Xenotransplantation 2014: 21: 291–300 doi: 10.1111/xen.12089

© 2014 John Wiley & Sons A/S XENOTRANSPLANTATION

Brief Communication

The combinational use of CRISPR/ Cas9-based gene editing and targeted toxin technology enables efficient biallelic knockout of the a-1,3-galactosyltransferase gene in porcine embryonic fibroblasts Sato M, Miyoshi K, Nagao Y, Nishi Y, Ohtsuka M, Nakamura S, Sakurai T, Watanabe S. The combinational use of CRISPR/Cas9based gene editing and targeted toxin technology enables efficient biallelic knockout of the a-1,3-galactosyltransferase gene in porcine embryonic fibroblasts. Xenotransplantation 2014: 21: 291–300. © 2014 John Wiley & Sons A/S. Abstract: Background: The recent development of the type II clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 system has enabled genome editing of mammalian genomes including those of mice and human; however, its applicability and efficiency in the pig have not been studied in depth. Here, using the CRISPR/Cas9 system, we aimed to destroy the function of the porcine a-1,3-galactosyltransferase (a-GalT) gene (GGTA1) whose product is responsible for the synthesis of the a-Gal epitope, a causative agent for hyperacute rejection upon pig-to-human xenotransplantation. Methods: Porcine embryonic fibroblasts were transfected with a Cas9 expression vector and guide RNA specifically designed to target GGTA1. At 4 days after transfection, the cells were incubated with IB4 conjugated with saporin (IB4SAP), which eliminates a-Gal epitopeexpressing cells. Therefore, the cells surviving after IB4SAP treatment would be those negative for a-Gal epitope expression, which in turn indicates the generation of GGTA1 biallelic knockout (KO) cells. Results: Of the 1.0 9 106 cells transfected, 10–33 colonies survived after IB4SAP treatment, and almost all colonies (approximately 90%) were negative for staining with red fluorescence-labeled IB4. Sequencing of the mutated portion of GGTA1 revealed a frameshift of the a-GalT protein. Porcine blastocysts derived from the somatic cell nuclear transfer of these a-Gal epitope-negative cells also lacked the a-Gal epitope on their surface. Conclusions: These results demonstrated that the CRISPR/Cas9 system can efficiently induce the biallelic conversion of GGTA1 in the resulting somatic cells and is thus a promising tool for the creation of KO cloned piglets.

Introduction

Gene targeting is now recognized as a powerful tool for the analysis of diverse aspects of gene function in vivo. Unfortunately, besides mice and rats, to date, no pluripotent stem cells with germline competency have been generated from large

Masahiro Sato,1 Kazuchika Miyoshi,2 Yozo Nagao,2 Yohei Nishi,2 Masato Ohtsuka,3 Shingo Nakamura,4 Takayuki Sakurai5 and Satoshi Watanabe6 1

Section of Gene Expression Regulation, Frontier Science Research Center, Kagoshima University, Kagoshima, Japan, 2Laboratory of Animal Reproduction, Faculty of Agriculture, Kagoshima University, Kagoshima, Japan, 3Division of Basic Molecular Science and Molecular Medicine, School of Medicine, Tokai University, Kanagawa, Japan, 4 Department of Surgery, National Defense Medical College, Saitama, Japan, 5Department of Organ Regeneration, Graduate School of Medicine, Shinshu University, Nagano, Japan, 6Animal Genome Research Unit, Division of Animal Science, National Institute of Agrobiological Sciences, Ibaraki, Japan

Key words: a-1 – 3-galactosyltransferase – biallelic knockout – CRISPR/Cas9 – porcine embryonic fibroblasts – somatic cell nuclear transfer – targeted toxin Address reprint requests to Masahiro Sato, Ph.D., Section of Gene Expression Regulation, Frontier Science Research Center, Kagoshima University, 835-1 Sakuragaoka, Kagoshima 890-8544, Japan (Email: [email protected]) Received 22 November 2013; Accepted 20 January 2014

animals. Therefore, gene targeting in these species has solely depended on somatic cell nuclear transfer (SCNT) of gene targeted (KO) cells using traditional gene targeting technology [1]. Due to the extremely low frequency of homologous recombination in somatic cells (< 10
6), this process is highly inefficient. To date, only a few successful 291

Sato et al. examples have been achieved, including cells lacking the a-1,3-galactosyltransferase (a-GalT) gene (GGTA1) [2–4], cystic fibrosis transmembrane conductance receptor gene [5,6], and X-linked interleukin-2 receptor gamma chain gene [7]. The low efficiency of generating cells with a biallelic (homozygous) KO phenotype is also one of the main hurdles for performing gene targeting using large animals because it requires mating between monoallelic (heterozygous) KO males and females to produce biallelic KO animals, which requires considerable cost and time. Recently, zinc-finger nucleases (ZFNs) and transcription activator-like effectors (TALENs) have emerged as powerful and efficient tools for gene editing [8]. To date, several studies have demonstrated high-efficiency gene targeting in various animals, for example, pigs [9–12], mice [13], rats [14], Xenopus [15], and zebrafish [16,17]. This gene editing system involves gene-specific alterations, including insertions or deletions (indels), via the generation of non-homologous end-joining (NHEJ) induced by ZFN- or TALEN-mediated double-strand breaks. However, these systems are complex and expensive to design, and their assembly is labor intensive [18]. Very recently, the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPRassociated (Cas) system has been developed as third generation technology for gene editing [19,20]. The Cas9 nuclease from Streptococcus pyogenes (hereafter referred to as Cas9) can be guided by simple base-pair complementarity between the first 20 nucleotides of an engineered guide RNA (gRNA) and a target genomic DNA sequence of sequence NGG, which is called a “protospacer adjacent motif” [20]. gRNA facilitates the fusion of interfering CRISPR RNA (crRNA) to transactivating crRNA, which is complementary to the repeat sequences in the pre-crRNA. The CRISPR/ Cas9 system mediates Cas9/gRNA-induced NHEJ-based gene disruption and oligo-mediated homology directed repair events, leading to highly efficient targeting in several species including yeast [21], Drosophila [22], mice [23–25], humans [26,27], and zebrafish [28]. However, the applicability and efficiency of this system in the pig have not been studied in depth. Here, we aimed to destroy the function of porcine GGTA1 as a model case using the CRISPR/ Cas9 system. a-GalT is involved in the synthesis of the a-Gal epitope, a causative agent of hyperacute rejection upon pig-to-human xenotransplantation [29]. In this case, we aimed to isolate cells lacking the expression of the a-Gal epitope through KO of GGTA1 using the CRISPR/Cas9 system, because it 292

