Genome editing with the donor plasmid equipped with synthetic crRNA-target sequence

CRISPR/Cas-mediated genome editing is a powerful tool for generating genetically mutated cells and organisms. Linearisation of donor cassettes with this system has been shown to facilitate both transgene donor insertion and targeted knock-in. Here, we developed a donor plasmid that we name pCriMGET (plasmid of synthetic CRISPR coded RNA target sequence-equipped donor plasmid-mediated gene targeting), in which an off-target free synthetic CRISPR coded RNA-target sequence (syn-crRNA-TS) is incorporated with a multi-cloning site, where a donor cassette can be inserted. With co-expression of Cas9 and the syn-crRNA-TS guide RNA (gRNA), pCriMGET provides a linearised donor cassette in vivo, thereby promoting the transgene donor insertion and targeted knock-in. When co-injected with Cas9 protein and gRNA into murine zygotes, pCriMGET yielded around 20% transgene insertion in embryos. This method also achieved more than 25% in-frame knock-in at the mouse Tbx3 gene locus without predicted insertion–deletion mutations using a transgene donor with 400-bp homology arms. pCriMGET is therefore useful as a versatile CRISPR/Cas9-cleavable donor plasmid for efficient integration and targeted knock-in of exogenous DNA in mice.


Intracellular cleavage of pCriMGET by CRISPR/Cas9. To test whether CRISPR/Cas9 induces DSBs
on pCriMGET at the syn-crRNA-TS in vivo, we transfected HEK293T cells with pCriMGET incorporated with the donor cassette encoding the EF1α promoter-hygromycin resistant gene-T2A-EGFP-bovine growth hormone polyadenylation signal sequence (bGHpA) (pCriMGET-EF1α-hygro-T2A-EGFP-pA), together with pX330 22 , a vector that expresses both humanised Staphylococcus pyogenes (sp) Cas9 protein and syn-crRNA-TS targeting sgRNA (pX330-syn-crRNA-TS-sgRNA) ( Fig. 2A). After analysing the transfection efficiency by EGFP expression using flow cytometry at 24 h, cells were selected using hygromycin for 14 days. Because linearisation of the donor plasmid promotes integration of the transgene into the host genome, pX330-syn-crRNA-TS-sgRNA is expected to increase the genomic integration efficiency of the transgene donor, thereby promoting hygromycinresistant colony formation. We found that the efficiency of colony formation was significantly increased when pCriMGET-EF1α-hygro-T2A-EGFP-pA was co-transfected with pX330-syn-crRNA-TS-sgRNA, compared with pX330 vector co-transfection (Fig. 2B,C). In addition, we confirmed genomic integration of the donor cassette (EF1α-hygro-T2A-EGFP-pA) in cells (Fig. 2D). These results indicate that CRISPR/Cas9 induces DSBs on pCriMGET at the syn-crRNA-TS in vivo.  www.nature.com/scientificreports/ Precise in-frame knock-in of exogenous DNA by pCriMGET in culture cells. We next examined whether pCriMGET can be applied for in-frame knock-in of exogenous DNA within culture cells. To this end, we established an mCherry/EGFP-homology-directed repair (HDR) reporter system, which monitors the in-frame knock-in of transgene donors into target genomic loci by mCherry/EGFP conversion (Fig. 3A), in accordance with the mCherry-HDR reporter system described previously 18 . First, we established a reporter HEK293T single-cell clone in which hygro-T2A-EGFP was genomically integrated (Fig. S2). We then constructed pCriMGET-resT2A-mCherry-stop, which incorporates the donor cassette encoding sgRNA resistant T2A-mCherry-3x stop codons, flanked by homology arms (Fig. 3A). To optimise the length of the homology arms, the donor cassette was flanked by homology