Here we present a protocol for developing genetically modified mouse models using embryonic stem cells, especially for large DNA knock-in (KI). This protocol is tuned up using CRISPR/Cas9 genome editing, resulting in significantly improved KI efficiency compared with the conventional homologous recombination-mediated linearized DNA targeting method.
The CRISPR/Cas9 system has made it possible to develop genetically modified mice by direct genome editing using fertilized zygotes. However, although the efficiency in developing gene-knockout mice by inducing small indel mutation would be sufficient enough, the efficiency of embryo genome editing for making large-size DNA knock-in (KI) is still low. Therefore, in contrast to the direct KI method in embryos, gene targeting using embryonic stem cells (ESCs) followed by embryo injection to develop chimera mice still has several advantages (e.g., high throughput targeting in vitro, multi-allele manipulation, and Cre and flox gene manipulation can be carried out in a short period). In addition, strains with difficult-to-handle embryos in vitro, such as BALB/c, can also be used for ESC targeting. This protocol describes the optimized method for large-size DNA (several kb) KI in ESCs by applying CRISPR/Cas9-mediated genome editing followed by chimera mice production to develop gene-manipulated mouse models.
Producing genetically modified mice and analyzing their phenotype enables us to understand specific gene functions in detail, in vivo. Numerous important findings have been uncovered using gene-modified animal models in the life science field. Furthermore, since the report of genome editing technology using CRISPR/Cas91, research using genetically modified mice has quickly spread to many laboratories2,3. Genome editing of mouse zygotes by CRISPR/Cas9 has achieved acceptable efficiency for developing short DNA modification, such as indel mutation-oriented gene knockout4, single nucleotide replacement, or short peptide-tag insertion using single-stranded oligonucleotides (ssODNs) as knock-in (KI) donors5. On the other hand, the KI of large DNA fragments into zygotes by genome editing remains at a low efficiency compared to the short-size DNA modification6,7. In addition, it is difficult to use mouse strains such as BALB/c, which is an important strain for specific research areas like immunology, for zygote-based genome editing because their preimplantation embryos are susceptible to in vitro manipulation.
Another way to develop genetically modified mouse models is to use the embryonic stem cell (ESC) targeting technique followed by ESC injection into the preimplantation embryo to produce chimeras8,9,10, which is still routinely used as a conventional method. Although the acquisition rate for obtaining accurate KI-ESC clones is not very high in conventional ESC targeting methods, ESC targeting offers some advantages compared to zygote genome editing, especially for long DNA KI. For example, the KI efficiency of long DNA fragments (> several kb) into the zygote genome is less evident6,7, and many zygotes are needed to develop even one line of KI mouse, which is undesirable in the current perspective of animal experiments. In contrast to zygote genome editing, long DNA targeting to ESCs followed by chimera production needs significantly fewer embryos than zygote genome editing. Furthermore, even though the preimplantation embryos from BALB/c are susceptible to in vitro manipulation, their ESCs can be maintained and handled in vitro11 as other competent 129 or F1 background ESCs, therefore, applicable for chimera productions. However, even though a targeting vector contains 5' and 3' homologous arms and drug resistance gene cassettes for positive or negative selection, the conventional KI efficiency of ESCs is generally insufficient, because of the high frequency of random genomic integration8,10, Thus, an improved method with precise ESC targeting efficiency is required. Recently, we reported a tuned-up ESC KI method using CRISPR/Cas9-based genome editing to achieve higher KI efficiency than conventional targeting methods11. The method we describe here is based on this procedure which enables long DNA (> several to 10 kb) KI to ESCs with acceptable efficiency for routine works without drug selection; thus, the vector construction procedure would be much easier and need a shorter period, or the cell culture period would also become significantly shorter.
All mouse experiments were approved by the Institutional Animal Care and Use Committee of the University of Tokyo (approval number PA17-63) and Osaka University (approval number Biken-AP-H30-01) and performed according to their guidelines as well as the ARRIVE guidelines (https://arriveguidelines.org).
1. Targeting vector construction
2. Preparation of the mouse embryonic fibroblast (MEF) as feeder cells for ESC
3. Cas9-RNP-DNA mixture preparation
4. Gene targeting of ESCs
5. PCR genotyping of targeted ESCs
6. Preparation of eight-cell stage embryo and microinjection of ESCs
We have targeted specific gene(s) in the ESC followed by chimera production to develop gene-manipulated mouse production according to our previous manuscript11. ESC genotyping (described in section 4) is routinely carried out by PCR using primers. The primers are designed on genomic sequences outside the homology arms and the specific sequences in the KI DNA fragment (Figure 2A). In that case, no wild-type allele is amplified, whereas a PCR amplicon of a specific size is detected only when the targeted exogenous DNA is KI at the target locus. Representative genotyping PCR results are shown in Figure 2B. Nine out of 22 clones (40.9%) showed a KI-specific band in this case. Three representative targeting results, including the result shown in Figure 2, are shown in Table 1. These results indicate that the method shown here is efficient and reproducible for gene KI without any drug selections.