allows the efficient production of cells for the creation of SCNT-mediated cloned piglets. We employed targeted toxin technology using isolectin BS-I-B4 conjugated with saporin (IB4SAP) [30] to eliminate untransfected and monoallelic KO porcine cells, all of which should express the a-Gal epitope on their surface [31]. The colonies surviving after treatment with IB4SAP are considered biallelic KOs for GGTA1.

Materials and methods Cell lines and culture

The porcine embryonic fibroblasts (PEFs) used throughout this study were primarily cultured from male fetuses of Clawn miniature swine (Japan Farm Ltd., Kagoshima, Japan) at 30 days after insemination. The cells were grown in PEF culture medium consisting of Dulbecco’s modified Eagle’s medium/Ham’s F-12 (#124; Wako Pure Chemical Industries Ltd., Osaka, Japan), 10% fetal bovine serum (FBS), and 1% antibiotic–antimycotic solution (#A5955; Sigma-Aldrich Co. Ltd., St. Louis, MO, USA) at 38.5 °C in a humidified atmosphere of 5% CO2 in air. The cells were passaged 5–7 times and then frozen. Frozen cells were thawed and passaged for 7–14 generations prior to transfection. Cas9/gRNA design

The hCas9 expression vector (Fig. S1) carrying a codon-optimized Cas9 gene was purchased from Addgene (http://www.addgene.org). The gRNA expression vector with the human U6 promoter and gRNA scaffold (Fig. S1) was also from Addgene. The construction of the gRNA expression vector was based on the method shown on the Addgene website (http://www.addgene.org/static/ data/85/85/e19394c4-5e76-11e2-a7c4-003048dd6500. pdf). A pair of oligos (pA3GalTgRNA1s: TTTCTT GGCTTTATATATCTTGTGGAAAGGACGAA ACACCGAGAAAATAATGAATGTCAA; pA3G alTgRNAr: GACTAGCCTTATTTTAACTTGCTATTTCTAGCTCTAAAACTTGACATTCATTATTTTCTC) was designed to cleave a region following the ATG codon of GGTA1, which can cause a frameshift of the a-GalT protein. These primers were annealed, phosphorylated, and ligated to the linearized vector using a Gibson Assemblyâ Cloning Kit (New England Biolabs Japan, Inc., Tokyo, Japan). These recombinant plasmids were amplified in Escherichia coli JM109 and finally purified using a QIAGEN Midi Kit (Hilden, Germany).

Simplified acquisition of a-GalT KO pig cells Transfection and subsequent selection

PEFs were transfected using an electroporationbased Lonza Nucleofector system (Lonza Biologics, Cologne, Germany), because a relatively high transfection efficiency (> 50%) has been achieved in PEFs using this system [32]. PEFs (1.0 9 106 cells) were electroporated in 100 ll nucleofector solution (for primary fibroblasts) containing the hCas9 expression vector + gRNA expression vector + pmaxGFP (Lonza Biologics) (each 2 lg). pmaxGFP was employed for monitoring transfection efficiency. After transfection, the cells were plated in a 100-mm tissue culture dish (Iwaki Glass Co. Ltd., Tokyo, Japan) containing PEF culture medium. At 2 days after transfection, the cells were trypsinized. We subjected approximately 4.0 9 105 cells to cytochemical staining using 2 lg/ml AF594-IB4 (#I21413; Invitrogen, Carlsbad, CA, USA) in phosphate-buffered saline without Ca2+ and Mg2+ (PBS[
]; pH 7.4), 2% FBS, and 1 mM CaCl2 for 2 h at 4 °C. The cells were then inspected under a fluorescence microscope, as described below. These stained cells were also subjected to FACS analysis, as described below. Furthermore, approximately 1.0 9 105 cells were subjected to the Surveyor nuclease assay, as described below. The remaining cells were plated in a 100-mm dish and cultured in PEF culture medium. At 4 days after transfection, the cells were trypsinized and divided into two parts: one was deepfrozen for back up, while the other (containing 1.0 9 105 cells) was subjected to IB4SAP treatment (#IT-10; Advanced Targeting Systems, Inc., San Diego, CA, USA). For IB4SAP treatment, the cells were incubated at 37 °C for 2 h in a solution (25 ll) containing 0.5 lg IB4-SAP in PBS/FBS/ CaCl2. The treated cells were returned directly to a 100-mm dish containing normal PEF culture medium and cultured for an additional 10 days. In the control, a small number of PEFs (1.0 9 104 cells) were plated in a 100-mm dish and cultured in normal medium for 10 days, at which time colonies should have emerged. At 10 days after IB4SAP treatment, the cells were fixed in the culture dish with 4% PFA in PBS [
] for 10 min at room temperature and then stained with AF594-IB4. For the propagation of colonies, emerging colonies were picked using a small paper disk (3 MM Whatman paper) that had been dipped in 0.25% trypsin/StemProâ Accutaseâ (Life Technologies Corporation, Carlsbad, CA, USA) and transferred directly into a 48-well plate (Iwaki Glass Co. Ltd.) containing PEF