arms of various lengths (Fig. 3B). The reporter HEK293T cells were transfected with pCriMGET-resT2A-mCherry-stop, together with pX330-syn-crRNA-TS-sgRNA and pX330-T2A-sgRNA, which are expected to induce DSBs on pCriMGET at syn-crRNA-TS and on the genomically integrated T2A site, respectively (Fig. 3A). The cleavage at the T2A site was confirmed by the emergence of an EGFP − cell population transfected with pX330-T2A-sgRNA alone, compared with mock-transfected cells ( Fig. 3B; no donor, indicated by a dashed line). When the cells were transfected with all three vectors, the mCherry + EGFP − cell population was increased in a manner dependent on the length of the homology arms, with a plateau at 400 bp, but to a lesser extent in the absence of pX330-syn-crRNA-TS-sgRNA (Fig. 3B,C). The mCherry + EGFP + cell population was increased in the transfected cells, which is presumably attributable to the multiple copy number of hygro-T2A-EGFP in the reporter HEK293T clone used in the assay and random integrations of the donor gene. Next, we sequenced the T2A site genomic region in transfected cells and confirmed knock-in of the donor gene (mCherry) with no indels or frame-shift in the 5′ and 3′ junction regions (Fig. 3D). These findings indicate that pCriMGET induced in-frame knock-in of the donor cassette with 400-bp homology arms. We further examined whether the pCriMGET/pX330 system could be applied for transgene knock-in into the endogenous AAVS1 site, a well-known safe harbour within the human PPP1R12C gene locus. To this end, we constructed pCriMGET-SA-neo-pA, which incorporates the donor cassette encoding AAVS1-sgRNA-targeting site-harbouring splicing acceptor (SA)-T2A-neomycin resistance gene (neo)-bGHpA, flanked by 400-bp homology arms (Fig. 4A). Then, we transfected HeLa cells with pCriMGET-SA-neo-pA, together with pX330-syn-crRNA-TS-sgRNA and pX330-AAVS1-sgRNA 23 , followed by G418 selection for 14 days. The colony formation efficiency was significantly increased when cells were transfected with all three vectors compared with that in the absence of pX330-AAVS1-sgRNA, pX330-syn-crRNA-TS-sgRNA or both ( Fig. 4B-D). These results indicate that the pCriMGET/pX330 system can induce in-frame exogenous transgene knock-in within culture cells via CRISPR/ Cas9-mediated cleavage of syn-crRNA-TS.
Efficient generation of transgenic mice with pCriMGET. We further attempted to generate transgenic mice with pCriMGET. We constructed pCriMGET that incorporates the donor cassette encoding CAG promoter-EGFP-pA (pCriMGET-pCAG-EGFP-pA) and microinjected it into the pronuclei of pronuclear-stage mouse zygotes together with syn-crRNA-TS-crRNA, tracrRNA and Cas9 protein (Fig. 5A). After transplantation into pseudopregnant mice, E17.5 embryos were collected and subjected to genotyping PCR (Fig. 5B, Fig. S3A). In addition, by using the genomes from heterozygous and homozygous H2B-EGFP mice as controls, we determined the copy number of the integrated donor cassette in each embryo (Fig. S3B). We defined a transgenic fetus as one that carries more than one copy of the integrated donor transgene (EGFP). The results showed that 20.3% (13/64) of the embryos were transgenic (Fig. 5C). We also confirmed EGFP protein expression in the transgenic embryos (Fig. 5D). Notably, the integration efficiency was much lower in the absence of syn-crRNA-TS-crRNA ( Fig. S4A-C), indicating that linearisation of the donor cassette with syn-crRNA-TS-crRNA enhances transgene integration. These results demonstrate that pCriMGET is useful for the efficient generation of transgenic mice.