ESC is picked up in an injection pipette, then a hole in the zona pellucida is made using a piezoelectric plus, and the ESC is released between the embryonic cells (Figure 3). Technically, this protocol to inject an ESC into an eight-cell or morula stage embryo is similar to the protocol for ESC injection into a blastocyst commonly used in many mouse facilities. Representative chimeric mice are shown in Figure 4. For coat color evaluation of chimerism, the embryo from the ICR strain (albino, white hair) was used as a recipient of B6 or B6-129 F1 ESC, and B6 embryos were used as a recipient of BALB/c ESCs (Figure 4).
Figure 1: A schematic of CRISPR/Cas9 ribonucleoprotein (RNP)-mediated circular plasmid integration into the specific locus of an ESC genome. Electroporation introduces a circular plasmid as a targeting vector into ESCs with Cas9-RNP. Abbreviations: GOI = gene of interest, HA = homology arm. Please click here to view a larger version of this figure.
Figure 2: Genomic PCR analyses for KI screening of targeted ESC clones. (A) KI-specific PCR primers are designed on genomic sequences outside the homology arms (forward) and in the specific sequences in the KI DNA fragment (reverse). (B) Representative genotyping PCR results. A wild-type genome was used as a negative control. Abbreviations: GOI = gene of interest, HA = homology arm. Please click here to view a larger version of this figure.
Figure 3: Representative images of eight-cell stage embryo injection. (A) For microinjection, three doublets of ESCs (six cells, arrowheads) are picked up. (B) A hole in the zona pellucida is made with a piezoelectric pulse, and ESCs are expelled in-between each blastomere. Scale bar shown is 50 µm. Please click here to view a larger version of this figure.
Figure 4: Representative images of chimera mice. (A,B) The ESC derived from C57BL/6N (JM8.A3, Agouti hair; A) or B6-129 F1 (Agouti hair; B) are injected into ICR (albino, white hair) embryos, followed by transferring the embryos to the mother surrogates. (C) The ESC derived from BALB/c (albino, white hair) is injected into C57BL/6J (black hair) embryos, followed by transferring the embryos to mother surrogates. Please click here to view a larger version of this figure.
Project ID | Choromosome | 5'HA length (bp) | 3'HA length (bp) | Insert size (bp) | Number of analyzed clones | Number of KI clones | Efficiency (%) | Remarks |
R26-CC* | Chr6 | 965 | 1006 | 5321 | 23 | 2 | 8.7% | – |
R4-03* | Chr8 | 1000 | 997 | 3070 | 22 | 6 | 27.3% | – |
P4-01* | Chr15 | 1000 | 1000 | 2569 | 22 | 9 | 40.9% | Shown in Figure 2 |
Table 1: KI efficiencies in three independent genomic loci in the ESC. *These projects have not been published so far. The name of these genes will be disclosed in independent manuscripts in the future.
Gene targeting of ESCs followed by chimera production has been conventionally used for developing gene-manipulated mice. Nevertheless, gene knock-in efficiency remains low even though the targeting vector contains long (> several kb in usual) homology arms with positive or negative drug selection gene cassettes. Our protocol introduced a tuned-up ESC KI method of long exogenous DNA using a circular plasmid without any drug selection cassettes as a targeting vector accompanied by Cas9-RNP-mediated genome editing at an acceptable efficiency for routine works. Thus, this protocol could help to substantially reduce the amount of time taken in producing genetically modified chimeric mice compared with the conventional ESC targeting.
In this protocol, we used CRISPR/Cas9-RNP to induce site-specific double-strand breaks in the genome, and a circular plasmid was used as the targeting vector instead of a linearized one. Linearized plasmids are conventionally used as a targeting vector for gene KI because of their increased efficiency of genomic integration8,9,10. However, many genomic integrations are nonspecific even though the vector contains homologous arms9. On the other hand, circular plasmids are rarely used as targeting vectors because of their hard-to-integrate feature into the genome of ESCs12 or fibroblasts13. Thus, it would be possible that applying a circular plasmid as a targeting vector accompanied by CRISPR/Cas9-mediated genome editing minimizes nonspecific random integration but maximizes the site-specific integration into the ESC genome. It would be notable that the induction efficiency of the double-strand break by CRISPR/Cas9 is quite important14, thus it would be essential to use gRNA which induces a double-strand break at high efficiency. It should be considered to re-design gRNA when the KI efficiency is too low. In the case shown in Figure 2, 40.9% of ESC clones showed a KI-specific band. In fact, as the core lab of the institute, we routinely perform gene targeting using ESCs by the method described here and have achieved KI efficiencies of 10%-50% for 1-2 kb sequences such as Cre, CreERT, or fluorescent reporters, although there are some variations depending on the gene locus. The longest DNA sequences we have successfully KI using this method is about 11.2 kb into the Rosa26 locus, and the efficiency was 12.2% (five KI colonies out of 41 analyzed colonies).