culture medium. The cells were cultured for 10– 20 days until confluency. Upon passage, a portion of the cells was subjected to cytochemical staining with AF594-IB4, as described previously. Detection of fluorescence

Fluorescence in the cells was examined using an Olympus BX60 fluorescence microscope (Olympus, Tokyo, Japan) with DM505 (BP460-490 and BA510IF; Olympus) and DM600 filters (BP545580 and BA6101F; Olympus), which were used to detect pmaxGFP-derived green fluorescence and AF594-derived red fluorescence, respectively. Microphotographs were taken using a digital camera (FUJIX HC-300/OL; Fuji Film, Tokyo, Japan) attached to the fluorescence microscope and printed using a Mitsubishi digital color printer (CP700DSA; Mitsubishi, Tokyo, Japan). Fluorescence activated cell sorting (FACS) analysis

Two days after transfection with CRISPR/Cas9related plasmids, cells were harvested with 0.25% trypsin/0.01% EDTA. One group was subjected to staining with AF594-IB4 lectin at 4 C for 1 h, washed with PBS[
]/FBS, resuspended in PBS[
]), and then sorted on ice before analysis. The other group was unstained and used as a negative control. The intact PEFs were also stained with AF594-IB4 lecti and used as a positive control. Analysis was performed using a FACSCalibur system (BD Biosciences, Heidelberg, Germany). Data were analyzed using CellQuest Pro software (BD Biosciences). Cas9/gRNA mutation screens by Sanger sequencing

Genomic DNA of PEFs was extracted according to the method of Blin and Stafford [33] with several modifications [34] and finally dissolved in 50 ll H2O. We subjected 1 ll of this solution (approximately 30 ng) to PCR/nested PCR for sequencing. For the isolation of genomic DNA from blastocysts, blastocysts were lysed in a drop (30 ll) of lysis buffer (30 mM Tris–Cl, 10 mM EDTA, and 1% SDS) containing 0.125 lg/ml proteinase K and 0.125 lg/ml Pronase E in a Terasaki microtest plate (Nunc, Roskilde, Denmark) and covered with paraffin oil to avoid evaporation in a 37 °C incubator overnight, and two drops containing a total of eight blastocysts were mixed. The solution was then phenol-extracted and subjected to ethanol precipitation. The pellet was dissolved in 10 ll dH2O. We subjected 1 ll of this solution (< 10 ng) to PCR/nested PCR for sequencing. 293

Sato et al. First-round PCR was performed using the above DNA as a template to amplify the genomic region surrounding the CRISPR target site for GGTA1. In brief, PCR amplification reactions were performed in a total volume of 20 ll containing 10 ll of 2 9 Ampdirect Plus (Shimadzu, Kyoto, Japan), 0.5 U Nova Taq TM Hot Start DNA polymerase (Novagen, UK, 5 pmol of a primer set (Ex4-S [50 GCAAATTAAGGTAGAACGCA-30 ]/Ex4-RV [50 GCTGCCCCTGAGCCACAACG-30 ]; Fig. 1A), and 1 ll genomic DNA (approximately 0.5 ng) using a thermal cycler (PC-708; Astec, Fukuoka, Japan). The following PCR conditions were used: initial denaturation at 95 °C for 10 min followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 45 s, and extension at 72 °C for 1 min. Nested PCR was performed with the same conditions used for the 1st PCR using 1 ll of the 1st PCR products and a primer set (Ex4-3S [50 AGAAAAGATATTGGTATAAG-30 ]/Ex4-3RV [50 CAGTTGAGACAAGCAGCATT-30 ]; Fig. 1A). The PCR products were separated by electrophoresis on 1.5% (w/v) agarose gels, stained with ethidium bromide, and visualized using a UV transilluminator. In some cases, the resulting nested PCR products were cloned into the pCR2.1 vector (Invitrogen), which was used for Sanger sequencing (Fasmac, Atsugi, Japan). Surveyor nuclease assay

The Surveyor nuclease assay was performed using a Surveyor Mutation Detection Kit (Transgenomic, Inc., Omaha, NE, USA). Briefly, genomic DNA from cells transfected with CRISPR/Cas9-related plasmids and from control cells was extracted. First-round PCR was performed using a primer set (Ex4-S/Ex4-RV) under the conditions mentioned previously. Nested PCR was with the same conditions used for the first-round PCR using 1 ll of the first-round PCR products and a primer set (Ex4-2S [50 -CTCCTTAGCGCTCGTTGGCT-30 ]/Ex4-2RV [50 -GCAACTCTCTGGAATGCTTT-30 ]; Fig. 1A). Equal amounts of sample DNA (200 ng) plus control DNA (200 ng) or control DNA (400 ng) alone were mixed and subjected to a re-annealing process to enable heteroduplex formation: 95 °C for 10 min, 95–85 °C ramping at 2 °C/s, 85–25 °C at 0.3 °C/s, and holding at 25 °C for 1 min. After re-annealing, the products were treated with Surveyor nuclease and Surveyor enhancer S, following the manufacturer’s recommended protocol, and analyzed on 3% agarose gels. The gels were stained with 0.5 lg/ml ethidium bromide and imaged with a gel-imaging system (Image J 294

software). Quantification was based on relative band intensity. Somatic cell nuclear transfer (SCNT)

SCNT was performed according to the method of Sato et al. [35] and Miyoshi et al. [36,37]. Each nucleus from the a-Gal epitope-negative clone CRISPR-2-1 or untransfected PEFs was introduced into a single enucleated oocyte using micromanipulators. The development of SCNTtreated embryos was evaluated by the rates of cleavage and blastocyst formation at 2 and 7 days of culture, respectively. Results

To test whether the CRISPR/Cas9 system could produce targeted cleavage in the porcine genome, we co-transfected a plasmid expressing mammalian-codon-optimized Cas9 and one expressing a gRNA targeting GGTA1 (Fig. 1; Fig. S1) together with pmaxGFP (Lonza GmbH, Wuppertal, Germany; for monitoring transfection efficiency) into PEFs. Two days after transfection, we determined the efficiency of targeted cleavage using the Surveyor assay. Transfection efficiency, as evaluated by counting fluorescent cells under a fluorescence microscope, was approximately 70% (data not shown). The transfectants displayed a cleavage efficiency of 7.9% at GGTA1 (Fig. 1B). FACS analysis of these cells stained with Alex Fluor 594conjugated IB4 (AF594-IB4), which specifically binds to the a-Gal epitope [31], revealed that approximately 13.2% of cells were a-Gal epitopenegative (Fig. 1C). The presence of a-Gal epitopenegative cells was also confirmed by inspection using a fluorescence microscope (Fig. S2). Next, we attempted to isolate a-Gal epitope-negative porcine cells by transfecting PEFs (1.0 9 106 cells) with the above three plasmids and subsequently treating them with IB4SAP. At 4 days after transfection, the cells were trypsinized. We incubated 1.0 9 105 cells in a solution (25 ll) containing 0.5 lg IB4SAP for 2 h at 37 °C and then plated the cells in a 100-mm dish containing normal medium. The culture was maintained for up to 10 days, at which time colonies of the appropriate size should have appeared. During this period, non-transfected cells and cells (probably monoallelic KO cells) that were successfully transfected, but still expressing the a-Gal epitope, die due to IB4SAP-mediated cytotoxicity, while cells in which both alleles of GGTA1 are completely knocked out by CRISPR/Cas9-mediated gene disruption survive, because IB4SAP is unable to bind these a-Gal

Simplified acquisition of a-GalT KO pig cells

Fig. 1. (A) Schematic representation of nucleotide sequences between the target locus (around exon 4 of GGTA1) and GGTA1targeting crRNA. The red arrow indicates the putative cleavage site. The primers used are shown above the sequence of GGTA1. (B) Surveyor assay for Cas9-mediated indels. A 320-bp PCR product is cleaved into two fragments, namely 120 bp and 100 bp. Lane 1, mixture of genomic DNA derived from PEFs transfected with CRISPR/Cas9 plasmids and from untransfected PEFs; Lane 2, genomic DNA from untransfected PEFs. m, 100-bp ladder marker. (C) FACS analysis of the a-Gal epitope on PEFs transfected with CRISPR/Cas9 plasmids. The transfected cells that have been stained with AF594-IB4 lectin are shown as red line. The transfected cells that have been unstained with the lectin are shown as light gray. The intact PEFs that have been stained with the lectin are shown as black. The result reveals that approximately 13.2% the transfected cells (red line) are located near the area where a-Gal epitope-negative cells (light gray) are located. (D) Staining of colonies with AF594-IB4. a, b: A colony surviving after IB4SAP treatment. c, d: Another colony surviving after IB4SAP treatment. e, f: A colony derived from intact PEFs. a, c, e: Photographs taken under light; b, d, f: photographs taken under UV light. Bars = 50 lm. (E) Staining of trypsinized cells with AF594-IB4. a–c: Cell clone (CRISPR-2-1) propagated from a colony surviving after IB4SAP treatment. d–f: Intact PEFs. a, d: Photographs taken under light; b, c, e, f: photographs taken under UV and light. Bars = 5 lm. (F) Upper panel, List of mutations in several clones. Inserted base is indicated by red. Red bars indicate deletion of base(s). PAM, protospacer adjacent motif. Lower panels, Sequence data from the clone CRISPR-2-1 and untransfected PEFs. The arrow indicates the insertion of one base (T) into the sequence of exon 4 of GGTA1 in CRISPR-2-1.