One-step generation of in-frame knock-in mice with pCriMGET. Given the highly efficient integration of a transgene donor in mice, we further examined whether pCriMGET is also useful for the in-frame knock-in of exogenous DNA in mice. We designed a strategy for the in-frame integration of a donor gene encoding 3×Flag-P2A-EGFP into the 3′-end of the Tbx3-coding sequence on exon 8 (Fig. 6A). We constructed pCriMGET-3×Flag-P2A-EGFP, which incorporates the donor gene encoding 3×Flag-P2A-EGFP flanked by 400-bp homology arms, and microinjected it into the pronuclei of pronuclear-stage mouse embryos together with syn-crRNA-TS-crRNA, Tbx3-crRNA 24 , tracrRNA and Cas9 protein (Fig. 6B). After transplantation into pseudopregnant mice, E15.5 embryos were collected and subjected to genotyping PCR (Fig. 6C). We also determined the copy number of the transgene in each donor gene-integrated embryo (Fig. S5). The results showed that, out of a total of 19 embryos, 8 embryos (42.1%) were transgenic and 5 (26.3%) were knock-in with more than one copy of the transgene (Fig. 6D). We detected no indels or frame-shift in the 5′ and 3′ junction regions of all knock-in embryos (Fig. 6E). We confirmed the expression of Tbx3-3×Flag protein and EGFP protein in the E15.5 knock-in embryo (No. 19) by western blotting and immunohistochemistry, respectively ( Fig. 7A-C). Additionally, we also obtained 4 knock-in pups out of 11 pups using the same strategy (Fig. S7A,B). Moreover, we confirmed that the transgene was inherited by F 1 pups in a Mendelian manner (Fig. S7C,D). We also checked for off-target effects of syn-crRNA-TS-crRNA. We searched putative off-target regions in the mouse genome that matched more than 17 out of the 20 bases of syn-crRNA-TS using the free software Cas-OFFinder, and identified five candidate regions. We sequenced these genomic regions and found no indel mutations in all donor geneintegrated embryos (Table S1).
Notably, the knock-in efficiency of the pCriMGET method was comparable to that of the Tild-CRISPR method, in which in vitro cleaved linearised donor cassette was used as a repair template (Fig. S6A,B,D). In addition, the knock-in efficiency was much lower in the absence of syn-crRNA-TS-crRNA ( Fig. S6A,C,D), indicating that linearisation of the pCriMGET donor cassette with syn-crRNA-TS-crRNA enhances knock-in efficiency.    www.nature.com/scientificreports/ Finally, to assess the versatility of the pCriMGET method for in-frame knock-in, we designed a strategy for the in-frame integration of a donor gene encoding IRES-hDTR-2A-EGFP into the Clec4f 3′UTR in the exon 7 genomic locus (Fig. S8A). We constructed pCriMGET-IRES-hDTR-2A-EGFP with 500-bp homology arms and microinjected it into the pronuclei of pronuclear-stage mouse embryos together with syn-crRNA-TS-crRNA, Clec4f-crRNA 25 , tracrRNA and Cas9 protein (Fig. S8B). After transplantation into pseudopregnant mice, 4-weekold male and female mice were collected and subjected to genotyping PCR (Fig. S8C). Out of a total of 17 pups, 2 pups (11.8%) were knock-in (Fig. S8D) with no indels or frame-shift in the 5′ and 3′ junction regions, implying   www.nature.com/scientificreports/ the versatility of the pCriMGET method in generating knock-in mice. These results taken together demonstrate that pCriMGET is useful for the precise and efficient generation of in-frame knock-in mice.

Discussion
In this study, we developed pCriMGET, a CRISPR/Cas9-cleavable donor plasmid equipped with syn-crRNA-TS of minimal off-target potential. Compared with the previously described HMEJ-based method, where crRNA on-target sequences sandwich a donor cassette 18,19 , pCriMGET is simple in construction because the donor cassette can be incorporated into MCS flanked by the syn-crRNA-TSs. Similar to the HMEJ-based and Tild-CRISPR method, the pCriMGET system achieved much higher targeted knock-in efficiency in mouse embryos than the conventional HR-based method 4,5 . Moreover, unlike the Easi-CRISPR method 14,16 , the pCriMGET system appears not to be particularly limited in terms of the transgene donor size: 3.0-3.3 kb of exogenous DNA (EF1α-hygro-T2A-EGFP-pA and Clec4f-hDTR-2A-EGFP) was successfully integrated into the genome.
