Heritable gene targeting in the mouse and rat using a CRISPR-Cas system

To the Editor:

CRISPR-Cas systems have been developed as an efficient gene editing technology in cells and model organisms. Here we use a CRISPR-Cas system to induce genomic DNA fragment deletion in mice by co-injecting two single-guide RNAs (sgRNAs) targeting the Uhrf2 locus with Cas9 mRNA. Furthermore, we report the generation of a Mc3R and Mc4R double-gene knockout rat by means of a single microinjection. High germline-transmission efficiency was observed in both mice and rats.

The clustered, regularly interspaced, short palindromic repeats (CRISPR)-associated protein (Cas) system has evolved in bacteria and archaea as an RNA-based adaptive immune system against viral and plasmid invasion1. Based on gene conservation and locus organization, three major types of CRISPR systems have been identified2,3. In the type II systems, the complex of a CRISPR RNA (crRNA) annealed to a trans-activating crRNA (tracrRNA) is sufficient to guide the Cas9 endonuclease to a specific genomic sequence to generate double-strand breaks in target DNA4. Previous studies established a strategy for multiplex genome engineering with the Cas9 RNA-guided endonuclease in mammalian cells5,6. Recently, efficient genome editing by the CRISPR-Cas system has been shown in multiple organisms, including zebrafish, mice and bacteria7,8,9. Several groups have demonstrated that compared with zinc finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs), CRISPR-Cas–mediated gene targeting has similar or greater efficiency in cells and zebrafish5,6,7,10. Although it has been demonstrated that multiple genes can be disrupted in individual mouse embryos using CRISPR-Cas–mediated systems9, germline transmission of Cas9-mediated mutations in animals has not yet been reported. In addition, whether long, specific, genomic DNA target fragments can be deleted by the CRISPR-Cas system is still unknown. Moreover, the utility of the CRISPR-Cas system for gene targeting in other mammalian models, for example the laboratory rat, still needs to be determined. Here, we report the generation of highly efficient, heritable, gene knockout in mice and rats by using a CRISPR-Cas system.

To test the activity of Cas9-mediated gene targeting in mice, a genomic Th site that has been previously targeted efficiently in mouse cells4,5 was selected for the initial experiments in knockout mouse generation. First, we injected different concentrations of linearized DNA (encoding humanized Cas9, target specific crRNA and tracrRNA) (Fig. 1a) into the male pronuclei of FVB strain mouse embryos and assayed the genomic mutation status of the Th locus in the resulting pups. Similar to ZFNs and TALENs, the CRISPR-Cas system induces double-strand DNA breaks that are repaired mainly by error-prone nonhomologous end joining (NHEJ). Only 1/11 (9%) of the pups generated from high-concentration DNA injection (2.5 ng/μl) was modified at the Th locus, with only wild-type pups generated by injection of a lower DNA concentration (Table 1 and Supplementary Fig. 1), suggesting that the mutation rate is low when the CRISPR-Cas system is delivered as linearized DNA.

Figure 1: Generation of gene mutant rats with a CRISPR-Cas system.

(a) Constructs of Cas9/RNA system used in this study for DNA (left) and RNA (right) injections. Spacer, nuclease guide sequence; DR, direct repeat to separate individual spacers. NLS, nuclear localization signal. (b) Detection of mutations in F0 rats generated by injection of gRNA:Cas9 targeting Mc4r before (−) or after (+) T7E1 digestion using PCR products amplified from Mc4r F0 rats tail genomic DNA. Arrowheads, mutant band. M, DNA molecular weight marker. (c) DNA sequences of Mc3r or Mc4r genomic loci in founders. Red boxes enclose nucleotide substitutions. The change in the base-pair sequence is shown at right. Six TA clones of the PCR products amplified from each founder were analyzed by DNA sequencing. The incidences of each genotype in six clones were listed at rightmost column.

Table 1 Generation of knockout mice and rats via the CRISPR-Cas system

To improve efficiency, we next injected RNAs synthesized in vitro. First, we constructed Cas9 expression vectors for in vitro transcription of Cas9 mRNA by subcloning a DNA fragment harboring the SP6 promoter sequence into a vector containing the nuclear-targeted humanized Cas9 coding sequence5. We also constructed a fusion of the crRNA and tracrRNA expression vectors that enable T7 promoter-driven production of a customizable synthetic (sgRNA) with 20 nucleotides of target-specific sequence followed by tracrRNA-derived sequences at the 3′ end (Fig. 1a). Concentrations of Cas9-encoding mRNA comparable to those used in TALEN studies11 (25 ng/μl), together with Th-targeting sgRNA (12.5 ng/μl), were microinjected into the cytoplasm of one-cell-stage C57BL/6 mouse embryos. Ninety percent (8 of 9) of the pups from RNA injection were founders bearing mutations at the Th locus, as determined by a T7 endonuclease I (T7EI) assay and DNA sequencing (Supplementary Fig. 1). The longest deletion was observed in founder 6, bearing a 70-bp deletion. Similarly, founder mice bearing insertion/deletion mutations (indels) at the Rheb genomic locus were generated with high efficiency using the same strategy (Table 1 and Supplementary Fig. 2).