295

Sato et al. that in the clone CRISPR-2-1. In the lower panels of Fig. 1F, the sequence data of the CRISPR-2-1 clone are shown as an example together with those of untransfected PEFs. The addition of one base (T) following the ATG would cause an insertioninduced frameshift that disables the encoded protein, such as MNVQRKSGSVNAACLN stop codon, which is clearly different from the wildtype amino acid sequence (MNVKGRVVLSMLL VST—). Notably, the former is expected to truncate the GGTA1 open reading frame prematurely. The a-Gal epitope-negative cell clone CRISPR2-1 was subjected to SCNT to examine whether it has the potential to support the development of cloned embryos and whether the resulting embryos (blastocysts) lack the expression of the a-Gal epitope. Untransfected PEFs were also subjected to SCNT as controls. At 7 days after SCNT, the number of developing blastocysts was scored. In the experimental group, in which the CRISPR-2-1 clone had been introduced, approximately 32% (25/79) of the treated embryos developed to blastocysts (Table 2). A similar developmental rate (30% [21/69]) was also obtained when untransfected PEFs were used as donors (Table 2). These blastocysts were then fixed and subsequently subjected to staining with AF594-IB4. In the experimental group, none of the embryos (six tested) were stained by the lectin (a and b of Fig. 2A). According to Chi et al. [38], weak staining was seen in the zona pellucida (ZP) of some parthenogenetic blastocysts. Consistent with this report, we observed weak staining in the ZP of one embryo (arrow in b of Fig. 2A), but the cell surface membrane was completely negative for such staining (arrowhead in b of Fig. 2A). In contrast, the control embryos exhibited strong staining for IB4 (c and d of Fig. 2A). To strengthen the difference in staining between the two groups, embryos from the two groups were mixed and then photographed under the same UV illumination. Clearly, embryos in the experimental group were negative for IB4 staining, while those in the control group were positive (e and f of Fig. 2A).

epitope-negative cells. To determine whether these surviving colonies were a-Gal epitope-negative, one dish was subjected to fixation with 4% paraformaldehyde (PFA) and subsequent staining with AF594-IB4. Of the 10 colonies examined, nine were completely negative for the a-Gal epitope (a– d of Fig. 1D) and one was positively stained (data not shown). All of the colonies that had been derived from sparsely plated untransfected PEFs exhibited positive staining for AF594-IB4 (e and f of Fig. 1D). These data suggest that 0.013% (= 100 9 9 9 0.7 9 10
5) of transfected PEFs were biallelic KO for GGTA1 (Table 1). Similar rates were also obtained when transfection was performed on different days of the 10-day incubation period (Table 1). Several of the colonies that survived after IB4SAP treatment were picked up using a paper method [32] and propagated by cultivating them in normal PEF medium. The other remaining colonies (< 10) were collected by trypsinization and then deep-frozen. We obtained 6 clones, called CRISPR-2-1 to -6. Staining with AF594-IB4 demonstrated that CRISPR-2-1 was completely negative for such staining (a–c of Fig. 1E), while untransfected PEFs were strongly stained by the lectin (d–f of Fig. 1E), as expected. Notably, there was no green fluorescence in these isolated clones, suggesting that transfection with pmaxGFP may have been transient (b of Fig. 1E). To check whether CRISPR/Cas9 system-mediated disruption occurs in the target gene, genomic DNA was isolated from CRISPR-2-1 to -4 and untransfected PEFs, and we performed PCR and subsequent nested PCR using the primers shown in Fig. 1A and the Materials and methods. The PCR products (approximately 140 bp) generated were inserted into the TA cloning vector pCR2.1. Sequencing analysis demonstrated that all of the clones tested had one or more nucleotide mutation (deletion or insertion) following the ATG in exon 4 of GGTA1 (upper panel in Fig. 1F). The site of mutation in the clone CRISPR-2-4 was the same found in the clone CRISPR-2-1. Another two clones exhibited different types of mutations from Table 1. Summary of in vitro acquisition of a–Gal epitope-negative colonies Experimentsa

No. of cells subjected to IB4SAP treatment

No. of colonies emerged 10 days IB4SAP after treatment

No. of a–Gal epitope- negative colonies/ no. colonies tested (%)

No. of a–Gal epitope- positive colonies/ no. colonies tested (%)

1 2 3

1 9 105 1 9 105 1 9 105

10 22 33

9/10 (90) 6/6 (100) –

1/10 (10) – –

aPEFs (1 9 106) were transfected with CRISPR/Cas9-related plasmids plus pmaxGFP. Two days later, cells were replated onto a new dish. Four days after transfection, cells (1 9 105) were subjected to IB4SAP treatment for a short period and then cultivated in normal medium. Ten days after IB4SAP treatment, the number of colonies was counted. In Experiment 1, the colonies emerged were all subjected to staining for AF594-IB4. In Experiment 2, some colonies were picked and then propagated for further characterization. Each experiment was performed at different days.

296

Simplified acquisition of a-GalT KO pig cells To confirm that the embryos in the experimental group were derived from oocytes reconstituted with CRISPR-2-1 nuclei, eight blastocysts were subjected to lysis and genomic DNA was isolated. Concomitantly, eight control embryos were treated similarly. The isolated DNA was subjected to PCR and subsequent nested PCR to amplify products of approximately 140 bp, as mentioned previously. Table 2. In vitro rates of development of SCNTa porcine embryos

Gel electrophoresis of the resulting PCR products revealed that the amplified product was almost the same size in SCNT-derived blastocysts and cultured cells for each group (Figure S3). Sequence analysis of the PCR-amplified products demonstrated that the position of the mutation following the ATG in exon 4 of GGTA1 was the same between both groups (Fig. 2B). These results indicate that these cloned blastocysts originated from nuclei of mutated cells unable to synthesize functional a-GalT protein.