The pCriMGET system accomplished the efficient generation of not only knock-in mice, but also transgenic mice. It has been reported that a linearised donor cassette is more efficiently integrated into genomic DNA than a circular plasmid 26,27 . In the pCriMGET system, the donor cassette is linearised in vivo by CRISPR/Cas9 via syn-crRNA-TS, enabling the frequency of transgene insertion into the genome to be increased. The efficiency of generation of transgenic mice by the pCriMGET system (> 20%) was comparable to that of the conventional linearised-donor cassette injection method (20-30%) 26 , highlighting its easy-to-manipulate linearised-donor gene preparation.
The pCriMGET system exhibited nontargeted donor gene integration with various copy numbers (Fig. 6D,  Fig. S5). Previous studies showed that the frequency of nontargeted donor gene integration correlated with the length of the homology arms [17][18][19] . In HMEJ-based and Tild-CRISPR methods, the optimised minimum length of homology arms was 800-900 bp. The homology arms used in this study were 400 bp in length, which would cause nontargeted donor integration. However, the pCriMGET system achieved 26.3% in-frame knock-in embryos, even using relatively short homology arms. In addition, Tbx3-3×Flag-P2A-EGFPKI mice generated by the pCriMGET system exhibited transgene germline transmission in accordance with Mendel's law, implying that backcrossing the mutant mice to the inbred control strain would eliminate the randomly integrated donor cassettes. Taking these findings together, the pCriMGET system is a simple, versatile, low-cost and efficient genome editing tool for the generation of transgenic and gene-targeted mice.

Methods
pCriMGET. The crRNA-TSs in SPA of the rabbit β-globin gene and putative off-target regions of each crRNA-TS in mouse (GRCm38/m10) and human (GRCh38/hg38) genomes were searched using the free software CRIS-PRdirect. The SPA sequence including syn-crRNA-TS was amplified by PCR using the single-stranded DNA (5′-GTT TTT TGT GTG AAT CGA TAG TAC TAA CAT ACG CTC TCC ATC AAA ACA AAA CGA AAC AAA ACA AAC TAG CAA AAT AGG CT-3′) as a template. To reduce CRISPR-Cas9 off-target effects, three nucleotide mutations (see Fig. S1) were introduced in SPA by reverse primers. KpnI/XhoI sites and BamHI/XbaI sites were introduced into the 5′-modified (m)SPA and 3′-mSPA, respectively, by PCR primers. Then, the 5′-mSPA and 3′-mSPA were inserted into KpnI-XhoI and BamHI-XbaI sites on the pBluescriptII SK(+) plasmid, respectively. We validated the insertion by sequencing.
pCriMGET-SA-neo-pA. The human genomic sequences for 400 bp upstream and downstream of the AAVS1 site within the PPP1R12C gene locus were PCR-amplified using HEK293T genomic DNA as a template and used as homology arms. The splicing acceptor (SA) sequence (5′-CTT CTG ACC TCT TCT CTT CCT CCC Scientific RepoRtS | (2020) 10:14120 | https://doi.org/10.1038/s41598-020-70804-6 www.nature.com/scientificreports/ ACAGG-3′) was fused to the 5′-end of T2A-EGFP cDNA by PCR amplification. The SA-T2A-EGFP coding sequence and the homology arms were fused and inserted into SalI-EcoRI sites of pCriMGET MCS by using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs), in accordance with the manufacturer's protocol. Then, the EGFP gene of pCriMGET-SA-T2A-EGFP was replaced by a neomycin resistance gene using NEBuilder HiFi DNA Assembly Master Mix.