It would also be of note that this protocol uses eight-cell or morula stage embryos as recipients for ESC microinjection but not blastocyst stage embryos. Although we have not conducted comparing experiments in this paper, several reports have shown that injection of ESCs into eight-cell or morula stage embryos significantly improves the efficiency of ESC contribution to chimeric offspring compared to the ESC injection into blastocysts15,16,17,18. Indeed, ESC injections of various ESC lines, including C57BL/6, B6-129 F1, and BALB/c, commonly resulted in high coat color chimera development in some, but not all, offspring (see Figure 4). The limitation of the method is that the ESCs injected into the preimplantation embryo could not always contribute to germ cells in a chimera19,20. Germline transmission of ESCs would depend on the quality of each ESC clone19. Therefore, it would be advantageous to develop multiple ESC clone lines as a backup for stable chimeric mouse production. In conclusion, the protocols presented here use simple targeting vectors which include only each homology arm and gene of interest for KI without any drug-resistant cassette. Therefore, the vector construction would be much easier, and the culture period could be shortened compared to the conventional ESC targeting technique. This would help in quickly and facilitatively producing various types of genetically modified mice for future analysis in life science.
The authors have nothing to disclose.
We thank Saki Nishioka in Osaka University, Biotechnology Research and Development (nonprofit organization), and Mio Kikuchi and Reiko Sakamoto in the Institute of Medical Science, The University of Tokyo, for their excellent technical assistance. This work was supported by: the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/Japan Society for the Promotion of Science (JSPS) KAKENHI grants to MI (JP19H05750, JP21H05033), and MO (20H03162); the Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST) grant to MI (JPMJCR21N1 ); the Eunice Kennedy Shriver National Institute of Child Health and Human Development to MI (R01HD088412); the Bill & Melinda Gates Foundation to MI (Grand Challenges Explorations grant INV-001902); and the Grant for Joint Research Project of the Research Institute for Microbial Diseases, Osaka University to MI, and MO.
BALB/c ESC | – | – | ESC developed from BALB/c strain |
Bambanker | Nippon Genetics | CS-02-001 | Cell-freezeing medium. Section 2.6 and elsewhere |
Cas9 Nuclease V3 | IDT | 1081059 | Section 3.2 and elsewhere. |
CHIR99021 | FUJIFILM Wako | 038-23101 | Section 4.3 |
CreERT gene fragment | GeneWiz | Section 1.1. | |
CRISPR-Cas9 crisprRNA | IDT | crisprRNA. Section 3.1 and elsewhere. | |
CRISPR-Cas9 tracrRNA | IDT | 1072534 | tracrRNA. Section 3.1 and elsewhere. |
DMEM | Nacalai | 08458-45 | MEF medium. Section 2.3 and elsewhere |
Duplex buffer | IDT | 1072534 | RNA dilution buffer. Section 3.1 and elsewhere. |
FastGene Gel/PCR Extraction Kit | Nippon Genetics | FG-91302 | Section 1.1 and 1.2. |
GlutaMax | Thermo Fisher | 35-050-061 | L-glutatime substrate |
hCG | ASKA Animal Health | Section 6.1. | |
In-Fusion HD Cloning Kit | Clontech | 639648 | DNA cloning kit. Section 1.3 and elsewhere |
JM8.A3 ESC | EuMMCR | – | ESC developed from C57BL/6N strain |
Knock-out DMEM | Thermo Fisher | 10829018 | Section 4.3 and elsewhere, DMEM-based modified commercial medium. |
KSOM | Merck | MR-121-D | Section 6.3 and 6.9. |
Leukemia inhibitory factor | FUJIFILM Wako | 125-05603 | Section 4.3. No unit concentration data is supplied by the provider. Used 1,000-fold dilution in this protocol. |
Neon Electroporation system | Thermo Fisher | MPK5000 | Section 3.2, 4.5 and elsewhere. The system containes electroporation buffer as well used in section 3.2. |
NucleoSpin Plasmid Transfection-grade | Takara | U0490B | Section 1.6. |
PD0325901 | FUJIFILM Wako | 162-25291 | Section 4.3 |
PMSG | ASKA Animal Health | Section 6.1. | |
Tail lysis buffer | Nacalai | 06169-95 | Section 5.5. |
Trypsin-EDTA | Nacalai | 32777-15 | Section 2.2 and elsewhere |
V6.5 ESC | – | – | ESC developed from B6J-129 F1 strain |
X-ray irradiation device | Hitachi | MBR-1618R-BE | Section 2.6. |