Rates of development (%) to

Donor cells Untransfected PEFs CRISPR-2-1 clone

Fusion rate (%)

2-cell stage(Day 2)

Blastocyst stage (Day 7)

69 (78)

53 (77)

21 (30)

79 (78)

58 (73)

25 (32)

aAs donor cells, CRISPR-2-1 clone, and untransfected PEFs were used for SCNT. After SCNT, electrical activation of SCNT embryos was performed. Development of SCNT-treated embryos was evaluated by the rates of cleavage and blastocyst formation at 2 and 7 days of culture, respectively. Parentheses indicate numbers of embryos with successful performance on development/number of embryos tested.

Discussion

The most prominent property of CRISPR/Cas9 system-mediated gene disruption is its ability to generate a biallelic KO phenotype efficiently through a single transfection event. For example, Wang et al. [24] demonstrated that between 50 and 90% of murine fertilized eggs injected with mRNAs for Cas9 and gRNA exhibited biallelic properties when their offspring (newborn pups) were analyzed at the molecular biological level.

Fig. 2. (A) a, b, Highly reduced expression of the a-Gal epitope in blastocysts derived from oocytes reconstituted with the CRISPR2-1 clone. When developing SCNT blastocysts were stained with AF594-IB4, no fluorescence is visible on the cell surface (arrowhead). In some embryos, slight red fluorescence is visible on the ZP (arrow). c, d, Control blastocysts derived from the oocytes reconstituted with the untransfected PEFs. e, f, Mixture of blastocysts derived from oocytes reconstituted with the CRISPR-2-1 clone and control blastocysts. Note that the former blastocysts exhibited no expression of the a-Gal epitope in contrast with the highly fluorescent control blastocysts. a, c, e, Microphotographs taken under light; b, d, f, microphotographs taken under UV + light. Scale bars = 100 lm. (B) Sequence data from blastocysts reconstituted with the clone CRISPR-2-1 or with untransfected PEFs. Note the presence of a 1 base (T)-inserted mutation in the genomic DNA of the former, which is also seen in the donor cell’s DNA (Fig. 1F).

297

Sato et al. Furthermore, they demonstrated that between 65 and 81% of embryonic stem cell colonies that survived in the presence of puromycin after transfection with an expression vector carrying the Cas9 gene and puromycin resistance gene together with a gRNA expression vector were biallelic for the target gene. Such a high incidence of the generation of a biallelic KO phenotype using the CRISPR/Cas9 system was also observed using porcine cells in the present study. We found that 13.2% of PEFs transfected with 2 CRISPR/Cas9related plasmids were biallelic for GGTA1 (see Fig. 1C). The Surveyor assay also supported this point, because 7.9% (indels) of genomic DNA was cleaved by the Surveyor nuclease (see Fig. 1B). The ultimate goal of this study was to produce genetically modified (GM) cloned pigs with a biallelic GGTA1 KO phenotype. For this purpose, the acquisition of biallelic KO PEF clones lacking aGal epitope expression on their surface is a prerequisite prior to SCNT. There are several methods for the in vitro isolation of a-Gal epitope-negative cells. Among these, a magnetic beads assay appears to have been employed frequently [11,12,39] to enrich a-Gal epitope-negative cells prior to SCNT. In this study, we used targeted toxin-based selection of a-Gal epitope-negative cells [30,40] because it is simpler and much more convenient than the magnetic beads approach. It requires only a short exposure of the cells to IB4SAP and their subsequent cultivation in normal medium. In this study, of the 1.0 9 105 cells transfected with the CRISPR/Cas9 plasmids, IB4SAP treatment resulted in the generation of 10–33 colonies (with an efficiency of 0.013–0.04%; see Table 1). This efficiency appears to be lower than that estimated from FACS analysis and the Surveyor assay, as mentioned previously. This may be due to the inability of IB4SAP-treated single or few porcine cells to survive and proliferate on a plain culture dish, as suggested by us [41]. The cultivation of such cells on a feeder cell layer (comprising mitotically arrested syngenic cells) would be one of the options that could be used to increase the rate of cells with a biallelic KO phenotype. Notably, approximately 90% of colonies surviving after IB4SAP treatment were a-Gal epitopenegative. The remaining colonies were a-Gal epitope-positive, which may have been derived from cell(s) occasionally escaping from IB4SAP binding. In our experience, such a-Gal epitopepositive cells can be eliminated by treating them with IB4SAP again [40]. The direct microinjection of CRISPR/Cas9based plasmids or mRNA into a single cell embryo allows successful genetic modification in mice [24]. 298

This approach appears to be superior to the SCNT-based production of GM pigs, because the former does not require the transfection of cells, selection of biallelic KO cells, and subsequent micromanipulator-aided SCNT. Notably, Carson et al. [42] reported that the cytoplasmic injection of TALEN mRNA into porcine zygotes was capable of inducing gene KO (including monoallelic and biallelic KO) in up to approximately 30% of embryos analyzed, suggesting that the injection of CRISPR/Cas9-related mRNA into porcine zygotes may be suitable for large animals. Recently, Fu et al. [43] pointed out that CRISPR/Cas9-mediated disruption of a target gene frequently causes mutations in unrelated loci in human cells, which is called “off-target mutagenesis.” Although we attempted to find similar sequences in the area surrounding the ATG in exon 4 of GGTA1 using the Sus scrofa genome database, there is no sequence that is closely matched to the target gene (data not shown). The detection of possible off site-targeted cleavage still remains our future subject. Interestingly, more recently, high levels of genome editing specificity can be obtained upon employing double nicking using RNA-guided CRISPR/Cas9 [44], which may enable us to perform CRISPR/Cas9-mediated disruption of a target gene in a more specific manner. In conclusion, this study is the first to show that the combined use of the CRISPR/Cas9 system with targeted toxin-based selection of a-Gal epitope-negative cells can accelerate the creation of cloned porcine blastocysts with biallelic disruption of GGTA1. It took us only 2 weeks to obtain biallelic KO cells for GGTA1. The ease in using CRISPR/Cas9-mediated genetic modification and its potential to disrupt multiple loci [24] may pave the way to create GM pigs for biomedical and agricultural applications. Acknowledgment

