragment length polymorphism (RFLP) analysis (lower). Genomic DNA was extracted from 5 to 10 embryos at the stages indicated at the top of the gel electrophoresis images. Arrows, open and closed arrowheads are described in Figure 3A. C and T indicate the uninjected embryos and the TALEN-injected embryos, respectively.

Targeted mutagenesis of both histone chaperone paralogues using two pairs of TALENs

To test whether low homology paralogues can be simultaneously mutated by two different pairs of TALENs, we introduced two pairs of TALENs that separately target exon 2 of both histone chaperone paralogues, no29 and npm3 (Motoi et al. 2011). We could not design a single pair of TALENs that recognize both paralogues because their coding regions are short and have polymorphisms. mRNAs of both pairs of no29 and npm3 TALENs (125 and 250 pg each) were co-injected into the fertilized eggs at the one-cell stage. Embryos were reared to the swimming tadpole stage and then genomic DNA was subjected to HMA and DNA sequencing. As the result of HMA, up-shifted bands were detected in the TALEN-injected tadpoles, whereas these bands were not found in the uninjected tadpole (Fig. 5A; arrows). The lower band of npm3 in the high dose (250 pg each) lane may be a large deletion product (Fig. 5A; an asterisk). As shown in Figure 5B, DNA sequencing analysis demonstrated high mutation rates of the two targeted paralogues in the 250 pg each TALEN-injected tadpole: 90% in no29 (total 10 clones) and 100% in npm3 (total nine clones). In the case of no29, seven and two clones had deletions and insertions, respectively. On the other hand, all nine clones of npm3 carried deletions containing two large deletion clones that may correspond to the lower band in HMA (Fig. 5A; an asterisk). In this case, major mutations of both paralogues were 6-bp deletions that did not cause frameshift.

Figure 5.

Transcription activator-like effector nuclease (TALEN)-mediated mutagenesis in both paralogues, no29 and npm3. (A) Detection of mutations by heteroduplex mobility assay (HMA). Genomic DNA was extracted from the uninjected (C) and the TALEN-injected tadpoles (125 pg each and 250 pg each) at the swimming tadpole stage. Arrows represent up-shifted bands. The band indicated by an asterisk may be a large deletion in the 250 pg each TALEN-injected tadpole. (B) Sequences of the no29 and npm3 target sites in the 250 pg each TALEN-injected tadpole. Wild-type sequences are shown at the top with the target sites represented by capital and highlighted letters. Sequences of mutations are aligned to the wild-type. Deletions and insertions are represented by dashed lines and blue characters, respectively. Polymorphisms between paralogues are indicated by red characters. Types of mutations and frequency of each are shown at the right of the panel.


Transcription activator-like effector nucleases-mediated mutagenesis has recently been used to functionally analyze genes of interest in various animals in vivo (Huang et al. 2011; Sander et al. 2011; Tesson et al. 2011; Liu et al. 2012; Sung et al. 2013). In amphibians, TALENs have been applied in Xenopus tropicalis, X. laevis, and Pleurodeles waltl (Ishibashi et al. 2012; Lei et al. 2012; Nakajima et al. 2013; Suzuki et al. 2013; Hayashi et al. 2013a). Fertilized eggs from these species can be easily manipulated and injected with high doses of TALEN mRNA, which contributes to highly-efficient targeted gene disruption. Thus, TALEN technology is suitable for various research fields using amphibians.

In this report, we showed that three individual genes, tyra, tyrb, and egfp, can be simultaneously and efficiently disrupted by injection of two pairs of TALENs into fertilized eggs. Consistent with over 50% of severe phenotypes in the TALEN-injected embryos, frameshift mutation rates of three genes were high: 50%, 80%, and 40% in tyra, tyrb, and egfp, respectively. Moreover, even though two pairs of TALEN mRNA were injected, there was no significant decrease in phenotype frequency or increase in toxicity, as compared with injection of only one pair (Sakuma et al. 2013a; Suzuki et al. 2013). Recently, many studies demonstrated that the CRISPR/Cas9 system is very convenient and effective for multiplexed genome editing in vivo (Jao et al. 2013; Wang et al. 2013). However, our results suggest that the TALEN technology could also be applied to multiple gene disruptions in X. laevis.

To address the issue regarding the redundancy of paralogues, we also attempted to simultaneously induce targeted mutagenesis of two paralogues using two different pairs of TALENs. As expected, two pairs of TALENs introduced mutations into target sites of no29 and npm3 with an efficiency of over 90%. Over half of the clones showed 6-bp deletions, consequently frameshift of these genes were not introduced in this case. Given that in-frame mutations may occur with triplet insertions or deletions, we need to carefully design TALEN recognition sites to efficiently disrupt target genes. For example, 5′- or 3′-splice sites of exon-intron junctions to disrupt mRNA splicing, and coding regions of crucial domains involved in protein function, may be good candidates for TALEN recognition sites.

Our results also indicate that TALEN-mediated mutagenesis begins to occur at least by the morula stage about 5 h after fertilization. Furthermore, in RFLP analysis, digested PCR products of tyra and tyrb were hardly detectable at the gastrula stage. This result suggests that TALENs possess the potential to disrupt target genes at relatively early stages, even though four copies of the genes exist in the genome. Thus, such early introduction of targeted mutagenesis by TALENs allows us to perform loss of function analysis in F0 X. laevis embryos and tadpoles.

Taken together, targeted mutagenesis using two pair TALENs enables us to analyze functions of genes in X. laevis, most of which are duplicated with redundancy. Furthermore, this technique has the potential to alter sequences of multiple genes in vivo.


This research was supported by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS), KAKENHI Grant Number 25124708 to K. T. S.

Competing interests

The authors declare no competing financial interests.