pCriMGET-3×Flag-P2A-EGFP. The mouse genomic sequences for 400 bp upstream and downstream of the stop codon of the Tbx3 gene were PCR-amplified using C57BL/6JJcl mouse genomic DNA as a template, and were used as left and right homology arms, respectively. The 3×Flag-tag coding sequence was fused to the 3′-end of the left homology arm by PCR. P2A coding sequence (5′-GGC TCG GGT GCC ACG AAT TTC TCA TTA CTG AAG CAG GCT GGA GAC GTG GAG GAG AAC CCT GGA CCT -3′) was fused to the 5′-end of the EGFP gene by PCR. Then, the fragments of the left arm-3×Flag, P2A-EGFP, and the right arm were fused and inserted into the EcoRV site of pCriMGET MCS using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs).
pCriMGET-Clec4f-IRES-hDTR-P2A-EGFP. The mouse genomic sequences for 500 bp upstream and downstream of the crRNA target site of the Clec4f gene were PCR-amplified using C57BL/6JJcl mouse genomic DNA as a template, and were used as left and right homology arms, respectively. Human diphtheria toxin receptor (hDTR) carrying I117V/L148V mutations was constructed using three fragments amplified from HEK293T genomic DNA by PCR, and these three fragments were fused and inserted into the EcoRI-BamHI site of pEGFP-N3 (Clontech) using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). The internal ribosome entry site (IRES) was amplified from pQCXIN X2/pTER shLUC (w347-1) (Addgene: #17489), hDTR carrying I117V/L148V was amplified from pEGFP-N3-hDTR, and P2A-EGFP was amplified from pCriMGET-Tbx3-3xFlag-P2A-EGFP plasmid by PCR. Then, the fragments of the left arm, IRES, hDTR carrying I117V/ L148V, P2A-EGFP, and the right arm were fused and inserted into the XhoI-EcoRI site of pCriMGET MCS using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs).
Genotyping pcR. For EGFP transgenic mice, genomic DNA extracted from E17.5 embryo tails was amplified by PCR with KOD Plus Neo (TOYOBO) using primer sets designed inside of the donor cassette. PCR amplification of the genomic region of the potential syn-crRNA-TS off-target sequence on chromosome 8 (see Table S1, CriMGET-OT1) was used as an internal control. copy number analysis of transgene donor cassette. For determination of the copy number of the transgene donor cassette in EGFP transgenic mice and Tbx3-3xFlag-2A-EGFP knock-in mice, real-time PCR was performed in the E17.5 and E15.5 embryonic tail genomes. qPCR was conducted in duplicate with 20 ng of DNA in a 20 µL reaction mixture using the THUNDERBIRD SYBR qPCR Mix (TOYOBO) and the Applied Biosystems 7500 Real-Time system (Applied Biosystems). R26R-H2B-EGFP heterozygous and homozygous knockin mouse tail genomes were used as references for copy number. DNA amount was calibrated using the PCR product of the CriMGET-OT1 region on chromosome 8 as an internal control. Copy number was defined as the value relative to the R26R-H2B-EGFP heterozygous knock-in sample.
Off-target analysis. A search for potential off-target sites in the mouse genome (GRCm38/mm10) was performed using Cas-OFFinder (https ://www.rgeno me.net/cas-offin der). Genomic regions that matched more than 17 out of 20 bases of the syn-crRNA-TS were PCR-amplified and sequenced to evaluate the off-target effects. The primers used in the analysis are listed in Table S3.
Scientific RepoRtS | (2020) 10:14120 | https://doi.org/10.1038/s41598-020-70804-6 www.nature.com/scientificreports/ Animals. C57BL/6JJmsSLC and Jcl:ICR mice were obtained from Japan SLC Inc. and CLEA Japan Inc., respectively. All experiments were performed in accordance with Kyoto University's Regulations on Animal Experimentation. All animal experiments in this study were approved by the Committee for Animal Experiments of the Institute for Frontier Life and Medical Sciences, Kyoto University.