This study was partly supported by a grant from The Ministry of Education, Science, Sports, and Culture, Japan. Author contributions

M. S. contributed to research design, acquisition of data, and writing. K. M., Y. N. and Y. N. involved in production of cloned blastocysts. M. O. involved in PCR-based amplification of mutated genes and sequencing. S. N. contributed to data analysis. T. S. performed Surveyor assay. S. W. involved in construction of vectors, FACS analysis and isolation of a-Gal epitope-negative clones.

Simplified acquisition of a-GalT KO pig cells Disclosure

The authors hereby declare no conflict of interest in this manuscript. References 1. PRATHER RS, SHEN M, DAI Y. Genetically modified pigs for medicine and agriculture. Biotechnol Genet Eng Rev 2008; 25: 245–265. 2. LAI L, KOLBER-SIMONDS D, PARK KW et al. Production of alpha-1,3-galactosyltransferase knockout pigs by nuclear transfer cloning. Science 2002; 295: 1089–1092. 3. DAI Y, VAUGHT TD, BOONE J et al. Targeted disruption of the alpha1,3-galactosyltransferase gene in cloned pigs. Nat Biotechnol 2002; 20: 251–255. 4. RAMSOONDAR JJ, MACHY Z, COSTA C et al. Production of alpha 1,3-galactosyltransferase-knockout cloned pigs expressing human alpha 1,2-fucosylosyltransferase. Biol Reprod 2003; 69: 437–445. 5. ROGERS CS, HAO Y, ROKHLINA T et al. Production of CFTR-null and CFTR-DeltaF508 heterozygous pigs by adeno-associated virus-mediated gene targeting and somatic cell nuclear transfer. J Clin Invest 2008; 118: 1571–1577. 6. ROGERS CS, STOLTZ DA, MEYERHOLZ DK et al. Disruption of the CFTR gene produces a model of cystic fibrosis in newborn pigs. Science 2008; 321: 1837–1841. 7. SUZUKI S, IWAMOTO M, SAITO Y et al. Il2rg gene-targeted severe combined immunodeficiency pigs. Cell Stem Cell 2012; 10: 753–758. 8. JOUNG JK, SANDER JD. TALENs: a widely applicable technology for targeted genome editing. Nat Rev Mol Cell Biol 2012; 14: 49–55. 9. YANG D, YANG H, LI W et al. Generation of PPARc mono-allelic knockout pigs via zinc-finger nucleases and nuclear transfer cloning. Cell Res 2011; 21: 979– 982. 10. WHYTE JJ, ZHAO J, WELLS KD et al. Gene targeting with zinc finger nucleases to produce cloned eGFP knockout pigs. Mol Reprod Dev 2011; 78: 2. 11. HAUSCHILD J, PETERSEN B, SANTIAGO Y et al. Efficient generation of a biallelic knockout in pigs using zinc-finger nucleases. Proc Natl Acad Sci USA 2011; 108: 12013– 12017. 12. LI P, ESTRADA JL, BURLAK C et al. Biallelic knockout of the a-1,3 galactosyltransferase gene in porcine liverderived cells using zinc finger nucleases. J Surg Res 2013; 181: e39–e45. 13. CUI X, JI D, FISHER DA et al. Targeted integration in rat and mouse embryos with zinc-finger nucleases. Nat Biotechnol 2011; 29: 64–67. 14. GEURTS AM, COST GJ, FREYVERT Y et al. Knockout rats via embryo microinjection of zinc-finger nucleases. Science 2009; 325: 433. 15. YOUNG JJ, CHERONE JM, DOYON Y et al. Efficient targeted gene disruption in the soma and germ line of the frog Xenopus tropicalis using engineered zinc-finger nucleases. Proc Natl Acad Sci USA 2011; 108: 7052–7057. 16. MENG X, NOYES MB, ZHU LJ et al. Targeted gene inactivation in zebrafish using engineered zinc-finger nucleases. Nat Biotechnol 2008; 26: 695–701. 17. DOYON Y, MCCAMMON JM, MILLER JC et al. Heritable targeted gene disruption in zebrafish using designed zincfinger nucleases. Nat Biotechnol 2008; 26: 702–708.

18. KLUG A. The discovery of zinc fingers and their applications in gene regulation and genome manipulation. Annu Rev Biochem 2010; 79: 213–231. 19. WIEDENHEFT B, STERNBERG SH, DOUDNA JA. RNA-guided genetic silencing systems in bacteria and archaea. Nature 2012; 482: 331–338. 20. JINEK M, CHYLINSKI K, FONFARA I et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 2012; 337: 816–821. 21. DICARLO JE, NORVILLE JE, MALI P et al. Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems. Nucleic Acids Res 2013; 41: 4336–4343. 22. GRATZ SJ, CUMMINGS AM, NGUYEN JN et al. Genome engineering of Drosophila with the CRISPR RNA-guided Cas9 nuclease. Genetics 2013; 194: 1029–1035. 23. SHEN B, ZHANG J, WU H et al. Generation of gene-modified mice via Cas9/RNA-mediated gene targeting. Cell Res 2013; 23: 720–723. 24. WANG H, YANG H, SHIVALILA CS et al. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell 2013; 153: 910–918. 25. FUJII W, KANO K, SUGIURA K et al. Efficient generation of large-scale genome-modified mice using gRNA and CAS9 endonuclease. Nucleic Acids Res 2013;. doi:10.1093/nar/ gkt772 26. CONG L, RAN FA, COX D et al. Multiplex genome engineering using CRISPR/Cas systems. Science 2013; 339: 819–823. 27. MALI P, YANG L, ESVELT KM et al. RNA-guided human genome engineering via Cas9. Science 2013; 339: 823–826. 28. CHANG N, SUN C, GAO L et al. Genome editing with RNA-guided Cas9 nuclease in zebrafish embryos. Cell Res 2013; 23: 465–472. 29. COOPER DKC, GOLLACKNER B, SACHS DH. Will the pig solve the transplantation backlog? Annu Rev Med 2002; 53: 133–147. 30. AKASAKA E, WATANABE S, HIMAKI T et al. Enrichment of xenograft-competent genetically modified pig cells using a targeted toxin, isolectin BS-I-B4 conjugate. Xenotransplantation 2010; 17: 81–89. 31. VAUGHAN HA, LOVELAND BE, SANDRIN MS. Gal(a1-3)Gal is the major xenoepitope expressed on pig endothelial cells recognized by naturally occurring cytotoxic human antibodies. Transplantation 1994; 58: 879–882. 32. NAKAYAMA A, SATO M, SHINOHARA M et al. Efficient transfection of primarily cultured porcine embryonic fibroblasts using the Amaxa nucleofection systemTM. Cloning Stem Cells 2007; 9: 523–534. 33. BLIN N, STAFFORD DW. A general method for isolation of high molecular weight DNA from eukaryotes. Nucleic Acids Res 1976; 3: 2303–2308. 34. SATO M, IWASE R, KASAI K et al. Direct injection of foreign DNA into mouse testis as a possible alternative of sperm-mediated gene transfer. Anim Biotechnol 1994; 5: 19–31. 35. SATO K, YOSHIDA M, MIYOSHI K. Utility of ultrasound stimulation for activation of pig oocytes matured in vitro. Mol Reprod Dev 2005; 72: 396–403. 36. MIYOSHI K, FUJIMOTO Y, MORI H et al. Activation and parthenogenetic development of pig oocytes exposed to ultrasound in media containing different concentrations of Ca2+. J Reprod Dev 2008; 54: 42–45. 37. MIYOSHI K, SATO K, YOSHIDA M. In vitro development of cloned embryos derived from miniature pig somatic cells

299

Sato et al.

38.

39.

40.

41.

42.

43.

44.

300

after activation by ultrasound stimulation. Cloning Stem Cells 2006; 8: 159–165. CHI H, SATO M, YOSHIDA M et al. Expression analysis of a a-1, 3-galactosyltransferase, an enzyme that creates xenotransplantation-related a-Gal epitope, in pig preimplantation embryos. Anim Sci J 2012; 83: 88–93. FUJIMURA T, TAKAHAGI Y, SHIGEHISA T et al. Production of a1,3-galactosyltransferase gene-deficient pigs by somatic cell nuclear transfer: a novel selection method for gala1,3-Gal antigen-deficient cells. Mol Reprod Dev 2008; 75: 1372–1378. SATO M, AKASAKA E, SAITOH I et al. Targeted toxinbased selectable drug-free enrichment of mammalian cells with high transgene expression. Biology 2013; 2: 341– 355. SATO M, YOSHIDA M, MIYOSHI K et al. Cultivation with untransfected fibroblasts can stimulate proliferation of a single gene-modified fibroblast derived from a Clawn miniature swine fetus. Reprod Dom Anim 2011; 46: 911–916. CARLSON DF, TAN W, LILLICO SG et al. Efficient TALENmediated gene knockout in livestock. Proc Natl Acad Sci USA 2012; 109: 17382–17387. FU Y, FODEN JA, KHAYTER C et al. High-frequency offtarget mutagenesis induced by CRISPR-Cas nucleases in human cells. Nat Biotechnol 2013; 31: 822–826. RAN FA, HSU PD, LIN C-Y et al. Double nicking by RNA-guided CRISPR Cas9 for enhanced genome editing specificity. Cell 2013; 154: 1380–1389.

Supporting Information

Additional Supporting Information may be found in the online version of this article: Figure S1. Schematic representation of the CRISPR/Cas9-related vector hCas9 and the guide RNA expression vector. CMV, cytomegalovirus; NLS, nuclear location signal; p(A), poly(A) sites. Figure S2. Inspection using a fluorescence microscope demonstrated the presence of an a-Gal epitope-negative cell (arrow) when the cells were subjected to staining with AF594-IB4 at 2 days after transfection. Note that a cell expressing pmaxGFP-derived green fluorescence is also a-Gal epitope-negative (arrow in b), suggesting biallelic KO for GGTA1. a, Photograph taken under light; b, c, photographs taken under light + UV. Scale bars = 70 lm. Figure S3. Gel electrophoresis of the nested PCR products derived from blastocysts reconstituted with the CRISPR-2-1 clone (lane 1) and those with untransfected PEFs (lane 2). The nested PCR products derived from the CRISPR-2-1 clone (lane 3) and untransfected PEFs (lane 4) are also shown. M, 100-bp ladder marker.