One of the most important advantages of CRISPR-Cas systems is that the Cas9 protein can be guided by individual gRNAs to modify multiple genomic target loci simultaneously5. To test this in mice, we then designed and injected two sgRNAs targeting adjacent sites spanning 86 bp in the Uhrf2 locus with Cas9 mRNA into embryos to make deletions. Eleven of 12 F0 mice had mutations in the Uhrf2 locus, and 6 of these founders had a total of 7 different disruptions of both these targets on the same allele (Supplementary Fig. 3). Three of the six founders modified at both sites had large deletions (Supplementary Fig. 3). These large deletions were probably generated by simultaneous DNA cleavage of these two sites, followed by end joining ligation of the broken ends.

The mutations generated by ZFNs and TALENs in founders are transmitted efficiently to the next generation11,12,13,14,15, but to the best of our knowledge, the germline transmission efficiency of a CRISPR-Cas system has not been reported in animals. To investigate this issue, we crossed Th founders to wild-type mice and the genotypes of the pups or fetuses were determined by T7E1 digestion or DNA sequencing. Although only two mutations were identified in the tail DNA of the founder generated by DNA injection, in 6 of 10 F1 fetuses we found a total of five different mutations (Supplementary Fig. 4). These data imply that the founder was a mosaic due to the delay in DNA cleavage by Cas9 in the embryos. It also suggests that sequencing of six clones of PCR products from founder tail DNA would not reveal all the mutations generated. Another two founders generated by mRNA injection also transmitted the mutation to the next generation (Supplementary Fig. 4). These data demonstrate that a CRISPR-Cas system is a useful genetic tool to generate heritable mutant mice with very high efficiency.

The laboratory rat is important for modeling diseases and has many advantages over mouse models in toxicology and pharmacology. Previous studies have successfully generated knockout rats through both ZFNs and TALENs14,15. Here, we attempt to generate knockout rats using a CRISPR-Cas system. Two sgRNAs targeting rat melanocortin 3 receptor (Mc3r) and melanocortin 4 receptor (Mc4r) were synthesized. Cas9 mRNA and a mixture of sgRNAs were injected into one-cell-stage Sprague-Dawley rat embryos, which were then transferred to pseudopregnant females. Pup genomic DNA was extracted for PCR amplification of the target loci. A T7EI assay and DNA sequencing data show that both the Mc3r and Mc4r loci were modified by our CRISPR-Cas system (Fig. 1b,c). However, the activities of these two Cas9-based nucleases are quite different. Thirteen of 15 F0 pups were identified as the founders of Mc4r mutant rats by T7E1 digestion and subsequent sequencing (Fig. 1b). Founders containing large deletions were easily detected by PCR without digestion. No founder with a Mc3r mutation was identified by T7E1 digestion, but one rat that had a Mc4r mutation, also had a single-nucleotide deletion in the Mc3r locus as determined by sequencing (Fig. 1c). We investigated the body weight, food intake, insulin level and leptin mRNA level of Mc4r founders and found that the biallelic mutants exhibited a similar phenotype to the N-ethyl-N-nitrosourea (ENU)-induced Mc4r mutant rat that has been used as an obesity animal model16 (Supplementary Fig. 5). These data suggest that a CRISPR-Cas system can generate gene knockout rats, and that the efficiency depends on the target site. In addition, a single injection is capable of inducing disruption of at least two different genes in the rat. To determine the germline transmission capability of Cas9-mediated gene mutation in rats, we crossed Mc4r mutant rat founder 12 with wild-type rat and determined the Mc4r target sequence of fetuses. Three of six fetuses containing two different mutations were identified (Supplementary Fig. 6), suggesting high efficiency of germline transmission in rats with Cas9-mediated gene mutation.

Another important issue for genome editing is the in vivo specificity. Previous studies indicated two possible rules for how mismatches affected Cas9-mediated DNA cleavage. One is that single-base mismatches up to 12-bp 5′ of the protospacer adjacent motif (PAM) completely abolished Cas9-mediated DNA cleavage. The other is that a stretch of at least a 13-bp match between gRNA and target DNA proximal to the PAM is required for efficient cleavage, and mismatches outside this motif can be tolerated4,5. We investigated the specificity of our CRISPR-Cas system to their targets by analyzing potential off-target sites in the mouse genome in these two categories. As it is difficult to analyze all potential off-target sites, the two sites with fewer than four mismatches or with a contiguous match to PAM motif of >11 bp were selected (Supplementary Fig. 7). No mutations were observed at these potential off-target sites in the 12 founders that were analyzed by sequencing (Supplementary Fig. 7).

In cells and zebrafish, CRISPR-Cas system–mediated gene disruption has similar activity to that of TALENs5,7; however, our study, together with a previous report9, suggests that the activity of the CRISPR-Cas system is greater than that of ZFNs or TALENs, at least in mice. The mutation rate of TALEN-modified mice ranges from 13% to 67% in our previous study11, but the gRNA:Cas9-induced mutation rate is usually >70%. We also note that the toxicity (referring to the viability immediately after microinjection) of a CRISPR-Cas system (Table 1) is a little bit greater than that of TALENs11,12,14. The germline transmission rate of the CRISPR-Cas system is similar to that of TALENs. As with TALENs, we also found that some targets cannot be disrupted by CRISPR-Cas for unknown reasons (data not shown), but the potential targets are more flexible than with TALENs. Although the targets are shorter than those of TALENs and ZFNs, no off-target mutation event using the CRISPR-Cas system was found in our study, suggesting it is a reliable technique for gene editing in animals. More comprehensive studies, such as the one recently published by Fu et al.17, will be needed to establish the relative advantages and disadvantages of the various genome editing systems.

In this study, we successfully generated specific gene knockout mice of distinct genetic backgrounds as well as a gene-specific knockout in Sprague-Dawley rats using a CRISPR-Cas system. Featuring highly efficient genomic modification activity and germline transmission, the RNA-guided CRISPR-Cas system is a potentially useful genetic tool for functional genomic research in mammalian organisms.


  1. 1

    Garneau, J.E. et al. Nature 468, 67–71 (2010).

    CAS  Article  Google Scholar 

  2. 2

    Makarova, K.S. et al. Nat. Rev. Microbiol. 9, 467–477 (2011).

    CAS  Article  Google Scholar 

  3. 3

    Wiedenheft, B., Sternberg, S.H. & Doudna, J.A. Nature 482, 331–338 (2012).

    CAS  Article  Google Scholar 

  4. 4

    Jinek, M. et al. Science 337, 816–821 (2012).

    CAS  Article  Google Scholar 

  5. 5

    Cong, L. et al. Science 339, 819–823 (2013).

    CAS  Article  Google Scholar 

  6. 6

    Mali, P. et al. Science 339, 823–826 (2013).

    CAS  Article  Google Scholar 

  7. 7

    Hwang, W.Y. et al. Nat. Biotechnol. 31, 227–229 (2013).

    CAS  Article  Google Scholar 

  8. 8

    Jiang, W., Bikard, D., Cox, D., Zhang, F. & Marraffini, L.A. Nat. Biotechnol. 31, 233–239 (2013).

    CAS  Article  Google Scholar 

  9. 9

    Wang, H. et al. Cell 153, 910–918 (2013).

    CAS  Article  Google Scholar 

  10. 10

    Cho, S.W., Kim, S., Kim, J.M. & Kim, J.S. Nat. Biotechnol. 31, 230–232 (2013).

    CAS  Article  Google Scholar 

  11. 11

    Qiu, Z. et al. Nucleic Acids Res. e120 (2013).

  12. 12

    Sung, Y.H. et al. Nat. Biotechnol. 31, 23–24 (2013).

    CAS  Article  Google Scholar 

  13. 13

    Cui, X. et al. Nat. Biotechnol. 29, 64–67 (2011).

    CAS  Article  Google Scholar 

  14. 14

    Tesson, L. et al. Nat. Biotechnol. 29, 695–696 (2011).

    CAS  Article  Google Scholar 

  15. 15

    Geurts, A.M. et al. Science 325, 433 (2009).

    CAS  Article  Google Scholar 

  16. 16

    Mul, J.D. et al. Obesity (Silver Spring) 20, 612–621 (2012).

    CAS  Article  Google Scholar 

  17. 17

    Fu, Y. et al. Nat. Biotechnol. advance online publication http://www.nature.com/doifinder/10.1038/nbt.2623 (23 June 2013).

Download references


We thank F. Zhang of the Broad Institute of MIT and Harvard for kindly providing us with the Cas9 expression vector. We thank S. Siwko for comments and advice. We also thank S.S. Bae and J.-S. Kim of Seoul National University for helping us to predict the potential off-target sites. This work was partially supported by grants from the State Key Development Programs of China (2012CB910400 to M.L., 2010CB945403 to D.L.), grants from the National Natural Science Foundation of China (31171318 to D.L. and 30930055 to M.L.), a grant from the Science and Technology Commission of Shanghai Municipality (11DZ2260300) and grants from the Program for Changjiang Scholars and Innovative Research Team in University (IRT1119 and IRT1128).

Author information



Corresponding authors

Correspondence to Dali Li or Yongxiang Zhao or Mingyao Liu.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Supplementary information

Supplementary Text and Figures

Supplementary Methods, Supplementary Figures 1–9 and Supplementary Table 1 (PDF 936 kb)

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Li, D., Qiu, Z., Shao, Y. et al. Heritable gene targeting in the mouse and rat using a CRISPR-Cas system. Nat Biotechnol 31, 681–683 (2013). https://doi.org/10.1038/nbt.2661

Download citation

Further reading


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing